diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,153327 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 2189,
+  "global_step": 21888,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 701.83056640625,
+      "learning_rate": 1.9999086257309942e-05,
+      "loss": 26.5271,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1200.2392578125,
+      "learning_rate": 1.9998172514619886e-05,
+      "loss": 26.7484,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1009.0261840820312,
+      "learning_rate": 1.9997258771929827e-05,
+      "loss": 25.6955,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2077.69677734375,
+      "learning_rate": 1.9996345029239767e-05,
+      "loss": 24.1503,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3223.77783203125,
+      "learning_rate": 1.9995431286549708e-05,
+      "loss": 24.3498,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1865.65673828125,
+      "learning_rate": 1.999451754385965e-05,
+      "loss": 25.3467,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 621.5952758789062,
+      "learning_rate": 1.9993603801169592e-05,
+      "loss": 23.6054,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2578.699951171875,
+      "learning_rate": 1.9992690058479533e-05,
+      "loss": 21.909,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 685.1784057617188,
+      "learning_rate": 1.9991776315789477e-05,
+      "loss": 21.5145,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 625.1463623046875,
+      "learning_rate": 1.9990862573099417e-05,
+      "loss": 20.4092,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1825.634033203125,
+      "learning_rate": 1.9989948830409358e-05,
+      "loss": 19.5223,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 522.1121215820312,
+      "learning_rate": 1.9989035087719298e-05,
+      "loss": 20.0322,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1119.1624755859375,
+      "learning_rate": 1.9988121345029242e-05,
+      "loss": 19.173,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 341.4478759765625,
+      "learning_rate": 1.9987207602339183e-05,
+      "loss": 18.2748,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1118.035888671875,
+      "learning_rate": 1.9986293859649123e-05,
+      "loss": 17.8722,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1043.7354736328125,
+      "learning_rate": 1.9985380116959067e-05,
+      "loss": 18.3362,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 623.2431030273438,
+      "learning_rate": 1.9984466374269008e-05,
+      "loss": 17.528,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 236.6895751953125,
+      "learning_rate": 1.9983552631578948e-05,
+      "loss": 16.753,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 225.5619659423828,
+      "learning_rate": 1.998263888888889e-05,
+      "loss": 16.9826,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 989.6156616210938,
+      "learning_rate": 1.9981725146198833e-05,
+      "loss": 16.4602,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 722.0059204101562,
+      "learning_rate": 1.9980811403508773e-05,
+      "loss": 16.8847,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 218.59686279296875,
+      "learning_rate": 1.9979897660818714e-05,
+      "loss": 16.8207,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 296.797119140625,
+      "learning_rate": 1.9978983918128657e-05,
+      "loss": 15.5995,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 323.1312255859375,
+      "learning_rate": 1.9978070175438598e-05,
+      "loss": 15.3953,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 423.7903747558594,
+      "learning_rate": 1.997715643274854e-05,
+      "loss": 15.238,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 617.6235961914062,
+      "learning_rate": 1.997624269005848e-05,
+      "loss": 15.6833,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 170.1925048828125,
+      "learning_rate": 1.9975328947368423e-05,
+      "loss": 14.6155,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 261.044921875,
+      "learning_rate": 1.9974415204678364e-05,
+      "loss": 14.628,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 369.1223449707031,
+      "learning_rate": 1.9973501461988304e-05,
+      "loss": 13.9809,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4648.13037109375,
+      "learning_rate": 1.9972587719298248e-05,
+      "loss": 14.7908,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1323.0887451171875,
+      "learning_rate": 1.997167397660819e-05,
+      "loss": 14.708,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 163.00051879882812,
+      "learning_rate": 1.997076023391813e-05,
+      "loss": 14.2975,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 199.43463134765625,
+      "learning_rate": 1.996984649122807e-05,
+      "loss": 14.1445,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 427.9976806640625,
+      "learning_rate": 1.9968932748538013e-05,
+      "loss": 14.0822,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 240.48971557617188,
+      "learning_rate": 1.9968019005847954e-05,
+      "loss": 13.5426,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 145.2312469482422,
+      "learning_rate": 1.9967105263157895e-05,
+      "loss": 13.0156,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 313.64337158203125,
+      "learning_rate": 1.996619152046784e-05,
+      "loss": 12.8278,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 266.9836120605469,
+      "learning_rate": 1.996527777777778e-05,
+      "loss": 13.3714,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 174.92787170410156,
+      "learning_rate": 1.996436403508772e-05,
+      "loss": 13.3778,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 477.0930480957031,
+      "learning_rate": 1.996345029239766e-05,
+      "loss": 13.0058,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1483.4581298828125,
+      "learning_rate": 1.9962536549707604e-05,
+      "loss": 12.7464,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 388.3049011230469,
+      "learning_rate": 1.9961622807017544e-05,
+      "loss": 12.4049,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 728.97119140625,
+      "learning_rate": 1.996070906432749e-05,
+      "loss": 12.3907,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 930.0480346679688,
+      "learning_rate": 1.995979532163743e-05,
+      "loss": 12.0784,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 273.0870666503906,
+      "learning_rate": 1.995888157894737e-05,
+      "loss": 12.2653,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2493.782470703125,
+      "learning_rate": 1.9957967836257313e-05,
+      "loss": 11.8293,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 657.7892456054688,
+      "learning_rate": 1.9957054093567254e-05,
+      "loss": 11.7462,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7545.38037109375,
+      "learning_rate": 1.9956140350877194e-05,
+      "loss": 11.876,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1491.529541015625,
+      "learning_rate": 1.9955226608187138e-05,
+      "loss": 12.2864,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 158.20066833496094,
+      "learning_rate": 1.995431286549708e-05,
+      "loss": 11.2565,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 156.17967224121094,
+      "learning_rate": 1.995339912280702e-05,
+      "loss": 11.4328,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 826.5393676757812,
+      "learning_rate": 1.995248538011696e-05,
+      "loss": 12.0166,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 291.9938659667969,
+      "learning_rate": 1.9951571637426904e-05,
+      "loss": 11.8023,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 169.12286376953125,
+      "learning_rate": 1.9950657894736844e-05,
+      "loss": 11.5029,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 203.3722686767578,
+      "learning_rate": 1.9949744152046785e-05,
+      "loss": 10.5545,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 136.9599609375,
+      "learning_rate": 1.994883040935673e-05,
+      "loss": 11.0836,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 76.5485610961914,
+      "learning_rate": 1.994791666666667e-05,
+      "loss": 10.5144,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 84.22417449951172,
+      "learning_rate": 1.994700292397661e-05,
+      "loss": 11.074,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 88.42478942871094,
+      "learning_rate": 1.994608918128655e-05,
+      "loss": 10.5162,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 144.1923370361328,
+      "learning_rate": 1.9945175438596494e-05,
+      "loss": 10.2883,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 176.57125854492188,
+      "learning_rate": 1.9944261695906435e-05,
+      "loss": 10.5996,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 199.21678161621094,
+      "learning_rate": 1.9943347953216375e-05,
+      "loss": 11.0329,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 69.54850006103516,
+      "learning_rate": 1.994243421052632e-05,
+      "loss": 9.9934,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 628.7960815429688,
+      "learning_rate": 1.994152046783626e-05,
+      "loss": 10.0064,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 84.74415588378906,
+      "learning_rate": 1.99406067251462e-05,
+      "loss": 9.931,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 135.67880249023438,
+      "learning_rate": 1.993969298245614e-05,
+      "loss": 10.3151,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 81.52942657470703,
+      "learning_rate": 1.9938779239766085e-05,
+      "loss": 9.3633,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 91.40180206298828,
+      "learning_rate": 1.9937865497076025e-05,
+      "loss": 9.821,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 100.1501235961914,
+      "learning_rate": 1.9936951754385966e-05,
+      "loss": 10.4403,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 72.85173797607422,
+      "learning_rate": 1.993603801169591e-05,
+      "loss": 9.6438,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 93.77415466308594,
+      "learning_rate": 1.993512426900585e-05,
+      "loss": 8.9354,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 97.62078094482422,
+      "learning_rate": 1.993421052631579e-05,
+      "loss": 8.8335,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 135.41534423828125,
+      "learning_rate": 1.993329678362573e-05,
+      "loss": 9.0866,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 60.743099212646484,
+      "learning_rate": 1.9932383040935675e-05,
+      "loss": 8.816,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 74.22317504882812,
+      "learning_rate": 1.9931469298245616e-05,
+      "loss": 8.5865,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 54.54646682739258,
+      "learning_rate": 1.9930555555555556e-05,
+      "loss": 8.433,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 72.51829528808594,
+      "learning_rate": 1.99296418128655e-05,
+      "loss": 8.8079,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 182.75042724609375,
+      "learning_rate": 1.992872807017544e-05,
+      "loss": 8.5358,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 65.38888549804688,
+      "learning_rate": 1.992781432748538e-05,
+      "loss": 9.0165,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 103.5543212890625,
+      "learning_rate": 1.9926900584795322e-05,
+      "loss": 7.7865,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 88.86284637451172,
+      "learning_rate": 1.9925986842105266e-05,
+      "loss": 7.9234,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 83.2920150756836,
+      "learning_rate": 1.9925073099415206e-05,
+      "loss": 8.2114,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 85.95154571533203,
+      "learning_rate": 1.9924159356725147e-05,
+      "loss": 7.8876,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 57.53984451293945,
+      "learning_rate": 1.992324561403509e-05,
+      "loss": 7.6025,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.21586608886719,
+      "learning_rate": 1.992233187134503e-05,
+      "loss": 7.6824,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 239.3819580078125,
+      "learning_rate": 1.9921418128654972e-05,
+      "loss": 7.6174,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.342735290527344,
+      "learning_rate": 1.9920504385964912e-05,
+      "loss": 8.1706,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.030391693115234,
+      "learning_rate": 1.9919590643274856e-05,
+      "loss": 7.5392,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.8525276184082,
+      "learning_rate": 1.9918676900584797e-05,
+      "loss": 7.6352,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.5462646484375,
+      "learning_rate": 1.9917763157894737e-05,
+      "loss": 7.3261,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.16826248168945,
+      "learning_rate": 1.991684941520468e-05,
+      "loss": 6.672,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.220985412597656,
+      "learning_rate": 1.991593567251462e-05,
+      "loss": 6.8084,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 142.97567749023438,
+      "learning_rate": 1.9915021929824562e-05,
+      "loss": 6.7216,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.219749450683594,
+      "learning_rate": 1.9914108187134503e-05,
+      "loss": 6.5632,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.457256317138672,
+      "learning_rate": 1.9913194444444447e-05,
+      "loss": 6.2604,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.317676544189453,
+      "learning_rate": 1.9912280701754387e-05,
+      "loss": 7.1737,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.08344078063965,
+      "learning_rate": 1.9911366959064328e-05,
+      "loss": 6.0266,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.90492630004883,
+      "learning_rate": 1.991045321637427e-05,
+      "loss": 6.5628,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.920045852661133,
+      "learning_rate": 1.9909539473684212e-05,
+      "loss": 6.5069,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 28.583229064941406,
+      "learning_rate": 1.9908625730994153e-05,
+      "loss": 7.0418,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.005748748779297,
+      "learning_rate": 1.9907711988304093e-05,
+      "loss": 6.0808,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.25435733795166,
+      "learning_rate": 1.9906798245614037e-05,
+      "loss": 6.501,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 113.401123046875,
+      "learning_rate": 1.9905884502923978e-05,
+      "loss": 6.1308,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.58757781982422,
+      "learning_rate": 1.9904970760233918e-05,
+      "loss": 5.7499,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.126625061035156,
+      "learning_rate": 1.9904057017543862e-05,
+      "loss": 5.6598,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.423065185546875,
+      "learning_rate": 1.9903143274853803e-05,
+      "loss": 5.2569,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.25627326965332,
+      "learning_rate": 1.9902229532163743e-05,
+      "loss": 5.3455,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.145015716552734,
+      "learning_rate": 1.9901315789473684e-05,
+      "loss": 5.522,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.936610221862793,
+      "learning_rate": 1.9900402046783628e-05,
+      "loss": 6.0443,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.364330291748047,
+      "learning_rate": 1.9899488304093568e-05,
+      "loss": 4.8274,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.064719200134277,
+      "learning_rate": 1.989857456140351e-05,
+      "loss": 5.4787,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.681669235229492,
+      "learning_rate": 1.9897660818713453e-05,
+      "loss": 5.0013,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.721562385559082,
+      "learning_rate": 1.9896747076023393e-05,
+      "loss": 5.8369,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.113441467285156,
+      "learning_rate": 1.9895833333333334e-05,
+      "loss": 5.2118,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.3292875289917,
+      "learning_rate": 1.9894919590643274e-05,
+      "loss": 4.9024,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.637796401977539,
+      "learning_rate": 1.9894005847953218e-05,
+      "loss": 4.885,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.46586036682129,
+      "learning_rate": 1.989309210526316e-05,
+      "loss": 5.8942,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.034916877746582,
+      "learning_rate": 1.98921783625731e-05,
+      "loss": 5.0978,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.645183563232422,
+      "learning_rate": 1.9891264619883043e-05,
+      "loss": 5.6964,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.670615196228027,
+      "learning_rate": 1.9890350877192984e-05,
+      "loss": 4.635,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.354955673217773,
+      "learning_rate": 1.9889437134502924e-05,
+      "loss": 4.6358,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.55729866027832,
+      "learning_rate": 1.9888523391812865e-05,
+      "loss": 4.8596,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.402263641357422,
+      "learning_rate": 1.988760964912281e-05,
+      "loss": 5.4823,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.1137113571167,
+      "learning_rate": 1.988669590643275e-05,
+      "loss": 5.2659,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.45231819152832,
+      "learning_rate": 1.988578216374269e-05,
+      "loss": 4.6422,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.504097938537598,
+      "learning_rate": 1.9884868421052633e-05,
+      "loss": 4.2845,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.255314826965332,
+      "learning_rate": 1.9883954678362574e-05,
+      "loss": 4.2374,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.717175483703613,
+      "learning_rate": 1.9883040935672515e-05,
+      "loss": 4.4061,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.204925060272217,
+      "learning_rate": 1.988212719298246e-05,
+      "loss": 4.9472,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.09031677246094,
+      "learning_rate": 1.98812134502924e-05,
+      "loss": 5.1678,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.497823715209961,
+      "learning_rate": 1.9880299707602343e-05,
+      "loss": 4.1907,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.823950290679932,
+      "learning_rate": 1.9879385964912283e-05,
+      "loss": 4.3924,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 44.2364387512207,
+      "learning_rate": 1.9878472222222224e-05,
+      "loss": 4.0112,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.884017467498779,
+      "learning_rate": 1.9877558479532168e-05,
+      "loss": 4.1533,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.491844177246094,
+      "learning_rate": 1.987664473684211e-05,
+      "loss": 4.0735,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.887130737304688,
+      "learning_rate": 1.987573099415205e-05,
+      "loss": 4.3003,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.853724479675293,
+      "learning_rate": 1.987481725146199e-05,
+      "loss": 3.9997,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 38.70127868652344,
+      "learning_rate": 1.9873903508771933e-05,
+      "loss": 4.6855,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.22600269317627,
+      "learning_rate": 1.9872989766081874e-05,
+      "loss": 4.8312,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.204143524169922,
+      "learning_rate": 1.9872076023391814e-05,
+      "loss": 4.0035,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.53924560546875,
+      "learning_rate": 1.9871162280701758e-05,
+      "loss": 3.9283,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.909905433654785,
+      "learning_rate": 1.98702485380117e-05,
+      "loss": 4.6494,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.622196197509766,
+      "learning_rate": 1.986933479532164e-05,
+      "loss": 4.283,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.451993465423584,
+      "learning_rate": 1.986842105263158e-05,
+      "loss": 4.325,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.205716133117676,
+      "learning_rate": 1.9867507309941524e-05,
+      "loss": 4.1427,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.988460540771484,
+      "learning_rate": 1.9866593567251464e-05,
+      "loss": 5.0233,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.678100109100342,
+      "learning_rate": 1.9865679824561405e-05,
+      "loss": 4.1243,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.146218299865723,
+      "learning_rate": 1.986476608187135e-05,
+      "loss": 4.157,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.47618579864502,
+      "learning_rate": 1.986385233918129e-05,
+      "loss": 4.3721,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.45893669128418,
+      "learning_rate": 1.986293859649123e-05,
+      "loss": 3.7447,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.511500358581543,
+      "learning_rate": 1.986202485380117e-05,
+      "loss": 4.4661,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.312541961669922,
+      "learning_rate": 1.9861111111111114e-05,
+      "loss": 4.3467,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.143101692199707,
+      "learning_rate": 1.9860197368421055e-05,
+      "loss": 4.2233,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.372185707092285,
+      "learning_rate": 1.9859283625730995e-05,
+      "loss": 4.5257,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.525723457336426,
+      "learning_rate": 1.985836988304094e-05,
+      "loss": 4.3266,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.03261947631836,
+      "learning_rate": 1.985745614035088e-05,
+      "loss": 4.2687,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.564573287963867,
+      "learning_rate": 1.985654239766082e-05,
+      "loss": 4.2196,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.399077892303467,
+      "learning_rate": 1.985562865497076e-05,
+      "loss": 4.3326,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.482998371124268,
+      "learning_rate": 1.9854714912280705e-05,
+      "loss": 4.7781,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.840038776397705,
+      "learning_rate": 1.9853801169590645e-05,
+      "loss": 3.7994,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.979851722717285,
+      "learning_rate": 1.9852887426900586e-05,
+      "loss": 3.1431,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6285271644592285,
+      "learning_rate": 1.985197368421053e-05,
+      "loss": 4.5495,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.800195693969727,
+      "learning_rate": 1.985105994152047e-05,
+      "loss": 3.5296,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.043542861938477,
+      "learning_rate": 1.985014619883041e-05,
+      "loss": 3.7554,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.695467472076416,
+      "learning_rate": 1.984923245614035e-05,
+      "loss": 4.9134,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.650739669799805,
+      "learning_rate": 1.9848318713450295e-05,
+      "loss": 3.5126,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.071653842926025,
+      "learning_rate": 1.9847404970760236e-05,
+      "loss": 3.5121,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.898786544799805,
+      "learning_rate": 1.9846491228070176e-05,
+      "loss": 3.8098,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.306551933288574,
+      "learning_rate": 1.984557748538012e-05,
+      "loss": 3.3916,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.225088119506836,
+      "learning_rate": 1.984466374269006e-05,
+      "loss": 3.8855,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.202965259552002,
+      "learning_rate": 1.984375e-05,
+      "loss": 3.4288,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.124242782592773,
+      "learning_rate": 1.9842836257309942e-05,
+      "loss": 4.0643,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.579541206359863,
+      "learning_rate": 1.9841922514619886e-05,
+      "loss": 3.2839,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.897547245025635,
+      "learning_rate": 1.9841008771929826e-05,
+      "loss": 3.9415,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.866248607635498,
+      "learning_rate": 1.9840095029239767e-05,
+      "loss": 3.8862,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 36.84048843383789,
+      "learning_rate": 1.983918128654971e-05,
+      "loss": 4.441,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.265117645263672,
+      "learning_rate": 1.983826754385965e-05,
+      "loss": 4.0433,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.287665367126465,
+      "learning_rate": 1.9837353801169592e-05,
+      "loss": 4.0286,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.007824420928955,
+      "learning_rate": 1.9836440058479532e-05,
+      "loss": 3.8144,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.272512912750244,
+      "learning_rate": 1.9835526315789476e-05,
+      "loss": 3.2789,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.081605911254883,
+      "learning_rate": 1.9834612573099417e-05,
+      "loss": 4.1524,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.091513633728027,
+      "learning_rate": 1.9833698830409357e-05,
+      "loss": 4.3063,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.41347599029541,
+      "learning_rate": 1.98327850877193e-05,
+      "loss": 3.7745,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.287306308746338,
+      "learning_rate": 1.983187134502924e-05,
+      "loss": 3.9965,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.184671401977539,
+      "learning_rate": 1.9830957602339182e-05,
+      "loss": 3.6561,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.290895938873291,
+      "learning_rate": 1.9830043859649123e-05,
+      "loss": 4.131,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.902993202209473,
+      "learning_rate": 1.9829130116959067e-05,
+      "loss": 4.3937,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 40.11512756347656,
+      "learning_rate": 1.9828216374269007e-05,
+      "loss": 3.9411,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.744055271148682,
+      "learning_rate": 1.9827302631578948e-05,
+      "loss": 3.8642,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.112586498260498,
+      "learning_rate": 1.982638888888889e-05,
+      "loss": 3.5025,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.545804023742676,
+      "learning_rate": 1.9825475146198832e-05,
+      "loss": 3.2997,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.986462593078613,
+      "learning_rate": 1.9824561403508773e-05,
+      "loss": 4.3294,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 144.84519958496094,
+      "learning_rate": 1.9823647660818713e-05,
+      "loss": 3.4648,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.959860324859619,
+      "learning_rate": 1.9822733918128657e-05,
+      "loss": 3.1202,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.575424194335938,
+      "learning_rate": 1.9821820175438598e-05,
+      "loss": 3.5466,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.180412769317627,
+      "learning_rate": 1.9820906432748538e-05,
+      "loss": 4.0408,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.927687644958496,
+      "learning_rate": 1.9819992690058482e-05,
+      "loss": 3.8454,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.450251579284668,
+      "learning_rate": 1.9819078947368423e-05,
+      "loss": 3.6041,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.439291477203369,
+      "learning_rate": 1.9818165204678363e-05,
+      "loss": 3.7548,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.258025169372559,
+      "learning_rate": 1.9817251461988304e-05,
+      "loss": 3.9363,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.24390411376953,
+      "learning_rate": 1.9816337719298248e-05,
+      "loss": 4.0805,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.970691204071045,
+      "learning_rate": 1.9815423976608188e-05,
+      "loss": 4.1865,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.895434379577637,
+      "learning_rate": 1.981451023391813e-05,
+      "loss": 3.7401,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.714330673217773,
+      "learning_rate": 1.9813596491228073e-05,
+      "loss": 3.3638,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4424943923950195,
+      "learning_rate": 1.9812682748538013e-05,
+      "loss": 3.0758,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.318918228149414,
+      "learning_rate": 1.9811769005847954e-05,
+      "loss": 3.1671,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.781872272491455,
+      "learning_rate": 1.9810855263157894e-05,
+      "loss": 3.8111,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.580051422119141,
+      "learning_rate": 1.9809941520467838e-05,
+      "loss": 3.6723,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.295648574829102,
+      "learning_rate": 1.980902777777778e-05,
+      "loss": 3.388,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.670519828796387,
+      "learning_rate": 1.980811403508772e-05,
+      "loss": 3.7016,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.83279037475586,
+      "learning_rate": 1.9807200292397663e-05,
+      "loss": 3.6307,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.685090065002441,
+      "learning_rate": 1.9806286549707604e-05,
+      "loss": 2.84,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.360495090484619,
+      "learning_rate": 1.9805372807017544e-05,
+      "loss": 3.6435,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.979997158050537,
+      "learning_rate": 1.9804459064327488e-05,
+      "loss": 3.2217,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.93638801574707,
+      "learning_rate": 1.980354532163743e-05,
+      "loss": 3.7364,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.778952598571777,
+      "learning_rate": 1.9802631578947372e-05,
+      "loss": 3.5462,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 61.82352066040039,
+      "learning_rate": 1.9801717836257313e-05,
+      "loss": 3.2641,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.050652980804443,
+      "learning_rate": 1.9800804093567253e-05,
+      "loss": 3.4419,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.222171783447266,
+      "learning_rate": 1.9799890350877194e-05,
+      "loss": 3.3621,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.589492321014404,
+      "learning_rate": 1.9798976608187138e-05,
+      "loss": 3.5176,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.41381549835205,
+      "learning_rate": 1.979806286549708e-05,
+      "loss": 3.4549,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.29306697845459,
+      "learning_rate": 1.979714912280702e-05,
+      "loss": 4.3479,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.587116241455078,
+      "learning_rate": 1.9796235380116963e-05,
+      "loss": 3.3659,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.63404655456543,
+      "learning_rate": 1.9795321637426903e-05,
+      "loss": 3.7832,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.391140937805176,
+      "learning_rate": 1.9794407894736844e-05,
+      "loss": 3.6167,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.925449848175049,
+      "learning_rate": 1.9793494152046784e-05,
+      "loss": 3.6021,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.100320339202881,
+      "learning_rate": 1.979258040935673e-05,
+      "loss": 3.7429,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.877830505371094,
+      "learning_rate": 1.979166666666667e-05,
+      "loss": 3.2923,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6128668785095215,
+      "learning_rate": 1.979075292397661e-05,
+      "loss": 3.3027,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.774679183959961,
+      "learning_rate": 1.9789839181286553e-05,
+      "loss": 4.0151,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.570647716522217,
+      "learning_rate": 1.9788925438596494e-05,
+      "loss": 3.5202,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.065339088439941,
+      "learning_rate": 1.9788011695906434e-05,
+      "loss": 3.1107,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.640481472015381,
+      "learning_rate": 1.9787097953216375e-05,
+      "loss": 3.7034,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.283030033111572,
+      "learning_rate": 1.978618421052632e-05,
+      "loss": 3.5194,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.714503288269043,
+      "learning_rate": 1.978527046783626e-05,
+      "loss": 4.2111,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.129957675933838,
+      "learning_rate": 1.97843567251462e-05,
+      "loss": 3.3454,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.56744909286499,
+      "learning_rate": 1.9783442982456144e-05,
+      "loss": 3.495,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.399701118469238,
+      "learning_rate": 1.9782529239766084e-05,
+      "loss": 3.2964,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.141587734222412,
+      "learning_rate": 1.9781615497076025e-05,
+      "loss": 3.8428,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.038739204406738,
+      "learning_rate": 1.9780701754385965e-05,
+      "loss": 3.4585,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.821687698364258,
+      "learning_rate": 1.977978801169591e-05,
+      "loss": 3.9688,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.458168029785156,
+      "learning_rate": 1.977887426900585e-05,
+      "loss": 3.8464,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5242462158203125,
+      "learning_rate": 1.977796052631579e-05,
+      "loss": 4.3641,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.950908660888672,
+      "learning_rate": 1.9777046783625734e-05,
+      "loss": 3.0835,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.411548614501953,
+      "learning_rate": 1.9776133040935675e-05,
+      "loss": 3.0273,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.073409557342529,
+      "learning_rate": 1.9775219298245615e-05,
+      "loss": 3.5751,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.594972133636475,
+      "learning_rate": 1.9774305555555556e-05,
+      "loss": 3.6512,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.357919692993164,
+      "learning_rate": 1.97733918128655e-05,
+      "loss": 3.3407,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.676962852478027,
+      "learning_rate": 1.977247807017544e-05,
+      "loss": 3.4543,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.49699068069458,
+      "learning_rate": 1.977156432748538e-05,
+      "loss": 3.8254,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.475175857543945,
+      "learning_rate": 1.9770650584795325e-05,
+      "loss": 3.4879,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.363021373748779,
+      "learning_rate": 1.9769736842105265e-05,
+      "loss": 3.4308,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6790771484375,
+      "learning_rate": 1.9768823099415206e-05,
+      "loss": 3.073,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.767417907714844,
+      "learning_rate": 1.9767909356725146e-05,
+      "loss": 3.1013,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.206299781799316,
+      "learning_rate": 1.976699561403509e-05,
+      "loss": 3.6468,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.512086391448975,
+      "learning_rate": 1.976608187134503e-05,
+      "loss": 3.0704,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.705033302307129,
+      "learning_rate": 1.976516812865497e-05,
+      "loss": 3.1298,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.925885200500488,
+      "learning_rate": 1.9764254385964915e-05,
+      "loss": 3.095,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.9268975257873535,
+      "learning_rate": 1.9763340643274856e-05,
+      "loss": 3.1744,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.818660736083984,
+      "learning_rate": 1.9762426900584796e-05,
+      "loss": 3.2069,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.300166606903076,
+      "learning_rate": 1.9761513157894737e-05,
+      "loss": 3.4016,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.3915252685546875,
+      "learning_rate": 1.976059941520468e-05,
+      "loss": 3.24,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.176346778869629,
+      "learning_rate": 1.975968567251462e-05,
+      "loss": 2.4937,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.932516574859619,
+      "learning_rate": 1.9758771929824562e-05,
+      "loss": 3.058,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.632996559143066,
+      "learning_rate": 1.9757858187134506e-05,
+      "loss": 3.3244,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.640111446380615,
+      "learning_rate": 1.9756944444444446e-05,
+      "loss": 2.9498,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.422543048858643,
+      "learning_rate": 1.9756030701754387e-05,
+      "loss": 3.5379,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.197525024414062,
+      "learning_rate": 1.9755116959064327e-05,
+      "loss": 2.7539,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.27121114730835,
+      "learning_rate": 1.975420321637427e-05,
+      "loss": 3.2346,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2417378425598145,
+      "learning_rate": 1.975328947368421e-05,
+      "loss": 3.3781,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.248963356018066,
+      "learning_rate": 1.9752375730994152e-05,
+      "loss": 3.4469,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.877746105194092,
+      "learning_rate": 1.9751461988304096e-05,
+      "loss": 3.046,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.440988063812256,
+      "learning_rate": 1.9750548245614037e-05,
+      "loss": 2.8763,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 288.98590087890625,
+      "learning_rate": 1.9749634502923977e-05,
+      "loss": 3.1553,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.371086120605469,
+      "learning_rate": 1.9748720760233918e-05,
+      "loss": 2.708,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.874140739440918,
+      "learning_rate": 1.974780701754386e-05,
+      "loss": 3.8295,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.718597888946533,
+      "learning_rate": 1.9746893274853802e-05,
+      "loss": 2.4922,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.550393104553223,
+      "learning_rate": 1.9745979532163743e-05,
+      "loss": 2.9043,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.715962886810303,
+      "learning_rate": 1.9745065789473687e-05,
+      "loss": 2.8081,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.642818450927734,
+      "learning_rate": 1.9744152046783627e-05,
+      "loss": 2.9363,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.299395561218262,
+      "learning_rate": 1.9743238304093568e-05,
+      "loss": 3.6243,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.45442533493042,
+      "learning_rate": 1.9742324561403508e-05,
+      "loss": 3.4569,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.40582799911499,
+      "learning_rate": 1.9741410818713452e-05,
+      "loss": 2.7873,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1408233642578125,
+      "learning_rate": 1.9740497076023393e-05,
+      "loss": 3.3256,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 49.20579528808594,
+      "learning_rate": 1.9739583333333333e-05,
+      "loss": 2.7399,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 76.68869018554688,
+      "learning_rate": 1.9738669590643277e-05,
+      "loss": 3.1363,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.037979602813721,
+      "learning_rate": 1.9737755847953218e-05,
+      "loss": 2.5785,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.954207897186279,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 2.8123,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.609184741973877,
+      "learning_rate": 1.97359283625731e-05,
+      "loss": 2.9975,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.424068927764893,
+      "learning_rate": 1.9735014619883043e-05,
+      "loss": 3.1003,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.72855758666992,
+      "learning_rate": 1.9734100877192983e-05,
+      "loss": 2.8038,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.438612461090088,
+      "learning_rate": 1.9733187134502924e-05,
+      "loss": 2.9442,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.934460639953613,
+      "learning_rate": 1.9732273391812868e-05,
+      "loss": 3.738,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.2489013671875,
+      "learning_rate": 1.9731359649122808e-05,
+      "loss": 3.9071,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.256026744842529,
+      "learning_rate": 1.973044590643275e-05,
+      "loss": 2.5559,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.560751438140869,
+      "learning_rate": 1.972953216374269e-05,
+      "loss": 3.3329,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 446.42303466796875,
+      "learning_rate": 1.9728618421052633e-05,
+      "loss": 2.8189,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.879525661468506,
+      "learning_rate": 1.9727704678362574e-05,
+      "loss": 2.8429,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.175571918487549,
+      "learning_rate": 1.9726790935672514e-05,
+      "loss": 3.2541,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.522804260253906,
+      "learning_rate": 1.9725877192982458e-05,
+      "loss": 2.6757,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.326038837432861,
+      "learning_rate": 1.9724963450292402e-05,
+      "loss": 2.6143,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.076855659484863,
+      "learning_rate": 1.9724049707602342e-05,
+      "loss": 3.1676,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.015915870666504,
+      "learning_rate": 1.9723135964912283e-05,
+      "loss": 3.0943,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.830132961273193,
+      "learning_rate": 1.9722222222222224e-05,
+      "loss": 2.9657,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.756165027618408,
+      "learning_rate": 1.9721308479532167e-05,
+      "loss": 2.9617,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 241.1164093017578,
+      "learning_rate": 1.9720394736842108e-05,
+      "loss": 3.3419,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.916194438934326,
+      "learning_rate": 1.971948099415205e-05,
+      "loss": 3.5704,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.941102981567383,
+      "learning_rate": 1.9718567251461992e-05,
+      "loss": 3.3643,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.602930068969727,
+      "learning_rate": 1.9717653508771933e-05,
+      "loss": 2.7272,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.0161285400390625,
+      "learning_rate": 1.9716739766081873e-05,
+      "loss": 2.8085,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.357957363128662,
+      "learning_rate": 1.9715826023391814e-05,
+      "loss": 2.8383,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 55.1662712097168,
+      "learning_rate": 1.9714912280701758e-05,
+      "loss": 2.9549,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.483248233795166,
+      "learning_rate": 1.97139985380117e-05,
+      "loss": 3.2391,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.943034648895264,
+      "learning_rate": 1.971308479532164e-05,
+      "loss": 3.0267,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.580035209655762,
+      "learning_rate": 1.9712171052631583e-05,
+      "loss": 3.1299,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.502040386199951,
+      "learning_rate": 1.9711257309941523e-05,
+      "loss": 2.7534,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.106404781341553,
+      "learning_rate": 1.9710343567251464e-05,
+      "loss": 3.1761,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.829550266265869,
+      "learning_rate": 1.9709429824561404e-05,
+      "loss": 2.9514,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.744872570037842,
+      "learning_rate": 1.970851608187135e-05,
+      "loss": 2.9684,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.841232776641846,
+      "learning_rate": 1.970760233918129e-05,
+      "loss": 2.9579,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.052304744720459,
+      "learning_rate": 1.970668859649123e-05,
+      "loss": 2.6828,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.731119632720947,
+      "learning_rate": 1.9705774853801173e-05,
+      "loss": 2.9526,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.191049098968506,
+      "learning_rate": 1.9704861111111114e-05,
+      "loss": 2.7007,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.264830589294434,
+      "learning_rate": 1.9703947368421054e-05,
+      "loss": 2.468,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.901215553283691,
+      "learning_rate": 1.9703033625730995e-05,
+      "loss": 3.192,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.332241058349609,
+      "learning_rate": 1.970211988304094e-05,
+      "loss": 2.6811,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.972348213195801,
+      "learning_rate": 1.970120614035088e-05,
+      "loss": 2.7548,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.208010196685791,
+      "learning_rate": 1.970029239766082e-05,
+      "loss": 2.5362,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.540053844451904,
+      "learning_rate": 1.9699378654970764e-05,
+      "loss": 2.6402,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.854733943939209,
+      "learning_rate": 1.9698464912280704e-05,
+      "loss": 2.5866,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.82240104675293,
+      "learning_rate": 1.9697551169590645e-05,
+      "loss": 2.7328,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.931463718414307,
+      "learning_rate": 1.9696637426900585e-05,
+      "loss": 2.9677,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.276640892028809,
+      "learning_rate": 1.969572368421053e-05,
+      "loss": 2.7238,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.499929428100586,
+      "learning_rate": 1.969480994152047e-05,
+      "loss": 3.0597,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 39.670249938964844,
+      "learning_rate": 1.969389619883041e-05,
+      "loss": 2.993,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.631933212280273,
+      "learning_rate": 1.9692982456140354e-05,
+      "loss": 3.2801,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.263147354125977,
+      "learning_rate": 1.9692068713450295e-05,
+      "loss": 3.0872,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.906842231750488,
+      "learning_rate": 1.9691154970760235e-05,
+      "loss": 2.6511,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.299118518829346,
+      "learning_rate": 1.9690241228070176e-05,
+      "loss": 2.3918,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.60709285736084,
+      "learning_rate": 1.968932748538012e-05,
+      "loss": 2.9502,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.624070167541504,
+      "learning_rate": 1.968841374269006e-05,
+      "loss": 2.5968,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 112.12809753417969,
+      "learning_rate": 1.96875e-05,
+      "loss": 3.1559,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.4475836753845215,
+      "learning_rate": 1.9686586257309945e-05,
+      "loss": 2.811,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.689187049865723,
+      "learning_rate": 1.9685672514619885e-05,
+      "loss": 3.7016,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.184946060180664,
+      "learning_rate": 1.9684758771929826e-05,
+      "loss": 2.2603,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.49462366104126,
+      "learning_rate": 1.9683845029239766e-05,
+      "loss": 3.3029,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.3217339515686035,
+      "learning_rate": 1.968293128654971e-05,
+      "loss": 3.0904,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.889744758605957,
+      "learning_rate": 1.968201754385965e-05,
+      "loss": 2.9266,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.916393280029297,
+      "learning_rate": 1.968110380116959e-05,
+      "loss": 3.6065,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.34841251373291,
+      "learning_rate": 1.9680190058479535e-05,
+      "loss": 2.8698,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.797161102294922,
+      "learning_rate": 1.9679276315789476e-05,
+      "loss": 2.5798,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.234203815460205,
+      "learning_rate": 1.9678362573099416e-05,
+      "loss": 2.7233,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.587125778198242,
+      "learning_rate": 1.9677448830409357e-05,
+      "loss": 3.0554,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.299403667449951,
+      "learning_rate": 1.96765350877193e-05,
+      "loss": 3.1106,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.419379234313965,
+      "learning_rate": 1.967562134502924e-05,
+      "loss": 3.0193,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.33437442779541,
+      "learning_rate": 1.9674707602339182e-05,
+      "loss": 2.7062,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.741194248199463,
+      "learning_rate": 1.9673793859649126e-05,
+      "loss": 2.8727,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.726256370544434,
+      "learning_rate": 1.9672880116959066e-05,
+      "loss": 2.3495,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.925276279449463,
+      "learning_rate": 1.9671966374269007e-05,
+      "loss": 2.5355,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.571691036224365,
+      "learning_rate": 1.9671052631578947e-05,
+      "loss": 2.7578,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.353464603424072,
+      "learning_rate": 1.967013888888889e-05,
+      "loss": 3.173,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.97727632522583,
+      "learning_rate": 1.966922514619883e-05,
+      "loss": 3.2094,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.645294189453125,
+      "learning_rate": 1.9668311403508772e-05,
+      "loss": 3.0737,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.198105812072754,
+      "learning_rate": 1.9667397660818716e-05,
+      "loss": 3.0111,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 237.36715698242188,
+      "learning_rate": 1.9666483918128657e-05,
+      "loss": 2.3141,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.555069446563721,
+      "learning_rate": 1.9665570175438597e-05,
+      "loss": 3.5556,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.799703598022461,
+      "learning_rate": 1.9664656432748538e-05,
+      "loss": 2.5052,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.250219821929932,
+      "learning_rate": 1.966374269005848e-05,
+      "loss": 2.7811,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.056912899017334,
+      "learning_rate": 1.9662828947368422e-05,
+      "loss": 2.9215,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.806487083435059,
+      "learning_rate": 1.9661915204678363e-05,
+      "loss": 3.0364,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.021186828613281,
+      "learning_rate": 1.9661001461988307e-05,
+      "loss": 3.0934,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.308094501495361,
+      "learning_rate": 1.9660087719298247e-05,
+      "loss": 3.4168,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.395194053649902,
+      "learning_rate": 1.9659173976608188e-05,
+      "loss": 2.6633,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.243008136749268,
+      "learning_rate": 1.9658260233918128e-05,
+      "loss": 2.9391,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.595361232757568,
+      "learning_rate": 1.9657346491228072e-05,
+      "loss": 2.7104,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.991573333740234,
+      "learning_rate": 1.9656432748538013e-05,
+      "loss": 2.7148,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.574895858764648,
+      "learning_rate": 1.9655519005847953e-05,
+      "loss": 2.6762,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.425365447998047,
+      "learning_rate": 1.9654605263157897e-05,
+      "loss": 2.7347,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.827467441558838,
+      "learning_rate": 1.9653691520467838e-05,
+      "loss": 2.6102,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.848750114440918,
+      "learning_rate": 1.9652777777777778e-05,
+      "loss": 2.5724,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.755354404449463,
+      "learning_rate": 1.965186403508772e-05,
+      "loss": 3.2623,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.142520904541016,
+      "learning_rate": 1.9650950292397663e-05,
+      "loss": 2.693,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.441346168518066,
+      "learning_rate": 1.9650036549707603e-05,
+      "loss": 2.5888,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.483672618865967,
+      "learning_rate": 1.9649122807017544e-05,
+      "loss": 2.752,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.001496315002441,
+      "learning_rate": 1.9648209064327488e-05,
+      "loss": 2.8665,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.7258195877075195,
+      "learning_rate": 1.9647295321637428e-05,
+      "loss": 2.7674,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.286237716674805,
+      "learning_rate": 1.9646381578947372e-05,
+      "loss": 3.1907,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.505671501159668,
+      "learning_rate": 1.9645467836257313e-05,
+      "loss": 3.0557,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.200510025024414,
+      "learning_rate": 1.9644554093567253e-05,
+      "loss": 2.8646,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9690544605255127,
+      "learning_rate": 1.9643640350877197e-05,
+      "loss": 2.4839,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.873897075653076,
+      "learning_rate": 1.9642726608187137e-05,
+      "loss": 2.7518,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.900351047515869,
+      "learning_rate": 1.9641812865497078e-05,
+      "loss": 3.0999,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.271961688995361,
+      "learning_rate": 1.964089912280702e-05,
+      "loss": 2.3854,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.035788059234619,
+      "learning_rate": 1.9639985380116962e-05,
+      "loss": 2.998,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.99265193939209,
+      "learning_rate": 1.9639071637426903e-05,
+      "loss": 2.2967,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.31301736831665,
+      "learning_rate": 1.9638157894736844e-05,
+      "loss": 2.5632,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.491322040557861,
+      "learning_rate": 1.9637244152046787e-05,
+      "loss": 2.2798,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.222789287567139,
+      "learning_rate": 1.9636330409356728e-05,
+      "loss": 2.7219,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.3516411781311035,
+      "learning_rate": 1.963541666666667e-05,
+      "loss": 2.8431,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.126443386077881,
+      "learning_rate": 1.963450292397661e-05,
+      "loss": 2.6267,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.083006858825684,
+      "learning_rate": 1.9633589181286553e-05,
+      "loss": 3.3718,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.604710578918457,
+      "learning_rate": 1.9632675438596493e-05,
+      "loss": 2.8671,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.865631103515625,
+      "learning_rate": 1.9631761695906434e-05,
+      "loss": 2.5089,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.647823810577393,
+      "learning_rate": 1.9630847953216378e-05,
+      "loss": 2.1846,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5284743309021,
+      "learning_rate": 1.962993421052632e-05,
+      "loss": 3.1125,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7988953590393066,
+      "learning_rate": 1.962902046783626e-05,
+      "loss": 2.4734,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.227822780609131,
+      "learning_rate": 1.96281067251462e-05,
+      "loss": 2.8242,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.366757869720459,
+      "learning_rate": 1.9627192982456143e-05,
+      "loss": 2.3924,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.720643520355225,
+      "learning_rate": 1.9626279239766084e-05,
+      "loss": 2.7471,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.600627422332764,
+      "learning_rate": 1.9625365497076024e-05,
+      "loss": 2.3123,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.650872707366943,
+      "learning_rate": 1.962445175438597e-05,
+      "loss": 2.7479,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.319598197937012,
+      "learning_rate": 1.962353801169591e-05,
+      "loss": 2.3772,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.218515396118164,
+      "learning_rate": 1.962262426900585e-05,
+      "loss": 2.3044,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.737773895263672,
+      "learning_rate": 1.962171052631579e-05,
+      "loss": 2.9196,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.187772274017334,
+      "learning_rate": 1.9620796783625734e-05,
+      "loss": 2.7344,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.082683563232422,
+      "learning_rate": 1.9619883040935674e-05,
+      "loss": 2.4471,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.861405372619629,
+      "learning_rate": 1.9618969298245615e-05,
+      "loss": 2.6142,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.603705406188965,
+      "learning_rate": 1.961805555555556e-05,
+      "loss": 2.7267,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.602875709533691,
+      "learning_rate": 1.96171418128655e-05,
+      "loss": 2.5339,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.97728157043457,
+      "learning_rate": 1.961622807017544e-05,
+      "loss": 2.6466,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.333723068237305,
+      "learning_rate": 1.961531432748538e-05,
+      "loss": 3.1901,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.458555221557617,
+      "learning_rate": 1.9614400584795324e-05,
+      "loss": 2.8775,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.334867477416992,
+      "learning_rate": 1.9613486842105265e-05,
+      "loss": 3.1638,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.815124034881592,
+      "learning_rate": 1.9612573099415205e-05,
+      "loss": 2.3771,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.994760274887085,
+      "learning_rate": 1.961165935672515e-05,
+      "loss": 2.5377,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.469785690307617,
+      "learning_rate": 1.961074561403509e-05,
+      "loss": 2.9703,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.10971212387085,
+      "learning_rate": 1.960983187134503e-05,
+      "loss": 2.8749,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.495964050292969,
+      "learning_rate": 1.960891812865497e-05,
+      "loss": 2.7875,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.26600980758667,
+      "learning_rate": 1.9608004385964915e-05,
+      "loss": 3.0774,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8729071617126465,
+      "learning_rate": 1.9607090643274855e-05,
+      "loss": 1.9917,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5201096534729,
+      "learning_rate": 1.9606176900584796e-05,
+      "loss": 2.5463,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.886353015899658,
+      "learning_rate": 1.960526315789474e-05,
+      "loss": 2.8336,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.621831893920898,
+      "learning_rate": 1.960434941520468e-05,
+      "loss": 2.8551,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.401886940002441,
+      "learning_rate": 1.960343567251462e-05,
+      "loss": 3.1916,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.768064022064209,
+      "learning_rate": 1.960252192982456e-05,
+      "loss": 2.9857,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.65527868270874,
+      "learning_rate": 1.9601608187134505e-05,
+      "loss": 2.4608,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7586448192596436,
+      "learning_rate": 1.9600694444444446e-05,
+      "loss": 2.356,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.426208972930908,
+      "learning_rate": 1.9599780701754386e-05,
+      "loss": 2.5356,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.758132457733154,
+      "learning_rate": 1.959886695906433e-05,
+      "loss": 2.8737,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.107352256774902,
+      "learning_rate": 1.959795321637427e-05,
+      "loss": 2.6076,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.891287326812744,
+      "learning_rate": 1.959703947368421e-05,
+      "loss": 2.6493,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0016069412231445,
+      "learning_rate": 1.9596125730994152e-05,
+      "loss": 2.4432,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.817814826965332,
+      "learning_rate": 1.9595211988304096e-05,
+      "loss": 2.5598,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.071478843688965,
+      "learning_rate": 1.9594298245614036e-05,
+      "loss": 2.6136,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.025794982910156,
+      "learning_rate": 1.9593384502923977e-05,
+      "loss": 2.723,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.40327787399292,
+      "learning_rate": 1.959247076023392e-05,
+      "loss": 2.7963,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.4485087394714355,
+      "learning_rate": 1.959155701754386e-05,
+      "loss": 2.8513,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.311450004577637,
+      "learning_rate": 1.9590643274853802e-05,
+      "loss": 2.5155,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.396361827850342,
+      "learning_rate": 1.9589729532163742e-05,
+      "loss": 2.8226,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.250613212585449,
+      "learning_rate": 1.9588815789473686e-05,
+      "loss": 3.1254,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.791965961456299,
+      "learning_rate": 1.9587902046783627e-05,
+      "loss": 2.7494,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.066056728363037,
+      "learning_rate": 1.9586988304093567e-05,
+      "loss": 2.1955,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8748276233673096,
+      "learning_rate": 1.958607456140351e-05,
+      "loss": 2.7604,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.429872989654541,
+      "learning_rate": 1.958516081871345e-05,
+      "loss": 2.7589,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.123318195343018,
+      "learning_rate": 1.9584247076023392e-05,
+      "loss": 2.2045,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8217899799346924,
+      "learning_rate": 1.9583333333333333e-05,
+      "loss": 2.5742,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.1973443031311035,
+      "learning_rate": 1.9582419590643277e-05,
+      "loss": 2.4113,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.499605178833008,
+      "learning_rate": 1.9581505847953217e-05,
+      "loss": 2.442,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.590133190155029,
+      "learning_rate": 1.9580592105263158e-05,
+      "loss": 2.673,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9618964195251465,
+      "learning_rate": 1.95796783625731e-05,
+      "loss": 2.4039,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.933688640594482,
+      "learning_rate": 1.9578764619883042e-05,
+      "loss": 2.6201,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.451448440551758,
+      "learning_rate": 1.9577850877192983e-05,
+      "loss": 2.6941,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.862902641296387,
+      "learning_rate": 1.9576937134502923e-05,
+      "loss": 2.8905,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9677035808563232,
+      "learning_rate": 1.9576023391812867e-05,
+      "loss": 2.5216,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.515180587768555,
+      "learning_rate": 1.9575109649122808e-05,
+      "loss": 2.8943,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.579672336578369,
+      "learning_rate": 1.9574195906432748e-05,
+      "loss": 2.5725,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.623291492462158,
+      "learning_rate": 1.9573282163742692e-05,
+      "loss": 3.1196,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.329836368560791,
+      "learning_rate": 1.9572368421052633e-05,
+      "loss": 2.5417,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.509422779083252,
+      "learning_rate": 1.9571454678362573e-05,
+      "loss": 2.5914,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.17478609085083,
+      "learning_rate": 1.9570540935672514e-05,
+      "loss": 2.4886,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.782310485839844,
+      "learning_rate": 1.9569627192982458e-05,
+      "loss": 2.7214,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.22948694229126,
+      "learning_rate": 1.95687134502924e-05,
+      "loss": 2.6065,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.478906631469727,
+      "learning_rate": 1.9567799707602342e-05,
+      "loss": 2.6752,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.466054439544678,
+      "learning_rate": 1.9566885964912283e-05,
+      "loss": 2.357,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.43820333480835,
+      "learning_rate": 1.9565972222222226e-05,
+      "loss": 2.3402,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.85646915435791,
+      "learning_rate": 1.9565058479532167e-05,
+      "loss": 2.8166,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.323126316070557,
+      "learning_rate": 1.9564144736842108e-05,
+      "loss": 2.5897,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.445167064666748,
+      "learning_rate": 1.9563230994152048e-05,
+      "loss": 2.5061,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.423470973968506,
+      "learning_rate": 1.9562317251461992e-05,
+      "loss": 2.7527,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.739559173583984,
+      "learning_rate": 1.9561403508771933e-05,
+      "loss": 2.8829,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.859517812728882,
+      "learning_rate": 1.9560489766081873e-05,
+      "loss": 2.3946,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.585390090942383,
+      "learning_rate": 1.9559576023391817e-05,
+      "loss": 2.6292,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 106.51385498046875,
+      "learning_rate": 1.9558662280701757e-05,
+      "loss": 2.8601,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.535924911499023,
+      "learning_rate": 1.9557748538011698e-05,
+      "loss": 2.8247,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.607716083526611,
+      "learning_rate": 1.955683479532164e-05,
+      "loss": 2.7399,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.7541184425354,
+      "learning_rate": 1.9555921052631582e-05,
+      "loss": 2.4216,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.372600555419922,
+      "learning_rate": 1.9555007309941523e-05,
+      "loss": 2.9986,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.269538879394531,
+      "learning_rate": 1.9554093567251464e-05,
+      "loss": 2.5627,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.888460874557495,
+      "learning_rate": 1.9553179824561407e-05,
+      "loss": 2.7021,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.6207122802734375,
+      "learning_rate": 1.9552266081871348e-05,
+      "loss": 2.3752,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9696249961853027,
+      "learning_rate": 1.955135233918129e-05,
+      "loss": 2.0694,
+      "step": 491
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.091122150421143,
+      "learning_rate": 1.955043859649123e-05,
+      "loss": 2.6496,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.353079795837402,
+      "learning_rate": 1.9549524853801173e-05,
+      "loss": 2.583,
+      "step": 493
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.025610446929932,
+      "learning_rate": 1.9548611111111113e-05,
+      "loss": 3.0876,
+      "step": 494
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 314.0523376464844,
+      "learning_rate": 1.9547697368421054e-05,
+      "loss": 2.4428,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.398311614990234,
+      "learning_rate": 1.9546783625730995e-05,
+      "loss": 2.4174,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.5869550704956055,
+      "learning_rate": 1.954586988304094e-05,
+      "loss": 2.8362,
+      "step": 497
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.53359317779541,
+      "learning_rate": 1.954495614035088e-05,
+      "loss": 2.6944,
+      "step": 498
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.071020126342773,
+      "learning_rate": 1.954404239766082e-05,
+      "loss": 2.3262,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.06372594833374,
+      "learning_rate": 1.9543128654970763e-05,
+      "loss": 2.2904,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.085155963897705,
+      "learning_rate": 1.9542214912280704e-05,
+      "loss": 2.1609,
+      "step": 501
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.70693039894104,
+      "learning_rate": 1.9541301169590644e-05,
+      "loss": 2.2613,
+      "step": 502
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.317453861236572,
+      "learning_rate": 1.9540387426900585e-05,
+      "loss": 2.5726,
+      "step": 503
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.968523979187012,
+      "learning_rate": 1.953947368421053e-05,
+      "loss": 2.5833,
+      "step": 504
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.28600549697876,
+      "learning_rate": 1.953855994152047e-05,
+      "loss": 2.7422,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.853191614151001,
+      "learning_rate": 1.953764619883041e-05,
+      "loss": 2.2769,
+      "step": 506
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.497349262237549,
+      "learning_rate": 1.9536732456140354e-05,
+      "loss": 1.9631,
+      "step": 507
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8684427738189697,
+      "learning_rate": 1.9535818713450294e-05,
+      "loss": 2.369,
+      "step": 508
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.45948600769043,
+      "learning_rate": 1.9534904970760235e-05,
+      "loss": 2.5314,
+      "step": 509
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.301353931427002,
+      "learning_rate": 1.9533991228070175e-05,
+      "loss": 2.5525,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.8236494064331055,
+      "learning_rate": 1.953307748538012e-05,
+      "loss": 2.6622,
+      "step": 511
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6134793758392334,
+      "learning_rate": 1.953216374269006e-05,
+      "loss": 1.9381,
+      "step": 512
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.106671333312988,
+      "learning_rate": 1.953125e-05,
+      "loss": 2.7809,
+      "step": 513
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.364893436431885,
+      "learning_rate": 1.9530336257309944e-05,
+      "loss": 2.592,
+      "step": 514
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.678206443786621,
+      "learning_rate": 1.9529422514619885e-05,
+      "loss": 2.6198,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.720276832580566,
+      "learning_rate": 1.9528508771929825e-05,
+      "loss": 2.4003,
+      "step": 516
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.384203910827637,
+      "learning_rate": 1.9527595029239766e-05,
+      "loss": 2.2136,
+      "step": 517
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.599427700042725,
+      "learning_rate": 1.952668128654971e-05,
+      "loss": 2.5992,
+      "step": 518
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.766898155212402,
+      "learning_rate": 1.952576754385965e-05,
+      "loss": 2.7103,
+      "step": 519
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.140356540679932,
+      "learning_rate": 1.952485380116959e-05,
+      "loss": 2.8018,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.603290557861328,
+      "learning_rate": 1.9523940058479535e-05,
+      "loss": 2.4347,
+      "step": 521
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.046136856079102,
+      "learning_rate": 1.9523026315789475e-05,
+      "loss": 2.4054,
+      "step": 522
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.187882423400879,
+      "learning_rate": 1.9522112573099416e-05,
+      "loss": 2.2638,
+      "step": 523
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.292525768280029,
+      "learning_rate": 1.9521198830409356e-05,
+      "loss": 2.6407,
+      "step": 524
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.590102195739746,
+      "learning_rate": 1.95202850877193e-05,
+      "loss": 2.3054,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.06143856048584,
+      "learning_rate": 1.951937134502924e-05,
+      "loss": 2.0866,
+      "step": 526
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8921730518341064,
+      "learning_rate": 1.951845760233918e-05,
+      "loss": 2.0887,
+      "step": 527
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.522274017333984,
+      "learning_rate": 1.9517543859649125e-05,
+      "loss": 2.1905,
+      "step": 528
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.864253520965576,
+      "learning_rate": 1.9516630116959066e-05,
+      "loss": 2.307,
+      "step": 529
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.337873935699463,
+      "learning_rate": 1.9515716374269006e-05,
+      "loss": 2.5669,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.349379539489746,
+      "learning_rate": 1.9514802631578947e-05,
+      "loss": 2.3586,
+      "step": 531
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.331864356994629,
+      "learning_rate": 1.951388888888889e-05,
+      "loss": 2.4985,
+      "step": 532
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.07124137878418,
+      "learning_rate": 1.951297514619883e-05,
+      "loss": 2.3061,
+      "step": 533
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.55294942855835,
+      "learning_rate": 1.9512061403508772e-05,
+      "loss": 2.6243,
+      "step": 534
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.87838077545166,
+      "learning_rate": 1.9511147660818716e-05,
+      "loss": 2.6709,
+      "step": 535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5878026485443115,
+      "learning_rate": 1.9510233918128656e-05,
+      "loss": 2.0347,
+      "step": 536
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.457275390625,
+      "learning_rate": 1.9509320175438597e-05,
+      "loss": 2.7895,
+      "step": 537
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.322775840759277,
+      "learning_rate": 1.9508406432748537e-05,
+      "loss": 3.1986,
+      "step": 538
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.804234743118286,
+      "learning_rate": 1.950749269005848e-05,
+      "loss": 2.3328,
+      "step": 539
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6262526512145996,
+      "learning_rate": 1.9506578947368422e-05,
+      "loss": 2.2842,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.129265785217285,
+      "learning_rate": 1.9505665204678362e-05,
+      "loss": 2.2555,
+      "step": 541
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8469393253326416,
+      "learning_rate": 1.9504751461988306e-05,
+      "loss": 2.5506,
+      "step": 542
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 23.765226364135742,
+      "learning_rate": 1.9503837719298247e-05,
+      "loss": 2.5251,
+      "step": 543
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8404078483581543,
+      "learning_rate": 1.9502923976608187e-05,
+      "loss": 2.0068,
+      "step": 544
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.255541801452637,
+      "learning_rate": 1.9502010233918128e-05,
+      "loss": 2.6767,
+      "step": 545
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.93123722076416,
+      "learning_rate": 1.950109649122807e-05,
+      "loss": 2.503,
+      "step": 546
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.429471969604492,
+      "learning_rate": 1.9500182748538012e-05,
+      "loss": 2.4065,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.593771934509277,
+      "learning_rate": 1.9499269005847953e-05,
+      "loss": 2.2405,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.293682098388672,
+      "learning_rate": 1.9498355263157897e-05,
+      "loss": 2.3366,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.523510932922363,
+      "learning_rate": 1.9497441520467837e-05,
+      "loss": 2.9902,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.801690101623535,
+      "learning_rate": 1.9496527777777778e-05,
+      "loss": 2.2476,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.251110553741455,
+      "learning_rate": 1.9495614035087718e-05,
+      "loss": 2.4312,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.342991828918457,
+      "learning_rate": 1.9494700292397662e-05,
+      "loss": 2.619,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2895917892456055,
+      "learning_rate": 1.9493786549707603e-05,
+      "loss": 2.5473,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.669871807098389,
+      "learning_rate": 1.9492872807017543e-05,
+      "loss": 2.3479,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.286612033843994,
+      "learning_rate": 1.9491959064327487e-05,
+      "loss": 2.6064,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.771662712097168,
+      "learning_rate": 1.949104532163743e-05,
+      "loss": 2.0511,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.106505870819092,
+      "learning_rate": 1.949013157894737e-05,
+      "loss": 2.1911,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.009293079376221,
+      "learning_rate": 1.9489217836257312e-05,
+      "loss": 3.0589,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.291355133056641,
+      "learning_rate": 1.9488304093567253e-05,
+      "loss": 2.7838,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.041572093963623,
+      "learning_rate": 1.9487390350877197e-05,
+      "loss": 2.1564,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.492610454559326,
+      "learning_rate": 1.9486476608187137e-05,
+      "loss": 2.5799,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.80037260055542,
+      "learning_rate": 1.9485562865497078e-05,
+      "loss": 2.5147,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.481323719024658,
+      "learning_rate": 1.948464912280702e-05,
+      "loss": 2.5359,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.260477066040039,
+      "learning_rate": 1.9483735380116962e-05,
+      "loss": 2.219,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.559218883514404,
+      "learning_rate": 1.9482821637426903e-05,
+      "loss": 2.8155,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.908571720123291,
+      "learning_rate": 1.9481907894736843e-05,
+      "loss": 2.0956,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.409866809844971,
+      "learning_rate": 1.9480994152046787e-05,
+      "loss": 2.6258,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.404524803161621,
+      "learning_rate": 1.9480080409356728e-05,
+      "loss": 2.4749,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.339915752410889,
+      "learning_rate": 1.9479166666666668e-05,
+      "loss": 2.5928,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.499112606048584,
+      "learning_rate": 1.9478252923976612e-05,
+      "loss": 2.8731,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.385985851287842,
+      "learning_rate": 1.9477339181286553e-05,
+      "loss": 2.3779,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.2399492263793945,
+      "learning_rate": 1.9476425438596493e-05,
+      "loss": 2.9127,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.935403347015381,
+      "learning_rate": 1.9475511695906434e-05,
+      "loss": 2.5599,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.447673797607422,
+      "learning_rate": 1.9474597953216377e-05,
+      "loss": 2.1238,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.354786396026611,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 2.7059,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.889080047607422,
+      "learning_rate": 1.947277046783626e-05,
+      "loss": 2.6908,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.019536018371582,
+      "learning_rate": 1.9471856725146202e-05,
+      "loss": 2.6101,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.781765937805176,
+      "learning_rate": 1.9470942982456143e-05,
+      "loss": 2.2567,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.108688354492188,
+      "learning_rate": 1.9470029239766084e-05,
+      "loss": 2.2919,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4754228591918945,
+      "learning_rate": 1.9469115497076024e-05,
+      "loss": 2.5909,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.068760395050049,
+      "learning_rate": 1.9468201754385968e-05,
+      "loss": 2.7352,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.6283369064331055,
+      "learning_rate": 1.946728801169591e-05,
+      "loss": 2.5089,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.853824138641357,
+      "learning_rate": 1.946637426900585e-05,
+      "loss": 2.5116,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.819953680038452,
+      "learning_rate": 1.9465460526315793e-05,
+      "loss": 2.2206,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.031167030334473,
+      "learning_rate": 1.9464546783625733e-05,
+      "loss": 2.6796,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.22852087020874,
+      "learning_rate": 1.9463633040935674e-05,
+      "loss": 2.2229,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.341584205627441,
+      "learning_rate": 1.9462719298245614e-05,
+      "loss": 2.3997,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.304408073425293,
+      "learning_rate": 1.946180555555556e-05,
+      "loss": 2.4105,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.545994758605957,
+      "learning_rate": 1.94608918128655e-05,
+      "loss": 2.7757,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.875184059143066,
+      "learning_rate": 1.945997807017544e-05,
+      "loss": 2.6851,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.675987720489502,
+      "learning_rate": 1.9459064327485383e-05,
+      "loss": 2.9652,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.276796340942383,
+      "learning_rate": 1.9458150584795324e-05,
+      "loss": 2.3452,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.735657691955566,
+      "learning_rate": 1.9457236842105264e-05,
+      "loss": 2.4661,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.818582057952881,
+      "learning_rate": 1.9456323099415205e-05,
+      "loss": 2.4313,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.424051284790039,
+      "learning_rate": 1.945540935672515e-05,
+      "loss": 2.6546,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.895254611968994,
+      "learning_rate": 1.945449561403509e-05,
+      "loss": 2.1784,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.102551460266113,
+      "learning_rate": 1.945358187134503e-05,
+      "loss": 2.371,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8944551944732666,
+      "learning_rate": 1.9452668128654974e-05,
+      "loss": 2.2813,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8904953002929688,
+      "learning_rate": 1.9451754385964914e-05,
+      "loss": 2.3755,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.560111999511719,
+      "learning_rate": 1.9450840643274855e-05,
+      "loss": 2.0048,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8486008644104,
+      "learning_rate": 1.9449926900584795e-05,
+      "loss": 2.8868,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.343141555786133,
+      "learning_rate": 1.944901315789474e-05,
+      "loss": 2.2551,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.5956130027771,
+      "learning_rate": 1.944809941520468e-05,
+      "loss": 2.2745,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.498810291290283,
+      "learning_rate": 1.944718567251462e-05,
+      "loss": 2.4855,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.192286014556885,
+      "learning_rate": 1.9446271929824564e-05,
+      "loss": 2.4536,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.21902084350586,
+      "learning_rate": 1.9445358187134505e-05,
+      "loss": 2.3112,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6286447048187256,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 1.9127,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.891916751861572,
+      "learning_rate": 1.9443530701754386e-05,
+      "loss": 2.5694,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.070334434509277,
+      "learning_rate": 1.944261695906433e-05,
+      "loss": 2.3956,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.221351623535156,
+      "learning_rate": 1.944170321637427e-05,
+      "loss": 2.6852,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.114386558532715,
+      "learning_rate": 1.944078947368421e-05,
+      "loss": 2.5146,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.00811767578125,
+      "learning_rate": 1.9439875730994155e-05,
+      "loss": 2.096,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.091344833374023,
+      "learning_rate": 1.9438961988304095e-05,
+      "loss": 2.8014,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.386213302612305,
+      "learning_rate": 1.9438048245614036e-05,
+      "loss": 2.48,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.060640335083008,
+      "learning_rate": 1.9437134502923976e-05,
+      "loss": 2.3226,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.478163242340088,
+      "learning_rate": 1.943622076023392e-05,
+      "loss": 2.0767,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.175593376159668,
+      "learning_rate": 1.943530701754386e-05,
+      "loss": 2.2414,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.503362655639648,
+      "learning_rate": 1.94343932748538e-05,
+      "loss": 2.3522,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9782774448394775,
+      "learning_rate": 1.9433479532163745e-05,
+      "loss": 2.1596,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.148620128631592,
+      "learning_rate": 1.9432565789473686e-05,
+      "loss": 2.1983,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1792216300964355,
+      "learning_rate": 1.9431652046783626e-05,
+      "loss": 2.2381,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.272927284240723,
+      "learning_rate": 1.9430738304093567e-05,
+      "loss": 2.5951,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.57019567489624,
+      "learning_rate": 1.942982456140351e-05,
+      "loss": 2.1534,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.206149578094482,
+      "learning_rate": 1.942891081871345e-05,
+      "loss": 2.3607,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.797969341278076,
+      "learning_rate": 1.9427997076023392e-05,
+      "loss": 2.601,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7328364849090576,
+      "learning_rate": 1.9427083333333336e-05,
+      "loss": 2.0871,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.305294513702393,
+      "learning_rate": 1.9426169590643276e-05,
+      "loss": 2.311,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.31925630569458,
+      "learning_rate": 1.9425255847953217e-05,
+      "loss": 2.3401,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.497191905975342,
+      "learning_rate": 1.9424342105263157e-05,
+      "loss": 2.2849,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.70668625831604,
+      "learning_rate": 1.94234283625731e-05,
+      "loss": 2.0245,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9598941802978516,
+      "learning_rate": 1.9422514619883042e-05,
+      "loss": 2.2062,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.308687686920166,
+      "learning_rate": 1.9421600877192982e-05,
+      "loss": 2.5603,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.955379009246826,
+      "learning_rate": 1.9420687134502926e-05,
+      "loss": 2.6172,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.761674404144287,
+      "learning_rate": 1.9419773391812867e-05,
+      "loss": 2.258,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.804222106933594,
+      "learning_rate": 1.9418859649122807e-05,
+      "loss": 2.4352,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0762224197387695,
+      "learning_rate": 1.9417945906432748e-05,
+      "loss": 2.0989,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.283390522003174,
+      "learning_rate": 1.941703216374269e-05,
+      "loss": 2.3389,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.582252025604248,
+      "learning_rate": 1.9416118421052632e-05,
+      "loss": 2.4365,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.160040855407715,
+      "learning_rate": 1.9415204678362573e-05,
+      "loss": 2.3807,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.865616321563721,
+      "learning_rate": 1.9414290935672517e-05,
+      "loss": 2.7938,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.511183738708496,
+      "learning_rate": 1.9413377192982457e-05,
+      "loss": 2.3469,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.543236494064331,
+      "learning_rate": 1.94124634502924e-05,
+      "loss": 2.0994,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.265830039978027,
+      "learning_rate": 1.941154970760234e-05,
+      "loss": 2.4604,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.590120792388916,
+      "learning_rate": 1.9410635964912282e-05,
+      "loss": 2.1281,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.529404640197754,
+      "learning_rate": 1.9409722222222226e-05,
+      "loss": 2.2467,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.20025372505188,
+      "learning_rate": 1.9408808479532167e-05,
+      "loss": 1.8312,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.3420023918151855,
+      "learning_rate": 1.9407894736842107e-05,
+      "loss": 2.3445,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.323243141174316,
+      "learning_rate": 1.9406980994152048e-05,
+      "loss": 2.8146,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.795651435852051,
+      "learning_rate": 1.940606725146199e-05,
+      "loss": 2.3384,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.361097812652588,
+      "learning_rate": 1.9405153508771932e-05,
+      "loss": 2.4825,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.981078624725342,
+      "learning_rate": 1.9404239766081873e-05,
+      "loss": 2.348,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9336636066436768,
+      "learning_rate": 1.9403326023391817e-05,
+      "loss": 2.1564,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.442413330078125,
+      "learning_rate": 1.9402412280701757e-05,
+      "loss": 1.8623,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.764566659927368,
+      "learning_rate": 1.9401498538011698e-05,
+      "loss": 1.9723,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.17547607421875,
+      "learning_rate": 1.9400584795321638e-05,
+      "loss": 2.2723,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8362903594970703,
+      "learning_rate": 1.9399671052631582e-05,
+      "loss": 1.6872,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.779014587402344,
+      "learning_rate": 1.9398757309941523e-05,
+      "loss": 2.3265,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.59075665473938,
+      "learning_rate": 1.9397843567251463e-05,
+      "loss": 2.057,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.746501922607422,
+      "learning_rate": 1.9396929824561407e-05,
+      "loss": 2.7284,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.377361297607422,
+      "learning_rate": 1.9396016081871348e-05,
+      "loss": 1.8527,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.766997337341309,
+      "learning_rate": 1.9395102339181288e-05,
+      "loss": 2.3959,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.725579261779785,
+      "learning_rate": 1.939418859649123e-05,
+      "loss": 2.5088,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.276793003082275,
+      "learning_rate": 1.9393274853801173e-05,
+      "loss": 2.7841,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.012879848480225,
+      "learning_rate": 1.9392361111111113e-05,
+      "loss": 2.4472,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9190621376037598,
+      "learning_rate": 1.9391447368421054e-05,
+      "loss": 2.4718,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.437495470046997,
+      "learning_rate": 1.9390533625730997e-05,
+      "loss": 2.1743,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.009895324707031,
+      "learning_rate": 1.9389619883040938e-05,
+      "loss": 2.2026,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.6044769287109375,
+      "learning_rate": 1.938870614035088e-05,
+      "loss": 2.432,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.838970184326172,
+      "learning_rate": 1.938779239766082e-05,
+      "loss": 2.1691,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.127239227294922,
+      "learning_rate": 1.9386878654970763e-05,
+      "loss": 2.6006,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.405420780181885,
+      "learning_rate": 1.9385964912280704e-05,
+      "loss": 2.2047,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.004523754119873,
+      "learning_rate": 1.9385051169590644e-05,
+      "loss": 2.0981,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 38.98711395263672,
+      "learning_rate": 1.9384137426900588e-05,
+      "loss": 2.2934,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.406396389007568,
+      "learning_rate": 1.938322368421053e-05,
+      "loss": 3.0515,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8955700397491455,
+      "learning_rate": 1.938230994152047e-05,
+      "loss": 2.1355,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.883955240249634,
+      "learning_rate": 1.938139619883041e-05,
+      "loss": 2.6208,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9370810985565186,
+      "learning_rate": 1.9380482456140353e-05,
+      "loss": 2.3313,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.93943452835083,
+      "learning_rate": 1.9379568713450294e-05,
+      "loss": 2.4528,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.126813888549805,
+      "learning_rate": 1.9378654970760234e-05,
+      "loss": 2.2503,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.755331516265869,
+      "learning_rate": 1.937774122807018e-05,
+      "loss": 2.3297,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.722839832305908,
+      "learning_rate": 1.937682748538012e-05,
+      "loss": 2.754,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.270683765411377,
+      "learning_rate": 1.937591374269006e-05,
+      "loss": 2.7498,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.823354721069336,
+      "learning_rate": 1.9375e-05,
+      "loss": 2.2593,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.926091194152832,
+      "learning_rate": 1.9374086257309944e-05,
+      "loss": 1.9119,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.157759666442871,
+      "learning_rate": 1.9373172514619884e-05,
+      "loss": 2.1532,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.634852409362793,
+      "learning_rate": 1.9372258771929825e-05,
+      "loss": 1.9574,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9825870990753174,
+      "learning_rate": 1.937134502923977e-05,
+      "loss": 1.9921,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.747121810913086,
+      "learning_rate": 1.937043128654971e-05,
+      "loss": 2.7728,
+      "step": 689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.56245756149292,
+      "learning_rate": 1.936951754385965e-05,
+      "loss": 2.469,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9575164318084717,
+      "learning_rate": 1.936860380116959e-05,
+      "loss": 1.5371,
+      "step": 691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.983729124069214,
+      "learning_rate": 1.9367690058479534e-05,
+      "loss": 2.2914,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9625766277313232,
+      "learning_rate": 1.9366776315789475e-05,
+      "loss": 2.0904,
+      "step": 693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1088786125183105,
+      "learning_rate": 1.9365862573099415e-05,
+      "loss": 2.6476,
+      "step": 694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.349859237670898,
+      "learning_rate": 1.936494883040936e-05,
+      "loss": 2.3426,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8201394081115723,
+      "learning_rate": 1.93640350877193e-05,
+      "loss": 2.3598,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.173504829406738,
+      "learning_rate": 1.936312134502924e-05,
+      "loss": 2.1844,
+      "step": 697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.621152877807617,
+      "learning_rate": 1.936220760233918e-05,
+      "loss": 2.8136,
+      "step": 698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.211901664733887,
+      "learning_rate": 1.9361293859649125e-05,
+      "loss": 2.1699,
+      "step": 699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.375916957855225,
+      "learning_rate": 1.9360380116959065e-05,
+      "loss": 1.9554,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.84232234954834,
+      "learning_rate": 1.9359466374269006e-05,
+      "loss": 2.5,
+      "step": 701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.56414794921875,
+      "learning_rate": 1.935855263157895e-05,
+      "loss": 2.429,
+      "step": 702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.742889642715454,
+      "learning_rate": 1.935763888888889e-05,
+      "loss": 1.9488,
+      "step": 703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.659942150115967,
+      "learning_rate": 1.935672514619883e-05,
+      "loss": 2.1723,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.302935600280762,
+      "learning_rate": 1.935581140350877e-05,
+      "loss": 2.0946,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.495572805404663,
+      "learning_rate": 1.9354897660818715e-05,
+      "loss": 1.973,
+      "step": 706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.440278053283691,
+      "learning_rate": 1.9353983918128656e-05,
+      "loss": 2.7605,
+      "step": 707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.431046009063721,
+      "learning_rate": 1.9353070175438596e-05,
+      "loss": 2.3478,
+      "step": 708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.062654495239258,
+      "learning_rate": 1.935215643274854e-05,
+      "loss": 2.1028,
+      "step": 709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.694772720336914,
+      "learning_rate": 1.935124269005848e-05,
+      "loss": 2.3053,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.713898181915283,
+      "learning_rate": 1.935032894736842e-05,
+      "loss": 2.3003,
+      "step": 711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.885684013366699,
+      "learning_rate": 1.9349415204678362e-05,
+      "loss": 2.1291,
+      "step": 712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.01513147354126,
+      "learning_rate": 1.9348501461988306e-05,
+      "loss": 2.0512,
+      "step": 713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.515771389007568,
+      "learning_rate": 1.9347587719298246e-05,
+      "loss": 2.141,
+      "step": 714
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.480216979980469,
+      "learning_rate": 1.9346673976608187e-05,
+      "loss": 2.6341,
+      "step": 715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.141611099243164,
+      "learning_rate": 1.934576023391813e-05,
+      "loss": 2.5158,
+      "step": 716
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.453416347503662,
+      "learning_rate": 1.934484649122807e-05,
+      "loss": 1.9661,
+      "step": 717
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.863529920578003,
+      "learning_rate": 1.9343932748538012e-05,
+      "loss": 2.3523,
+      "step": 718
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.348851680755615,
+      "learning_rate": 1.9343019005847952e-05,
+      "loss": 2.4887,
+      "step": 719
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.251592636108398,
+      "learning_rate": 1.9342105263157896e-05,
+      "loss": 2.4089,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.802089214324951,
+      "learning_rate": 1.9341191520467837e-05,
+      "loss": 2.2884,
+      "step": 721
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.13915491104126,
+      "learning_rate": 1.9340277777777777e-05,
+      "loss": 2.0632,
+      "step": 722
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4935972690582275,
+      "learning_rate": 1.933936403508772e-05,
+      "loss": 2.3017,
+      "step": 723
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.427822113037109,
+      "learning_rate": 1.9338450292397662e-05,
+      "loss": 2.2842,
+      "step": 724
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.3092041015625,
+      "learning_rate": 1.9337536549707602e-05,
+      "loss": 2.4402,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.913783550262451,
+      "learning_rate": 1.9336622807017543e-05,
+      "loss": 2.4018,
+      "step": 726
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.411746501922607,
+      "learning_rate": 1.9335709064327487e-05,
+      "loss": 2.2419,
+      "step": 727
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.262309551239014,
+      "learning_rate": 1.933479532163743e-05,
+      "loss": 2.3064,
+      "step": 728
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.20076847076416,
+      "learning_rate": 1.933388157894737e-05,
+      "loss": 1.8508,
+      "step": 729
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8839688301086426,
+      "learning_rate": 1.933296783625731e-05,
+      "loss": 2.4128,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.801154613494873,
+      "learning_rate": 1.9332054093567256e-05,
+      "loss": 2.3358,
+      "step": 731
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.894882917404175,
+      "learning_rate": 1.9331140350877196e-05,
+      "loss": 1.8692,
+      "step": 732
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.676366329193115,
+      "learning_rate": 1.9330226608187137e-05,
+      "loss": 2.2993,
+      "step": 733
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9296953678131104,
+      "learning_rate": 1.9329312865497077e-05,
+      "loss": 2.077,
+      "step": 734
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.38944673538208,
+      "learning_rate": 1.932839912280702e-05,
+      "loss": 2.6447,
+      "step": 735
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.881602764129639,
+      "learning_rate": 1.932748538011696e-05,
+      "loss": 2.5183,
+      "step": 736
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.246567249298096,
+      "learning_rate": 1.9326571637426902e-05,
+      "loss": 2.0654,
+      "step": 737
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.292662620544434,
+      "learning_rate": 1.9325657894736846e-05,
+      "loss": 2.8658,
+      "step": 738
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.420701503753662,
+      "learning_rate": 1.9324744152046787e-05,
+      "loss": 2.5813,
+      "step": 739
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5555291175842285,
+      "learning_rate": 1.9323830409356727e-05,
+      "loss": 2.2008,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.603286266326904,
+      "learning_rate": 1.9322916666666668e-05,
+      "loss": 2.4215,
+      "step": 741
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.481932640075684,
+      "learning_rate": 1.932200292397661e-05,
+      "loss": 2.0125,
+      "step": 742
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 77.36470794677734,
+      "learning_rate": 1.9321089181286552e-05,
+      "loss": 2.2598,
+      "step": 743
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2836391925811768,
+      "learning_rate": 1.9320175438596493e-05,
+      "loss": 1.8098,
+      "step": 744
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5476577281951904,
+      "learning_rate": 1.9319261695906437e-05,
+      "loss": 1.7372,
+      "step": 745
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.245576858520508,
+      "learning_rate": 1.9318347953216377e-05,
+      "loss": 2.7617,
+      "step": 746
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.158528804779053,
+      "learning_rate": 1.9317434210526318e-05,
+      "loss": 2.2115,
+      "step": 747
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.686639308929443,
+      "learning_rate": 1.9316520467836258e-05,
+      "loss": 2.3236,
+      "step": 748
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.483212947845459,
+      "learning_rate": 1.9315606725146202e-05,
+      "loss": 1.8025,
+      "step": 749
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4013144969940186,
+      "learning_rate": 1.9314692982456143e-05,
+      "loss": 1.5576,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.808257579803467,
+      "learning_rate": 1.9313779239766083e-05,
+      "loss": 2.2654,
+      "step": 751
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3494467735290527,
+      "learning_rate": 1.9312865497076027e-05,
+      "loss": 2.0231,
+      "step": 752
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.541708946228027,
+      "learning_rate": 1.9311951754385968e-05,
+      "loss": 2.6318,
+      "step": 753
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.058699131011963,
+      "learning_rate": 1.9311038011695908e-05,
+      "loss": 2.4577,
+      "step": 754
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.029106616973877,
+      "learning_rate": 1.931012426900585e-05,
+      "loss": 2.1604,
+      "step": 755
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.00335693359375,
+      "learning_rate": 1.9309210526315793e-05,
+      "loss": 2.283,
+      "step": 756
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.712960720062256,
+      "learning_rate": 1.9308296783625733e-05,
+      "loss": 2.2101,
+      "step": 757
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8647968769073486,
+      "learning_rate": 1.9307383040935674e-05,
+      "loss": 2.106,
+      "step": 758
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.953418016433716,
+      "learning_rate": 1.9306469298245617e-05,
+      "loss": 2.1787,
+      "step": 759
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.113554000854492,
+      "learning_rate": 1.9305555555555558e-05,
+      "loss": 2.1108,
+      "step": 760
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.299625396728516,
+      "learning_rate": 1.93046418128655e-05,
+      "loss": 2.5059,
+      "step": 761
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.584951400756836,
+      "learning_rate": 1.930372807017544e-05,
+      "loss": 2.4414,
+      "step": 762
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.764194011688232,
+      "learning_rate": 1.9302814327485383e-05,
+      "loss": 2.5397,
+      "step": 763
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.439151763916016,
+      "learning_rate": 1.9301900584795324e-05,
+      "loss": 2.1266,
+      "step": 764
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.0771942138671875,
+      "learning_rate": 1.9300986842105264e-05,
+      "loss": 2.3389,
+      "step": 765
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.161471366882324,
+      "learning_rate": 1.9300073099415208e-05,
+      "loss": 2.1373,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2186808586120605,
+      "learning_rate": 1.929915935672515e-05,
+      "loss": 2.343,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7741081714630127,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 2.3068,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.175380706787109,
+      "learning_rate": 1.929733187134503e-05,
+      "loss": 2.3707,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3715341091156006,
+      "learning_rate": 1.9296418128654973e-05,
+      "loss": 1.9781,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.141051769256592,
+      "learning_rate": 1.9295504385964914e-05,
+      "loss": 2.348,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.3509392738342285,
+      "learning_rate": 1.9294590643274854e-05,
+      "loss": 2.7627,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.697583198547363,
+      "learning_rate": 1.92936769005848e-05,
+      "loss": 2.6687,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.005764484405518,
+      "learning_rate": 1.929276315789474e-05,
+      "loss": 1.9967,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.52233362197876,
+      "learning_rate": 1.929184941520468e-05,
+      "loss": 2.7502,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.934687614440918,
+      "learning_rate": 1.929093567251462e-05,
+      "loss": 2.3875,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.94195556640625,
+      "learning_rate": 1.9290021929824564e-05,
+      "loss": 2.1966,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.936178684234619,
+      "learning_rate": 1.9289108187134504e-05,
+      "loss": 1.7578,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.012917518615723,
+      "learning_rate": 1.9288194444444445e-05,
+      "loss": 2.2334,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.472999095916748,
+      "learning_rate": 1.928728070175439e-05,
+      "loss": 1.7817,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.207167625427246,
+      "learning_rate": 1.928636695906433e-05,
+      "loss": 2.3206,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.439852237701416,
+      "learning_rate": 1.928545321637427e-05,
+      "loss": 2.5925,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.064619541168213,
+      "learning_rate": 1.928453947368421e-05,
+      "loss": 1.8311,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.789416790008545,
+      "learning_rate": 1.9283625730994154e-05,
+      "loss": 2.576,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.075992584228516,
+      "learning_rate": 1.9282711988304095e-05,
+      "loss": 1.9102,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.666959047317505,
+      "learning_rate": 1.9281798245614035e-05,
+      "loss": 1.9299,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.03481388092041,
+      "learning_rate": 1.928088450292398e-05,
+      "loss": 2.575,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.271378517150879,
+      "learning_rate": 1.927997076023392e-05,
+      "loss": 2.1847,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.770169734954834,
+      "learning_rate": 1.927905701754386e-05,
+      "loss": 2.4084,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.389711380004883,
+      "learning_rate": 1.92781432748538e-05,
+      "loss": 2.1427,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.621871471405029,
+      "learning_rate": 1.9277229532163745e-05,
+      "loss": 2.718,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7207767963409424,
+      "learning_rate": 1.9276315789473685e-05,
+      "loss": 2.2326,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.286648750305176,
+      "learning_rate": 1.9275402046783626e-05,
+      "loss": 2.0673,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6563689708709717,
+      "learning_rate": 1.927448830409357e-05,
+      "loss": 2.2213,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 23.63245964050293,
+      "learning_rate": 1.927357456140351e-05,
+      "loss": 2.2595,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.178287506103516,
+      "learning_rate": 1.927266081871345e-05,
+      "loss": 2.1046,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.932530403137207,
+      "learning_rate": 1.927174707602339e-05,
+      "loss": 2.4403,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5830094814300537,
+      "learning_rate": 1.9270833333333335e-05,
+      "loss": 2.0644,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.445899963378906,
+      "learning_rate": 1.9269919590643276e-05,
+      "loss": 2.6698,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.304088592529297,
+      "learning_rate": 1.9269005847953216e-05,
+      "loss": 1.745,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7515411376953125,
+      "learning_rate": 1.926809210526316e-05,
+      "loss": 1.926,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.342683792114258,
+      "learning_rate": 1.92671783625731e-05,
+      "loss": 2.1579,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.652381658554077,
+      "learning_rate": 1.926626461988304e-05,
+      "loss": 2.1265,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.279080390930176,
+      "learning_rate": 1.9265350877192982e-05,
+      "loss": 2.2928,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5693817138671875,
+      "learning_rate": 1.9264437134502926e-05,
+      "loss": 2.1937,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2837252616882324,
+      "learning_rate": 1.9263523391812866e-05,
+      "loss": 1.8554,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.39755916595459,
+      "learning_rate": 1.9262609649122807e-05,
+      "loss": 2.4749,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.752129554748535,
+      "learning_rate": 1.926169590643275e-05,
+      "loss": 2.4332,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8616416454315186,
+      "learning_rate": 1.926078216374269e-05,
+      "loss": 2.1259,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.38275146484375,
+      "learning_rate": 1.9259868421052632e-05,
+      "loss": 2.381,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.921184539794922,
+      "learning_rate": 1.9258954678362572e-05,
+      "loss": 2.8489,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5407490730285645,
+      "learning_rate": 1.9258040935672516e-05,
+      "loss": 2.409,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.426090717315674,
+      "learning_rate": 1.925712719298246e-05,
+      "loss": 2.258,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.885023593902588,
+      "learning_rate": 1.92562134502924e-05,
+      "loss": 2.2047,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.983090400695801,
+      "learning_rate": 1.925529970760234e-05,
+      "loss": 2.7147,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.606155872344971,
+      "learning_rate": 1.9254385964912282e-05,
+      "loss": 2.8497,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.179603576660156,
+      "learning_rate": 1.9253472222222226e-05,
+      "loss": 2.2854,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6164438724517822,
+      "learning_rate": 1.9252558479532166e-05,
+      "loss": 2.3499,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8573403358459473,
+      "learning_rate": 1.9251644736842107e-05,
+      "loss": 2.3882,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.195144176483154,
+      "learning_rate": 1.925073099415205e-05,
+      "loss": 2.5851,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.477236270904541,
+      "learning_rate": 1.924981725146199e-05,
+      "loss": 2.7115,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.672609806060791,
+      "learning_rate": 1.924890350877193e-05,
+      "loss": 2.4586,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.454754829406738,
+      "learning_rate": 1.9247989766081872e-05,
+      "loss": 2.0091,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.107194423675537,
+      "learning_rate": 1.9247076023391816e-05,
+      "loss": 2.3164,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.751315116882324,
+      "learning_rate": 1.9246162280701757e-05,
+      "loss": 2.7901,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.22568941116333,
+      "learning_rate": 1.9245248538011697e-05,
+      "loss": 2.4532,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0836968421936035,
+      "learning_rate": 1.924433479532164e-05,
+      "loss": 2.3597,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.558022499084473,
+      "learning_rate": 1.924342105263158e-05,
+      "loss": 2.5008,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9329171180725098,
+      "learning_rate": 1.9242507309941522e-05,
+      "loss": 2.3146,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.64417028427124,
+      "learning_rate": 1.9241593567251463e-05,
+      "loss": 2.6355,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3803181648254395,
+      "learning_rate": 1.9240679824561407e-05,
+      "loss": 1.912,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.8305888175964355,
+      "learning_rate": 1.9239766081871347e-05,
+      "loss": 2.7015,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.524272918701172,
+      "learning_rate": 1.9238852339181288e-05,
+      "loss": 2.0063,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.175962924957275,
+      "learning_rate": 1.923793859649123e-05,
+      "loss": 2.3746,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.845829486846924,
+      "learning_rate": 1.9237024853801172e-05,
+      "loss": 2.3332,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.689274549484253,
+      "learning_rate": 1.9236111111111113e-05,
+      "loss": 1.8561,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.597241401672363,
+      "learning_rate": 1.9235197368421053e-05,
+      "loss": 2.2031,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.458482265472412,
+      "learning_rate": 1.9234283625730997e-05,
+      "loss": 2.3432,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.228919506072998,
+      "learning_rate": 1.9233369883040938e-05,
+      "loss": 2.3911,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8157684803009033,
+      "learning_rate": 1.9232456140350878e-05,
+      "loss": 2.1148,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.529042720794678,
+      "learning_rate": 1.9231542397660822e-05,
+      "loss": 2.4145,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.884305953979492,
+      "learning_rate": 1.9230628654970763e-05,
+      "loss": 2.3204,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6283633708953857,
+      "learning_rate": 1.9229714912280703e-05,
+      "loss": 2.1999,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.017271518707275,
+      "learning_rate": 1.9228801169590644e-05,
+      "loss": 2.453,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.148788928985596,
+      "learning_rate": 1.9227887426900588e-05,
+      "loss": 2.0824,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6241416931152344,
+      "learning_rate": 1.9226973684210528e-05,
+      "loss": 2.1984,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.21389627456665,
+      "learning_rate": 1.922605994152047e-05,
+      "loss": 1.7485,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.273962020874023,
+      "learning_rate": 1.9225146198830413e-05,
+      "loss": 2.4031,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.498214244842529,
+      "learning_rate": 1.9224232456140353e-05,
+      "loss": 1.7404,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.28731107711792,
+      "learning_rate": 1.9223318713450294e-05,
+      "loss": 2.4738,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2411723136901855,
+      "learning_rate": 1.9222404970760234e-05,
+      "loss": 1.844,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.19752311706543,
+      "learning_rate": 1.9221491228070178e-05,
+      "loss": 2.2522,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5070960521698,
+      "learning_rate": 1.922057748538012e-05,
+      "loss": 1.7368,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8987984657287598,
+      "learning_rate": 1.921966374269006e-05,
+      "loss": 2.3,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.277067184448242,
+      "learning_rate": 1.9218750000000003e-05,
+      "loss": 2.6557,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.706038475036621,
+      "learning_rate": 1.9217836257309944e-05,
+      "loss": 2.0293,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.77439022064209,
+      "learning_rate": 1.9216922514619884e-05,
+      "loss": 2.2294,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9565327167510986,
+      "learning_rate": 1.9216008771929825e-05,
+      "loss": 2.1453,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.128267288208008,
+      "learning_rate": 1.921509502923977e-05,
+      "loss": 2.384,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.05579948425293,
+      "learning_rate": 1.921418128654971e-05,
+      "loss": 2.3313,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.182831764221191,
+      "learning_rate": 1.921326754385965e-05,
+      "loss": 1.8804,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.047855377197266,
+      "learning_rate": 1.9212353801169593e-05,
+      "loss": 2.5693,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.1031365394592285,
+      "learning_rate": 1.9211440058479534e-05,
+      "loss": 2.2826,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.27340841293335,
+      "learning_rate": 1.9210526315789474e-05,
+      "loss": 1.8529,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8945047855377197,
+      "learning_rate": 1.9209612573099415e-05,
+      "loss": 2.1626,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.635134220123291,
+      "learning_rate": 1.920869883040936e-05,
+      "loss": 2.263,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.263993263244629,
+      "learning_rate": 1.92077850877193e-05,
+      "loss": 2.1422,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8482468128204346,
+      "learning_rate": 1.920687134502924e-05,
+      "loss": 2.0531,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.382257461547852,
+      "learning_rate": 1.9205957602339184e-05,
+      "loss": 2.5282,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5798611640930176,
+      "learning_rate": 1.9205043859649124e-05,
+      "loss": 2.0615,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.658118724822998,
+      "learning_rate": 1.9204130116959065e-05,
+      "loss": 2.8118,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.04719352722168,
+      "learning_rate": 1.9203216374269005e-05,
+      "loss": 2.2682,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.888798952102661,
+      "learning_rate": 1.920230263157895e-05,
+      "loss": 2.1793,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6961638927459717,
+      "learning_rate": 1.920138888888889e-05,
+      "loss": 2.1554,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.592283248901367,
+      "learning_rate": 1.920047514619883e-05,
+      "loss": 2.4754,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5103678703308105,
+      "learning_rate": 1.9199561403508774e-05,
+      "loss": 2.5799,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8973448276519775,
+      "learning_rate": 1.9198647660818715e-05,
+      "loss": 1.9853,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8423757553100586,
+      "learning_rate": 1.9197733918128655e-05,
+      "loss": 2.3449,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7061686515808105,
+      "learning_rate": 1.9196820175438596e-05,
+      "loss": 2.5734,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8876824378967285,
+      "learning_rate": 1.919590643274854e-05,
+      "loss": 2.2716,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.271540641784668,
+      "learning_rate": 1.919499269005848e-05,
+      "loss": 2.3541,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.269762992858887,
+      "learning_rate": 1.919407894736842e-05,
+      "loss": 2.2293,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3506076335906982,
+      "learning_rate": 1.9193165204678365e-05,
+      "loss": 1.8384,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.189403057098389,
+      "learning_rate": 1.9192251461988305e-05,
+      "loss": 2.4684,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3017659187316895,
+      "learning_rate": 1.9191337719298246e-05,
+      "loss": 2.6695,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.268389701843262,
+      "learning_rate": 1.9190423976608186e-05,
+      "loss": 2.5871,
+      "step": 886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6177544593811035,
+      "learning_rate": 1.918951023391813e-05,
+      "loss": 2.0279,
+      "step": 887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8025765419006348,
+      "learning_rate": 1.918859649122807e-05,
+      "loss": 2.1161,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7674031257629395,
+      "learning_rate": 1.918768274853801e-05,
+      "loss": 2.4905,
+      "step": 889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.353452205657959,
+      "learning_rate": 1.9186769005847955e-05,
+      "loss": 2.3153,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 117.96253204345703,
+      "learning_rate": 1.9185855263157896e-05,
+      "loss": 2.0071,
+      "step": 891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.274765491485596,
+      "learning_rate": 1.9184941520467836e-05,
+      "loss": 2.4241,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.963358879089355,
+      "learning_rate": 1.9184027777777777e-05,
+      "loss": 2.0647,
+      "step": 893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.72590970993042,
+      "learning_rate": 1.918311403508772e-05,
+      "loss": 2.2389,
+      "step": 894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.103352546691895,
+      "learning_rate": 1.918220029239766e-05,
+      "loss": 2.2199,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7910122871398926,
+      "learning_rate": 1.9181286549707602e-05,
+      "loss": 2.1051,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.565003395080566,
+      "learning_rate": 1.9180372807017546e-05,
+      "loss": 2.2757,
+      "step": 897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8677494525909424,
+      "learning_rate": 1.917945906432749e-05,
+      "loss": 1.6927,
+      "step": 898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.736361503601074,
+      "learning_rate": 1.917854532163743e-05,
+      "loss": 2.7571,
+      "step": 899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.788877010345459,
+      "learning_rate": 1.917763157894737e-05,
+      "loss": 2.2425,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6981027126312256,
+      "learning_rate": 1.917671783625731e-05,
+      "loss": 2.4186,
+      "step": 901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.488523483276367,
+      "learning_rate": 1.9175804093567255e-05,
+      "loss": 2.3425,
+      "step": 902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.481853485107422,
+      "learning_rate": 1.9174890350877196e-05,
+      "loss": 2.4695,
+      "step": 903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.2883758544921875,
+      "learning_rate": 1.9173976608187136e-05,
+      "loss": 2.6333,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.997525453567505,
+      "learning_rate": 1.917306286549708e-05,
+      "loss": 2.3604,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.800597667694092,
+      "learning_rate": 1.917214912280702e-05,
+      "loss": 2.0409,
+      "step": 906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.294788837432861,
+      "learning_rate": 1.917123538011696e-05,
+      "loss": 2.3631,
+      "step": 907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5270345211029053,
+      "learning_rate": 1.9170321637426902e-05,
+      "loss": 2.1254,
+      "step": 908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.222897052764893,
+      "learning_rate": 1.9169407894736846e-05,
+      "loss": 2.3615,
+      "step": 909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.461884498596191,
+      "learning_rate": 1.9168494152046786e-05,
+      "loss": 2.334,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.932321548461914,
+      "learning_rate": 1.9167580409356727e-05,
+      "loss": 2.037,
+      "step": 911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.341099262237549,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 1.6966,
+      "step": 912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.21416711807251,
+      "learning_rate": 1.916575292397661e-05,
+      "loss": 2.9359,
+      "step": 913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.77032208442688,
+      "learning_rate": 1.916483918128655e-05,
+      "loss": 1.9312,
+      "step": 914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.752573013305664,
+      "learning_rate": 1.9163925438596492e-05,
+      "loss": 2.347,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3194613456726074,
+      "learning_rate": 1.9163011695906436e-05,
+      "loss": 2.048,
+      "step": 916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.304385662078857,
+      "learning_rate": 1.9162097953216377e-05,
+      "loss": 2.2912,
+      "step": 917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3184661865234375,
+      "learning_rate": 1.9161184210526317e-05,
+      "loss": 2.111,
+      "step": 918
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9930551052093506,
+      "learning_rate": 1.916027046783626e-05,
+      "loss": 2.4548,
+      "step": 919
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.105517864227295,
+      "learning_rate": 1.91593567251462e-05,
+      "loss": 1.8557,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.431919097900391,
+      "learning_rate": 1.9158442982456142e-05,
+      "loss": 2.1904,
+      "step": 921
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.850925922393799,
+      "learning_rate": 1.9157529239766083e-05,
+      "loss": 2.4602,
+      "step": 922
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.444172382354736,
+      "learning_rate": 1.9156615497076027e-05,
+      "loss": 2.171,
+      "step": 923
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.173868656158447,
+      "learning_rate": 1.9155701754385967e-05,
+      "loss": 2.4185,
+      "step": 924
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4473824501037598,
+      "learning_rate": 1.9154788011695908e-05,
+      "loss": 1.8914,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.597036361694336,
+      "learning_rate": 1.915387426900585e-05,
+      "loss": 2.3405,
+      "step": 926
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3072476387023926,
+      "learning_rate": 1.9152960526315792e-05,
+      "loss": 2.0777,
+      "step": 927
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.441699981689453,
+      "learning_rate": 1.9152046783625733e-05,
+      "loss": 2.5961,
+      "step": 928
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.561361789703369,
+      "learning_rate": 1.9151133040935673e-05,
+      "loss": 1.9869,
+      "step": 929
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.16067361831665,
+      "learning_rate": 1.9150219298245617e-05,
+      "loss": 2.3257,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.214109420776367,
+      "learning_rate": 1.9149305555555558e-05,
+      "loss": 2.2544,
+      "step": 931
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.262564659118652,
+      "learning_rate": 1.9148391812865498e-05,
+      "loss": 2.1121,
+      "step": 932
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.351069688796997,
+      "learning_rate": 1.9147478070175442e-05,
+      "loss": 1.8093,
+      "step": 933
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.456480026245117,
+      "learning_rate": 1.9146564327485383e-05,
+      "loss": 1.988,
+      "step": 934
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.609579563140869,
+      "learning_rate": 1.9145650584795323e-05,
+      "loss": 2.2061,
+      "step": 935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.440683364868164,
+      "learning_rate": 1.9144736842105264e-05,
+      "loss": 2.3493,
+      "step": 936
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.073796272277832,
+      "learning_rate": 1.9143823099415208e-05,
+      "loss": 2.5598,
+      "step": 937
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4021201133728027,
+      "learning_rate": 1.9142909356725148e-05,
+      "loss": 1.911,
+      "step": 938
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9411051273345947,
+      "learning_rate": 1.914199561403509e-05,
+      "loss": 1.8005,
+      "step": 939
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.108205318450928,
+      "learning_rate": 1.9141081871345033e-05,
+      "loss": 1.9213,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.816187858581543,
+      "learning_rate": 1.9140168128654973e-05,
+      "loss": 2.0325,
+      "step": 941
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2830629348754883,
+      "learning_rate": 1.9139254385964914e-05,
+      "loss": 1.9429,
+      "step": 942
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9726951122283936,
+      "learning_rate": 1.9138340643274854e-05,
+      "loss": 1.8168,
+      "step": 943
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.138609886169434,
+      "learning_rate": 1.9137426900584798e-05,
+      "loss": 2.1464,
+      "step": 944
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.869192600250244,
+      "learning_rate": 1.913651315789474e-05,
+      "loss": 2.2918,
+      "step": 945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.241971254348755,
+      "learning_rate": 1.913559941520468e-05,
+      "loss": 1.9193,
+      "step": 946
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2145960330963135,
+      "learning_rate": 1.913468567251462e-05,
+      "loss": 1.4563,
+      "step": 947
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.035637378692627,
+      "learning_rate": 1.9133771929824564e-05,
+      "loss": 2.7332,
+      "step": 948
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6907827854156494,
+      "learning_rate": 1.9132858187134504e-05,
+      "loss": 1.8464,
+      "step": 949
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.727969169616699,
+      "learning_rate": 1.9131944444444445e-05,
+      "loss": 2.2838,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.519984722137451,
+      "learning_rate": 1.913103070175439e-05,
+      "loss": 2.2353,
+      "step": 951
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9727723598480225,
+      "learning_rate": 1.913011695906433e-05,
+      "loss": 2.1459,
+      "step": 952
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.782010078430176,
+      "learning_rate": 1.912920321637427e-05,
+      "loss": 2.373,
+      "step": 953
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.505768775939941,
+      "learning_rate": 1.912828947368421e-05,
+      "loss": 1.9886,
+      "step": 954
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.413275241851807,
+      "learning_rate": 1.9127375730994154e-05,
+      "loss": 2.7303,
+      "step": 955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.475909233093262,
+      "learning_rate": 1.9126461988304094e-05,
+      "loss": 1.8183,
+      "step": 956
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.251417398452759,
+      "learning_rate": 1.9125548245614035e-05,
+      "loss": 1.9281,
+      "step": 957
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6955249309539795,
+      "learning_rate": 1.912463450292398e-05,
+      "loss": 2.2656,
+      "step": 958
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.847608804702759,
+      "learning_rate": 1.912372076023392e-05,
+      "loss": 1.7367,
+      "step": 959
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.650590419769287,
+      "learning_rate": 1.912280701754386e-05,
+      "loss": 2.2316,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.172611713409424,
+      "learning_rate": 1.91218932748538e-05,
+      "loss": 2.3718,
+      "step": 961
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6343131065368652,
+      "learning_rate": 1.9120979532163744e-05,
+      "loss": 1.8087,
+      "step": 962
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.967189311981201,
+      "learning_rate": 1.9120065789473685e-05,
+      "loss": 2.2248,
+      "step": 963
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.631379127502441,
+      "learning_rate": 1.9119152046783625e-05,
+      "loss": 2.1244,
+      "step": 964
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5074462890625,
+      "learning_rate": 1.911823830409357e-05,
+      "loss": 1.796,
+      "step": 965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6078860759735107,
+      "learning_rate": 1.911732456140351e-05,
+      "loss": 1.7831,
+      "step": 966
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.584834575653076,
+      "learning_rate": 1.911641081871345e-05,
+      "loss": 2.5384,
+      "step": 967
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.633042097091675,
+      "learning_rate": 1.911549707602339e-05,
+      "loss": 1.9808,
+      "step": 968
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.903585910797119,
+      "learning_rate": 1.9114583333333335e-05,
+      "loss": 1.8777,
+      "step": 969
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.041944980621338,
+      "learning_rate": 1.9113669590643275e-05,
+      "loss": 1.8118,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8259356021881104,
+      "learning_rate": 1.9112755847953216e-05,
+      "loss": 1.7802,
+      "step": 971
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.573541164398193,
+      "learning_rate": 1.911184210526316e-05,
+      "loss": 2.0247,
+      "step": 972
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3990559577941895,
+      "learning_rate": 1.91109283625731e-05,
+      "loss": 2.1439,
+      "step": 973
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.101099967956543,
+      "learning_rate": 1.911001461988304e-05,
+      "loss": 1.9131,
+      "step": 974
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5512924194335938,
+      "learning_rate": 1.910910087719298e-05,
+      "loss": 2.103,
+      "step": 975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6550965309143066,
+      "learning_rate": 1.9108187134502925e-05,
+      "loss": 1.7742,
+      "step": 976
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.993306875228882,
+      "learning_rate": 1.9107273391812866e-05,
+      "loss": 1.9776,
+      "step": 977
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.526031494140625,
+      "learning_rate": 1.9106359649122806e-05,
+      "loss": 1.8546,
+      "step": 978
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.201831340789795,
+      "learning_rate": 1.910544590643275e-05,
+      "loss": 1.9286,
+      "step": 979
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.714323043823242,
+      "learning_rate": 1.910453216374269e-05,
+      "loss": 1.8302,
+      "step": 980
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.87203311920166,
+      "learning_rate": 1.910361842105263e-05,
+      "loss": 2.1973,
+      "step": 981
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8676538467407227,
+      "learning_rate": 1.9102704678362572e-05,
+      "loss": 2.0018,
+      "step": 982
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.681328773498535,
+      "learning_rate": 1.9101790935672516e-05,
+      "loss": 2.2346,
+      "step": 983
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.559359073638916,
+      "learning_rate": 1.910087719298246e-05,
+      "loss": 2.474,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.519443035125732,
+      "learning_rate": 1.90999634502924e-05,
+      "loss": 2.6404,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.190274238586426,
+      "learning_rate": 1.909904970760234e-05,
+      "loss": 2.2686,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3926849365234375,
+      "learning_rate": 1.9098135964912285e-05,
+      "loss": 2.5915,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.502401351928711,
+      "learning_rate": 1.9097222222222225e-05,
+      "loss": 2.2792,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.537962913513184,
+      "learning_rate": 1.9096308479532166e-05,
+      "loss": 2.4233,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6795568466186523,
+      "learning_rate": 1.9095394736842106e-05,
+      "loss": 2.1365,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5512542724609375,
+      "learning_rate": 1.909448099415205e-05,
+      "loss": 2.5385,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6758036613464355,
+      "learning_rate": 1.909356725146199e-05,
+      "loss": 2.0862,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.98296856880188,
+      "learning_rate": 1.909265350877193e-05,
+      "loss": 2.0288,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.339117050170898,
+      "learning_rate": 1.9091739766081875e-05,
+      "loss": 2.3703,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.676278114318848,
+      "learning_rate": 1.9090826023391816e-05,
+      "loss": 2.1305,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4482717514038086,
+      "learning_rate": 1.9089912280701756e-05,
+      "loss": 1.9838,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.638920783996582,
+      "learning_rate": 1.9088998538011697e-05,
+      "loss": 2.2719,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4708774089813232,
+      "learning_rate": 1.908808479532164e-05,
+      "loss": 1.7595,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4779558181762695,
+      "learning_rate": 1.908717105263158e-05,
+      "loss": 2.167,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.1510419845581055,
+      "learning_rate": 1.9086257309941522e-05,
+      "loss": 2.1847,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.960857629776001,
+      "learning_rate": 1.9085343567251466e-05,
+      "loss": 1.9408,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.747143745422363,
+      "learning_rate": 1.9084429824561406e-05,
+      "loss": 2.1857,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.905744791030884,
+      "learning_rate": 1.9083516081871347e-05,
+      "loss": 2.3941,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5020554065704346,
+      "learning_rate": 1.9082602339181287e-05,
+      "loss": 2.0396,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.519199848175049,
+      "learning_rate": 1.908168859649123e-05,
+      "loss": 2.2244,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.202764511108398,
+      "learning_rate": 1.908077485380117e-05,
+      "loss": 2.1798,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.655405044555664,
+      "learning_rate": 1.9079861111111112e-05,
+      "loss": 2.093,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9121992588043213,
+      "learning_rate": 1.9078947368421056e-05,
+      "loss": 2.0526,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6697449684143066,
+      "learning_rate": 1.9078033625730997e-05,
+      "loss": 2.1624,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.104246139526367,
+      "learning_rate": 1.9077119883040937e-05,
+      "loss": 1.919,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.391928672790527,
+      "learning_rate": 1.9076206140350878e-05,
+      "loss": 2.0336,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.035699844360352,
+      "learning_rate": 1.907529239766082e-05,
+      "loss": 2.5734,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.314977169036865,
+      "learning_rate": 1.9074378654970762e-05,
+      "loss": 2.2745,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.086503982543945,
+      "learning_rate": 1.9073464912280703e-05,
+      "loss": 2.0752,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.993229866027832,
+      "learning_rate": 1.9072551169590647e-05,
+      "loss": 1.8427,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.281862258911133,
+      "learning_rate": 1.9071637426900587e-05,
+      "loss": 2.3523,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.799807071685791,
+      "learning_rate": 1.9070723684210528e-05,
+      "loss": 2.2065,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.660357713699341,
+      "learning_rate": 1.9069809941520468e-05,
+      "loss": 2.0259,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.409531831741333,
+      "learning_rate": 1.9068896198830412e-05,
+      "loss": 1.5721,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7202935218811035,
+      "learning_rate": 1.9067982456140353e-05,
+      "loss": 1.945,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.501166820526123,
+      "learning_rate": 1.9067068713450293e-05,
+      "loss": 2.0702,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.8482441902160645,
+      "learning_rate": 1.9066154970760237e-05,
+      "loss": 2.1739,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.265207290649414,
+      "learning_rate": 1.9065241228070178e-05,
+      "loss": 2.2069,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5383334159851074,
+      "learning_rate": 1.9064327485380118e-05,
+      "loss": 2.2069,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.981412649154663,
+      "learning_rate": 1.906341374269006e-05,
+      "loss": 2.0204,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7555694580078125,
+      "learning_rate": 1.9062500000000003e-05,
+      "loss": 2.3209,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.023899555206299,
+      "learning_rate": 1.9061586257309943e-05,
+      "loss": 2.6473,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3747143745422363,
+      "learning_rate": 1.9060672514619884e-05,
+      "loss": 1.7128,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.801265239715576,
+      "learning_rate": 1.9059758771929828e-05,
+      "loss": 2.1179,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8765008449554443,
+      "learning_rate": 1.9058845029239768e-05,
+      "loss": 2.0346,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.850428342819214,
+      "learning_rate": 1.905793128654971e-05,
+      "loss": 1.9827,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9616787433624268,
+      "learning_rate": 1.905701754385965e-05,
+      "loss": 1.5861,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.988034963607788,
+      "learning_rate": 1.9056103801169593e-05,
+      "loss": 2.5328,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7351388931274414,
+      "learning_rate": 1.9055190058479534e-05,
+      "loss": 2.0645,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4185831546783447,
+      "learning_rate": 1.9054276315789474e-05,
+      "loss": 2.0659,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5192694664001465,
+      "learning_rate": 1.9053362573099418e-05,
+      "loss": 1.7903,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.264528751373291,
+      "learning_rate": 1.905244883040936e-05,
+      "loss": 1.9456,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.408207416534424,
+      "learning_rate": 1.90515350877193e-05,
+      "loss": 2.1891,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7601332664489746,
+      "learning_rate": 1.905062134502924e-05,
+      "loss": 2.1977,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.707606554031372,
+      "learning_rate": 1.9049707602339184e-05,
+      "loss": 2.1994,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.28793478012085,
+      "learning_rate": 1.9048793859649124e-05,
+      "loss": 2.3832,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.762112617492676,
+      "learning_rate": 1.9047880116959065e-05,
+      "loss": 2.0087,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4595963954925537,
+      "learning_rate": 1.904696637426901e-05,
+      "loss": 2.1626,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9789469242095947,
+      "learning_rate": 1.904605263157895e-05,
+      "loss": 2.0358,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8171684741973877,
+      "learning_rate": 1.904513888888889e-05,
+      "loss": 2.1039,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.009982585906982,
+      "learning_rate": 1.904422514619883e-05,
+      "loss": 1.934,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.62654709815979,
+      "learning_rate": 1.9043311403508774e-05,
+      "loss": 2.0292,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.111857891082764,
+      "learning_rate": 1.9042397660818714e-05,
+      "loss": 2.0293,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1074323654174805,
+      "learning_rate": 1.9041483918128655e-05,
+      "loss": 2.0349,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8492023944854736,
+      "learning_rate": 1.90405701754386e-05,
+      "loss": 1.954,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9887311458587646,
+      "learning_rate": 1.903965643274854e-05,
+      "loss": 2.3238,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.235262632369995,
+      "learning_rate": 1.903874269005848e-05,
+      "loss": 1.8725,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.477421760559082,
+      "learning_rate": 1.903782894736842e-05,
+      "loss": 2.5491,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9157755374908447,
+      "learning_rate": 1.9036915204678364e-05,
+      "loss": 1.8955,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.217145919799805,
+      "learning_rate": 1.9036001461988305e-05,
+      "loss": 2.0716,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4757111072540283,
+      "learning_rate": 1.9035087719298245e-05,
+      "loss": 2.2322,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.56860613822937,
+      "learning_rate": 1.903417397660819e-05,
+      "loss": 2.0672,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0114312171936035,
+      "learning_rate": 1.903326023391813e-05,
+      "loss": 1.8989,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.87302827835083,
+      "learning_rate": 1.903234649122807e-05,
+      "loss": 2.3503,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.531928539276123,
+      "learning_rate": 1.903143274853801e-05,
+      "loss": 2.0102,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.346701622009277,
+      "learning_rate": 1.9030519005847955e-05,
+      "loss": 2.3235,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.4518351554870605,
+      "learning_rate": 1.9029605263157895e-05,
+      "loss": 2.4721,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.700350284576416,
+      "learning_rate": 1.9028691520467836e-05,
+      "loss": 2.3185,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3473920822143555,
+      "learning_rate": 1.902777777777778e-05,
+      "loss": 2.1042,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.911104679107666,
+      "learning_rate": 1.902686403508772e-05,
+      "loss": 1.9479,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1877875328063965,
+      "learning_rate": 1.902595029239766e-05,
+      "loss": 1.8379,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8794305324554443,
+      "learning_rate": 1.90250365497076e-05,
+      "loss": 1.9586,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.831916332244873,
+      "learning_rate": 1.9024122807017545e-05,
+      "loss": 2.3091,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.073638916015625,
+      "learning_rate": 1.902320906432749e-05,
+      "loss": 2.353,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.797715187072754,
+      "learning_rate": 1.902229532163743e-05,
+      "loss": 2.4478,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.244030475616455,
+      "learning_rate": 1.902138157894737e-05,
+      "loss": 2.2964,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.38600492477417,
+      "learning_rate": 1.902046783625731e-05,
+      "loss": 1.8168,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.945760726928711,
+      "learning_rate": 1.9019554093567255e-05,
+      "loss": 2.4256,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.050909519195557,
+      "learning_rate": 1.9018640350877195e-05,
+      "loss": 2.5673,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4666361808776855,
+      "learning_rate": 1.9017726608187136e-05,
+      "loss": 2.4608,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.773751735687256,
+      "learning_rate": 1.901681286549708e-05,
+      "loss": 1.9652,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9890389442443848,
+      "learning_rate": 1.901589912280702e-05,
+      "loss": 2.3712,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5691959857940674,
+      "learning_rate": 1.901498538011696e-05,
+      "loss": 2.13,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.183465003967285,
+      "learning_rate": 1.90140716374269e-05,
+      "loss": 1.931,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.429266452789307,
+      "learning_rate": 1.9013157894736845e-05,
+      "loss": 2.1171,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.812577247619629,
+      "learning_rate": 1.9012244152046786e-05,
+      "loss": 2.2873,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.280272960662842,
+      "learning_rate": 1.9011330409356726e-05,
+      "loss": 2.6724,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3942134380340576,
+      "learning_rate": 1.901041666666667e-05,
+      "loss": 1.7027,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.879467248916626,
+      "learning_rate": 1.900950292397661e-05,
+      "loss": 1.9698,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.865872383117676,
+      "learning_rate": 1.900858918128655e-05,
+      "loss": 2.3416,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7740535736083984,
+      "learning_rate": 1.9007675438596492e-05,
+      "loss": 2.0699,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3177647590637207,
+      "learning_rate": 1.9006761695906436e-05,
+      "loss": 1.9649,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.515688419342041,
+      "learning_rate": 1.9005847953216376e-05,
+      "loss": 1.8964,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.963352680206299,
+      "learning_rate": 1.9004934210526317e-05,
+      "loss": 1.8569,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.616734027862549,
+      "learning_rate": 1.900402046783626e-05,
+      "loss": 2.0368,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.117925643920898,
+      "learning_rate": 1.90031067251462e-05,
+      "loss": 2.0714,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.472536087036133,
+      "learning_rate": 1.9002192982456142e-05,
+      "loss": 2.114,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.457908868789673,
+      "learning_rate": 1.9001279239766082e-05,
+      "loss": 2.1797,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8006069660186768,
+      "learning_rate": 1.9000365497076026e-05,
+      "loss": 2.0155,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.103565216064453,
+      "learning_rate": 1.8999451754385967e-05,
+      "loss": 1.9967,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8436527252197266,
+      "learning_rate": 1.8998538011695907e-05,
+      "loss": 2.2092,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.150453567504883,
+      "learning_rate": 1.899762426900585e-05,
+      "loss": 2.108,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.776523590087891,
+      "learning_rate": 1.899671052631579e-05,
+      "loss": 2.0854,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2069144248962402,
+      "learning_rate": 1.8995796783625732e-05,
+      "loss": 1.8228,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.593660593032837,
+      "learning_rate": 1.8994883040935673e-05,
+      "loss": 1.7513,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.544040679931641,
+      "learning_rate": 1.8993969298245617e-05,
+      "loss": 2.2767,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7124412059783936,
+      "learning_rate": 1.8993055555555557e-05,
+      "loss": 2.2375,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.1874613761901855,
+      "learning_rate": 1.8992141812865498e-05,
+      "loss": 2.2561,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.670504093170166,
+      "learning_rate": 1.899122807017544e-05,
+      "loss": 2.1949,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.556002140045166,
+      "learning_rate": 1.8990314327485382e-05,
+      "loss": 1.9465,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.766432762145996,
+      "learning_rate": 1.8989400584795323e-05,
+      "loss": 2.7163,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6667733192443848,
+      "learning_rate": 1.8988486842105263e-05,
+      "loss": 2.2987,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.082210540771484,
+      "learning_rate": 1.8987573099415207e-05,
+      "loss": 2.0768,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.488269329071045,
+      "learning_rate": 1.8986659356725148e-05,
+      "loss": 1.997,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9037983417510986,
+      "learning_rate": 1.8985745614035088e-05,
+      "loss": 1.8015,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.420522689819336,
+      "learning_rate": 1.8984831871345032e-05,
+      "loss": 1.9505,
+      "step": 1111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.212322235107422,
+      "learning_rate": 1.8983918128654973e-05,
+      "loss": 1.7673,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4913277626037598,
+      "learning_rate": 1.8983004385964913e-05,
+      "loss": 1.5814,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8928399085998535,
+      "learning_rate": 1.8982090643274854e-05,
+      "loss": 1.5774,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.268044471740723,
+      "learning_rate": 1.8981176900584798e-05,
+      "loss": 2.4792,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.09326696395874,
+      "learning_rate": 1.8980263157894738e-05,
+      "loss": 1.9114,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3353095054626465,
+      "learning_rate": 1.897934941520468e-05,
+      "loss": 2.0247,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.1051154136657715,
+      "learning_rate": 1.8978435672514623e-05,
+      "loss": 2.4194,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.868288516998291,
+      "learning_rate": 1.8977521929824563e-05,
+      "loss": 2.5106,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8706986904144287,
+      "learning_rate": 1.8976608187134504e-05,
+      "loss": 2.0426,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8199541568756104,
+      "learning_rate": 1.8975694444444444e-05,
+      "loss": 2.2069,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.296692371368408,
+      "learning_rate": 1.8974780701754388e-05,
+      "loss": 2.3418,
+      "step": 1122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.179008483886719,
+      "learning_rate": 1.897386695906433e-05,
+      "loss": 2.4008,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7358851432800293,
+      "learning_rate": 1.897295321637427e-05,
+      "loss": 2.242,
+      "step": 1124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.26728630065918,
+      "learning_rate": 1.8972039473684213e-05,
+      "loss": 2.1339,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.287548065185547,
+      "learning_rate": 1.8971125730994154e-05,
+      "loss": 2.4128,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.942791938781738,
+      "learning_rate": 1.8970211988304094e-05,
+      "loss": 2.474,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.386000871658325,
+      "learning_rate": 1.8969298245614035e-05,
+      "loss": 2.038,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4568562507629395,
+      "learning_rate": 1.896838450292398e-05,
+      "loss": 2.1549,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.097240924835205,
+      "learning_rate": 1.896747076023392e-05,
+      "loss": 2.09,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.527532577514648,
+      "learning_rate": 1.896655701754386e-05,
+      "loss": 2.0886,
+      "step": 1131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.865006685256958,
+      "learning_rate": 1.8965643274853804e-05,
+      "loss": 1.958,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6839888095855713,
+      "learning_rate": 1.8964729532163744e-05,
+      "loss": 2.0896,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.222008466720581,
+      "learning_rate": 1.8963815789473685e-05,
+      "loss": 1.842,
+      "step": 1134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5155394077301025,
+      "learning_rate": 1.8962902046783625e-05,
+      "loss": 1.9761,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7227892875671387,
+      "learning_rate": 1.896198830409357e-05,
+      "loss": 2.0484,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0542497634887695,
+      "learning_rate": 1.896107456140351e-05,
+      "loss": 2.0639,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.821378707885742,
+      "learning_rate": 1.896016081871345e-05,
+      "loss": 2.3966,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.265622854232788,
+      "learning_rate": 1.8959247076023394e-05,
+      "loss": 1.7194,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.694140672683716,
+      "learning_rate": 1.8958333333333334e-05,
+      "loss": 2.483,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4101192951202393,
+      "learning_rate": 1.8957419590643275e-05,
+      "loss": 2.1813,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.568486213684082,
+      "learning_rate": 1.8956505847953216e-05,
+      "loss": 1.6204,
+      "step": 1142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3263025283813477,
+      "learning_rate": 1.895559210526316e-05,
+      "loss": 1.8654,
+      "step": 1143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6062896251678467,
+      "learning_rate": 1.89546783625731e-05,
+      "loss": 1.9421,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3594796657562256,
+      "learning_rate": 1.895376461988304e-05,
+      "loss": 2.2164,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.096215724945068,
+      "learning_rate": 1.8952850877192984e-05,
+      "loss": 2.057,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.079684734344482,
+      "learning_rate": 1.8951937134502925e-05,
+      "loss": 2.0115,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0841965675354,
+      "learning_rate": 1.8951023391812865e-05,
+      "loss": 1.7117,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.998613357543945,
+      "learning_rate": 1.8950109649122806e-05,
+      "loss": 2.1313,
+      "step": 1149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8772940635681152,
+      "learning_rate": 1.894919590643275e-05,
+      "loss": 2.2208,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.042503356933594,
+      "learning_rate": 1.894828216374269e-05,
+      "loss": 2.1524,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.504063129425049,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 2.3263,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8735170364379883,
+      "learning_rate": 1.8946454678362575e-05,
+      "loss": 1.8999,
+      "step": 1153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.1463141441345215,
+      "learning_rate": 1.8945540935672515e-05,
+      "loss": 1.8916,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.302323579788208,
+      "learning_rate": 1.894462719298246e-05,
+      "loss": 1.8872,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8732757568359375,
+      "learning_rate": 1.89437134502924e-05,
+      "loss": 1.9749,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.299610614776611,
+      "learning_rate": 1.894279970760234e-05,
+      "loss": 2.0297,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.966254472732544,
+      "learning_rate": 1.8941885964912284e-05,
+      "loss": 2.2164,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.595834732055664,
+      "learning_rate": 1.8940972222222225e-05,
+      "loss": 2.0226,
+      "step": 1159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.753574848175049,
+      "learning_rate": 1.8940058479532165e-05,
+      "loss": 2.0095,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.193459987640381,
+      "learning_rate": 1.893914473684211e-05,
+      "loss": 2.2699,
+      "step": 1161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.253948926925659,
+      "learning_rate": 1.893823099415205e-05,
+      "loss": 1.6358,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9676618576049805,
+      "learning_rate": 1.893731725146199e-05,
+      "loss": 1.9217,
+      "step": 1163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.63472580909729,
+      "learning_rate": 1.893640350877193e-05,
+      "loss": 1.9178,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.088320255279541,
+      "learning_rate": 1.8935489766081875e-05,
+      "loss": 2.0602,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.187202453613281,
+      "learning_rate": 1.8934576023391815e-05,
+      "loss": 2.2063,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6920628547668457,
+      "learning_rate": 1.8933662280701756e-05,
+      "loss": 2.085,
+      "step": 1167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.714535713195801,
+      "learning_rate": 1.89327485380117e-05,
+      "loss": 2.2257,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.324184894561768,
+      "learning_rate": 1.893183479532164e-05,
+      "loss": 1.9611,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.077911376953125,
+      "learning_rate": 1.893092105263158e-05,
+      "loss": 1.9483,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.210391044616699,
+      "learning_rate": 1.893000730994152e-05,
+      "loss": 2.2306,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7274816036224365,
+      "learning_rate": 1.8929093567251465e-05,
+      "loss": 2.1501,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.624011993408203,
+      "learning_rate": 1.8928179824561406e-05,
+      "loss": 2.023,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.278721809387207,
+      "learning_rate": 1.8927266081871346e-05,
+      "loss": 2.3959,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.192224979400635,
+      "learning_rate": 1.892635233918129e-05,
+      "loss": 1.9791,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.793865442276001,
+      "learning_rate": 1.892543859649123e-05,
+      "loss": 1.8525,
+      "step": 1176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.075912952423096,
+      "learning_rate": 1.892452485380117e-05,
+      "loss": 2.1107,
+      "step": 1177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5858750343322754,
+      "learning_rate": 1.8923611111111112e-05,
+      "loss": 2.0004,
+      "step": 1178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.4264020919799805,
+      "learning_rate": 1.8922697368421056e-05,
+      "loss": 2.3289,
+      "step": 1179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4843993186950684,
+      "learning_rate": 1.8921783625730996e-05,
+      "loss": 1.7352,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5505943298339844,
+      "learning_rate": 1.8920869883040937e-05,
+      "loss": 1.8941,
+      "step": 1181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.207243919372559,
+      "learning_rate": 1.891995614035088e-05,
+      "loss": 2.3656,
+      "step": 1182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.749342679977417,
+      "learning_rate": 1.891904239766082e-05,
+      "loss": 1.8096,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 14.770569801330566,
+      "learning_rate": 1.8918128654970762e-05,
+      "loss": 2.0026,
+      "step": 1184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.47472882270813,
+      "learning_rate": 1.8917214912280702e-05,
+      "loss": 1.9954,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.37821102142334,
+      "learning_rate": 1.8916301169590646e-05,
+      "loss": 2.4093,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.525505304336548,
+      "learning_rate": 1.8915387426900587e-05,
+      "loss": 1.8236,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.056930065155029,
+      "learning_rate": 1.8914473684210527e-05,
+      "loss": 1.917,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9050662517547607,
+      "learning_rate": 1.891355994152047e-05,
+      "loss": 2.3434,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.8885791301727295,
+      "learning_rate": 1.891264619883041e-05,
+      "loss": 2.2334,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.311828136444092,
+      "learning_rate": 1.8911732456140352e-05,
+      "loss": 2.2472,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4389681816101074,
+      "learning_rate": 1.8910818713450293e-05,
+      "loss": 1.8408,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0956130027771,
+      "learning_rate": 1.8909904970760237e-05,
+      "loss": 2.1504,
+      "step": 1193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.720426082611084,
+      "learning_rate": 1.8908991228070177e-05,
+      "loss": 2.2545,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.656968116760254,
+      "learning_rate": 1.8908077485380118e-05,
+      "loss": 1.9117,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.078965663909912,
+      "learning_rate": 1.890716374269006e-05,
+      "loss": 2.3354,
+      "step": 1196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9284451007843018,
+      "learning_rate": 1.8906250000000002e-05,
+      "loss": 2.0852,
+      "step": 1197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.34766149520874,
+      "learning_rate": 1.8905336257309943e-05,
+      "loss": 1.8123,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.502547264099121,
+      "learning_rate": 1.8904422514619883e-05,
+      "loss": 2.0325,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.181613445281982,
+      "learning_rate": 1.8903508771929827e-05,
+      "loss": 2.3239,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6656765937805176,
+      "learning_rate": 1.8902595029239768e-05,
+      "loss": 2.0455,
+      "step": 1201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.819164752960205,
+      "learning_rate": 1.8901681286549708e-05,
+      "loss": 2.1526,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.851900577545166,
+      "learning_rate": 1.8900767543859652e-05,
+      "loss": 2.516,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.050770282745361,
+      "learning_rate": 1.8899853801169593e-05,
+      "loss": 2.2412,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6035029888153076,
+      "learning_rate": 1.8898940058479533e-05,
+      "loss": 2.1114,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.751007080078125,
+      "learning_rate": 1.8898026315789474e-05,
+      "loss": 2.141,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.897411346435547,
+      "learning_rate": 1.8897112573099418e-05,
+      "loss": 2.1194,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0904998779296875,
+      "learning_rate": 1.8896198830409358e-05,
+      "loss": 2.0088,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.611757278442383,
+      "learning_rate": 1.88952850877193e-05,
+      "loss": 2.045,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5273518562316895,
+      "learning_rate": 1.8894371345029243e-05,
+      "loss": 2.3753,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8047032356262207,
+      "learning_rate": 1.8893457602339183e-05,
+      "loss": 1.8399,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.163215637207031,
+      "learning_rate": 1.8892543859649124e-05,
+      "loss": 1.9,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3016839027404785,
+      "learning_rate": 1.8891630116959064e-05,
+      "loss": 2.4318,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.226953983306885,
+      "learning_rate": 1.8890716374269008e-05,
+      "loss": 2.0588,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.143353462219238,
+      "learning_rate": 1.888980263157895e-05,
+      "loss": 1.849,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.307502746582031,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 2.3729,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.447706937789917,
+      "learning_rate": 1.8887975146198833e-05,
+      "loss": 1.7901,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.676396131515503,
+      "learning_rate": 1.8887061403508774e-05,
+      "loss": 2.221,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.771751642227173,
+      "learning_rate": 1.8886147660818714e-05,
+      "loss": 2.0087,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8646907806396484,
+      "learning_rate": 1.8885233918128655e-05,
+      "loss": 2.0492,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.312722682952881,
+      "learning_rate": 1.88843201754386e-05,
+      "loss": 2.4455,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3021955490112305,
+      "learning_rate": 1.888340643274854e-05,
+      "loss": 2.0345,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4053890705108643,
+      "learning_rate": 1.888249269005848e-05,
+      "loss": 2.0004,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.608340263366699,
+      "learning_rate": 1.8881578947368423e-05,
+      "loss": 1.9328,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9883816242218018,
+      "learning_rate": 1.8880665204678364e-05,
+      "loss": 2.1015,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6569955348968506,
+      "learning_rate": 1.8879751461988305e-05,
+      "loss": 1.9948,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8155338764190674,
+      "learning_rate": 1.8878837719298245e-05,
+      "loss": 2.351,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4120020866394043,
+      "learning_rate": 1.887792397660819e-05,
+      "loss": 1.927,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.258161544799805,
+      "learning_rate": 1.887701023391813e-05,
+      "loss": 1.8001,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.757899045944214,
+      "learning_rate": 1.887609649122807e-05,
+      "loss": 2.0777,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.248706817626953,
+      "learning_rate": 1.8875182748538014e-05,
+      "loss": 2.215,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9470319747924805,
+      "learning_rate": 1.8874269005847954e-05,
+      "loss": 1.7326,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.072483062744141,
+      "learning_rate": 1.8873355263157895e-05,
+      "loss": 2.3485,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6576788425445557,
+      "learning_rate": 1.8872441520467836e-05,
+      "loss": 1.9947,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6526615619659424,
+      "learning_rate": 1.887152777777778e-05,
+      "loss": 2.1878,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.47247314453125,
+      "learning_rate": 1.887061403508772e-05,
+      "loss": 1.5772,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.6715312004089355,
+      "learning_rate": 1.886970029239766e-05,
+      "loss": 2.0269,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.142459869384766,
+      "learning_rate": 1.8868786549707604e-05,
+      "loss": 2.371,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.58030366897583,
+      "learning_rate": 1.8867872807017545e-05,
+      "loss": 1.6529,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9417614936828613,
+      "learning_rate": 1.886695906432749e-05,
+      "loss": 2.2491,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.651772975921631,
+      "learning_rate": 1.886604532163743e-05,
+      "loss": 2.4008,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3290977478027344,
+      "learning_rate": 1.886513157894737e-05,
+      "loss": 1.5943,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.430382490158081,
+      "learning_rate": 1.8864217836257314e-05,
+      "loss": 1.8531,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7331933975219727,
+      "learning_rate": 1.8863304093567254e-05,
+      "loss": 2.2528,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5899975299835205,
+      "learning_rate": 1.8862390350877195e-05,
+      "loss": 1.7576,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.78801155090332,
+      "learning_rate": 1.8861476608187135e-05,
+      "loss": 2.3613,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.485844612121582,
+      "learning_rate": 1.886056286549708e-05,
+      "loss": 2.5217,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.293576717376709,
+      "learning_rate": 1.885964912280702e-05,
+      "loss": 1.713,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8360581398010254,
+      "learning_rate": 1.885873538011696e-05,
+      "loss": 1.9357,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7601840496063232,
+      "learning_rate": 1.8857821637426904e-05,
+      "loss": 1.8915,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8597588539123535,
+      "learning_rate": 1.8856907894736845e-05,
+      "loss": 2.0276,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.315775394439697,
+      "learning_rate": 1.8855994152046785e-05,
+      "loss": 2.0851,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5954129695892334,
+      "learning_rate": 1.8855080409356726e-05,
+      "loss": 1.6287,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.255249500274658,
+      "learning_rate": 1.885416666666667e-05,
+      "loss": 1.8501,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.011921405792236,
+      "learning_rate": 1.885325292397661e-05,
+      "loss": 2.3839,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5348594188690186,
+      "learning_rate": 1.885233918128655e-05,
+      "loss": 1.5821,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.067683696746826,
+      "learning_rate": 1.8851425438596495e-05,
+      "loss": 2.6096,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.796114444732666,
+      "learning_rate": 1.8850511695906435e-05,
+      "loss": 2.0407,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.879784107208252,
+      "learning_rate": 1.8849597953216376e-05,
+      "loss": 2.1768,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.245840311050415,
+      "learning_rate": 1.8848684210526316e-05,
+      "loss": 1.9122,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5824108123779297,
+      "learning_rate": 1.884777046783626e-05,
+      "loss": 1.9534,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.700308322906494,
+      "learning_rate": 1.88468567251462e-05,
+      "loss": 2.058,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.364081382751465,
+      "learning_rate": 1.884594298245614e-05,
+      "loss": 2.5896,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.692622184753418,
+      "learning_rate": 1.8845029239766085e-05,
+      "loss": 2.2184,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6238961219787598,
+      "learning_rate": 1.8844115497076026e-05,
+      "loss": 1.9893,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6622474193573,
+      "learning_rate": 1.8843201754385966e-05,
+      "loss": 2.1086,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8726348876953125,
+      "learning_rate": 1.8842288011695907e-05,
+      "loss": 2.3752,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5059752464294434,
+      "learning_rate": 1.884137426900585e-05,
+      "loss": 1.9329,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.321408271789551,
+      "learning_rate": 1.884046052631579e-05,
+      "loss": 1.8702,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7025256156921387,
+      "learning_rate": 1.8839546783625732e-05,
+      "loss": 2.0045,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4326376914978027,
+      "learning_rate": 1.8838633040935676e-05,
+      "loss": 1.7933,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1965012550354004,
+      "learning_rate": 1.8837719298245616e-05,
+      "loss": 1.8102,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.873558282852173,
+      "learning_rate": 1.8836805555555557e-05,
+      "loss": 2.1892,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5769221782684326,
+      "learning_rate": 1.8835891812865497e-05,
+      "loss": 2.2567,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7142059803009033,
+      "learning_rate": 1.883497807017544e-05,
+      "loss": 2.1169,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7580926418304443,
+      "learning_rate": 1.8834064327485382e-05,
+      "loss": 1.9878,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.427607774734497,
+      "learning_rate": 1.8833150584795322e-05,
+      "loss": 1.782,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5653645992279053,
+      "learning_rate": 1.8832236842105266e-05,
+      "loss": 1.851,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9319913387298584,
+      "learning_rate": 1.8831323099415207e-05,
+      "loss": 2.0907,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.592641592025757,
+      "learning_rate": 1.8830409356725147e-05,
+      "loss": 2.0166,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.151914596557617,
+      "learning_rate": 1.8829495614035088e-05,
+      "loss": 1.4751,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6666171550750732,
+      "learning_rate": 1.882858187134503e-05,
+      "loss": 2.0105,
+      "step": 1282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4818623065948486,
+      "learning_rate": 1.8827668128654972e-05,
+      "loss": 1.9362,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.447360992431641,
+      "learning_rate": 1.8826754385964913e-05,
+      "loss": 2.3969,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4069623947143555,
+      "learning_rate": 1.8825840643274857e-05,
+      "loss": 1.8346,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2201669216156006,
+      "learning_rate": 1.8824926900584797e-05,
+      "loss": 1.847,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3543758392333984,
+      "learning_rate": 1.8824013157894738e-05,
+      "loss": 2.0233,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5395774841308594,
+      "learning_rate": 1.8823099415204678e-05,
+      "loss": 2.1734,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.156424045562744,
+      "learning_rate": 1.8822185672514622e-05,
+      "loss": 1.7051,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4344656467437744,
+      "learning_rate": 1.8821271929824563e-05,
+      "loss": 1.7084,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.761688232421875,
+      "learning_rate": 1.8820358187134503e-05,
+      "loss": 2.0467,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.704547882080078,
+      "learning_rate": 1.8819444444444447e-05,
+      "loss": 2.3146,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.293471336364746,
+      "learning_rate": 1.8818530701754388e-05,
+      "loss": 2.3824,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.219223499298096,
+      "learning_rate": 1.8817616959064328e-05,
+      "loss": 1.9311,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8534963130950928,
+      "learning_rate": 1.881670321637427e-05,
+      "loss": 2.0213,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.480644702911377,
+      "learning_rate": 1.8815789473684213e-05,
+      "loss": 2.1556,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.014294624328613,
+      "learning_rate": 1.8814875730994153e-05,
+      "loss": 1.8694,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.66390323638916,
+      "learning_rate": 1.8813961988304094e-05,
+      "loss": 2.1128,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3098442554473877,
+      "learning_rate": 1.8813048245614038e-05,
+      "loss": 2.1994,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8393728733062744,
+      "learning_rate": 1.8812134502923978e-05,
+      "loss": 2.2456,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.756659984588623,
+      "learning_rate": 1.881122076023392e-05,
+      "loss": 2.275,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.659334659576416,
+      "learning_rate": 1.881030701754386e-05,
+      "loss": 2.0197,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7700746059417725,
+      "learning_rate": 1.8809393274853803e-05,
+      "loss": 2.2139,
+      "step": 1303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3587896823883057,
+      "learning_rate": 1.8808479532163744e-05,
+      "loss": 1.7662,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.255063056945801,
+      "learning_rate": 1.8807565789473684e-05,
+      "loss": 2.42,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.69321870803833,
+      "learning_rate": 1.8806652046783628e-05,
+      "loss": 1.7326,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9151110649108887,
+      "learning_rate": 1.880573830409357e-05,
+      "loss": 1.8073,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.020442962646484,
+      "learning_rate": 1.880482456140351e-05,
+      "loss": 1.9495,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.025590896606445,
+      "learning_rate": 1.880391081871345e-05,
+      "loss": 1.8763,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.516815185546875,
+      "learning_rate": 1.8802997076023394e-05,
+      "loss": 1.962,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0324788093566895,
+      "learning_rate": 1.8802083333333334e-05,
+      "loss": 1.9893,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.392843008041382,
+      "learning_rate": 1.8801169590643275e-05,
+      "loss": 1.9206,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.676548719406128,
+      "learning_rate": 1.880025584795322e-05,
+      "loss": 1.401,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.662692070007324,
+      "learning_rate": 1.879934210526316e-05,
+      "loss": 2.1145,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.08876895904541,
+      "learning_rate": 1.87984283625731e-05,
+      "loss": 2.1304,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8529982566833496,
+      "learning_rate": 1.879751461988304e-05,
+      "loss": 2.0105,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7096526622772217,
+      "learning_rate": 1.8796600877192984e-05,
+      "loss": 1.6891,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0712242126464844,
+      "learning_rate": 1.8795687134502925e-05,
+      "loss": 1.7556,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.035948276519775,
+      "learning_rate": 1.8794773391812865e-05,
+      "loss": 2.0941,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3396143913269043,
+      "learning_rate": 1.879385964912281e-05,
+      "loss": 1.7976,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.949076175689697,
+      "learning_rate": 1.879294590643275e-05,
+      "loss": 1.7703,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.872591972351074,
+      "learning_rate": 1.879203216374269e-05,
+      "loss": 1.9925,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.743661642074585,
+      "learning_rate": 1.879111842105263e-05,
+      "loss": 2.2369,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.099400997161865,
+      "learning_rate": 1.8790204678362574e-05,
+      "loss": 2.3841,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3526675701141357,
+      "learning_rate": 1.8789290935672515e-05,
+      "loss": 1.7008,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.966040849685669,
+      "learning_rate": 1.878837719298246e-05,
+      "loss": 2.4052,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7510557174682617,
+      "learning_rate": 1.87874634502924e-05,
+      "loss": 1.9957,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.5986127853393555,
+      "learning_rate": 1.8786549707602343e-05,
+      "loss": 2.4715,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.050223350524902,
+      "learning_rate": 1.8785635964912284e-05,
+      "loss": 2.0581,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.749553918838501,
+      "learning_rate": 1.8784722222222224e-05,
+      "loss": 2.1499,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2598252296447754,
+      "learning_rate": 1.8783808479532165e-05,
+      "loss": 1.9777,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.889674186706543,
+      "learning_rate": 1.878289473684211e-05,
+      "loss": 2.1208,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.151425361633301,
+      "learning_rate": 1.878198099415205e-05,
+      "loss": 1.8812,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1569294929504395,
+      "learning_rate": 1.878106725146199e-05,
+      "loss": 2.2527,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.503685474395752,
+      "learning_rate": 1.8780153508771934e-05,
+      "loss": 2.123,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.955512285232544,
+      "learning_rate": 1.8779239766081874e-05,
+      "loss": 1.9752,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.021307468414307,
+      "learning_rate": 1.8778326023391815e-05,
+      "loss": 2.2169,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2151072025299072,
+      "learning_rate": 1.8777412280701755e-05,
+      "loss": 1.7083,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.426212787628174,
+      "learning_rate": 1.87764985380117e-05,
+      "loss": 2.2605,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9317562580108643,
+      "learning_rate": 1.877558479532164e-05,
+      "loss": 1.9762,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8856468200683594,
+      "learning_rate": 1.877467105263158e-05,
+      "loss": 2.1909,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9630613327026367,
+      "learning_rate": 1.8773757309941524e-05,
+      "loss": 2.6237,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0652689933776855,
+      "learning_rate": 1.8772843567251465e-05,
+      "loss": 2.3774,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3550643920898438,
+      "learning_rate": 1.8771929824561405e-05,
+      "loss": 1.7824,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5247199535369873,
+      "learning_rate": 1.8771016081871346e-05,
+      "loss": 1.9524,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.394230365753174,
+      "learning_rate": 1.877010233918129e-05,
+      "loss": 1.9315,
+      "step": 1346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.190184593200684,
+      "learning_rate": 1.876918859649123e-05,
+      "loss": 2.2925,
+      "step": 1347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.617635488510132,
+      "learning_rate": 1.876827485380117e-05,
+      "loss": 2.246,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4309041500091553,
+      "learning_rate": 1.8767361111111115e-05,
+      "loss": 1.9943,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.665670871734619,
+      "learning_rate": 1.8766447368421055e-05,
+      "loss": 2.3686,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.048494815826416,
+      "learning_rate": 1.8765533625730996e-05,
+      "loss": 1.828,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.504638671875,
+      "learning_rate": 1.8764619883040936e-05,
+      "loss": 2.1397,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.203843116760254,
+      "learning_rate": 1.876370614035088e-05,
+      "loss": 1.8944,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.410677194595337,
+      "learning_rate": 1.876279239766082e-05,
+      "loss": 2.0938,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.074373483657837,
+      "learning_rate": 1.876187865497076e-05,
+      "loss": 1.696,
+      "step": 1355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.226317882537842,
+      "learning_rate": 1.8760964912280705e-05,
+      "loss": 1.7198,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.901991128921509,
+      "learning_rate": 1.8760051169590646e-05,
+      "loss": 1.6433,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9010589122772217,
+      "learning_rate": 1.8759137426900586e-05,
+      "loss": 2.0103,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2058699131011963,
+      "learning_rate": 1.8758223684210527e-05,
+      "loss": 1.71,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.209739923477173,
+      "learning_rate": 1.875730994152047e-05,
+      "loss": 1.7256,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.150022506713867,
+      "learning_rate": 1.875639619883041e-05,
+      "loss": 2.2441,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5766470432281494,
+      "learning_rate": 1.8755482456140352e-05,
+      "loss": 1.808,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6616876125335693,
+      "learning_rate": 1.8754568713450296e-05,
+      "loss": 1.8771,
+      "step": 1363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.333174705505371,
+      "learning_rate": 1.8753654970760236e-05,
+      "loss": 1.9844,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.232007026672363,
+      "learning_rate": 1.8752741228070177e-05,
+      "loss": 2.1689,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4261250495910645,
+      "learning_rate": 1.8751827485380117e-05,
+      "loss": 2.0104,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8575427532196045,
+      "learning_rate": 1.875091374269006e-05,
+      "loss": 1.9049,
+      "step": 1367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9201924800872803,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 2.0018,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.294146537780762,
+      "learning_rate": 1.8749086257309942e-05,
+      "loss": 2.423,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.32291316986084,
+      "learning_rate": 1.8748172514619886e-05,
+      "loss": 1.8936,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.281188488006592,
+      "learning_rate": 1.8747258771929827e-05,
+      "loss": 1.9972,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7568271160125732,
+      "learning_rate": 1.8746345029239767e-05,
+      "loss": 2.2086,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.536242485046387,
+      "learning_rate": 1.8745431286549708e-05,
+      "loss": 2.0257,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6521639823913574,
+      "learning_rate": 1.874451754385965e-05,
+      "loss": 2.2027,
+      "step": 1374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5532405376434326,
+      "learning_rate": 1.8743603801169592e-05,
+      "loss": 1.9537,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8316056728363037,
+      "learning_rate": 1.8742690058479533e-05,
+      "loss": 2.2751,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.600461959838867,
+      "learning_rate": 1.8741776315789477e-05,
+      "loss": 1.788,
+      "step": 1377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5620508193969727,
+      "learning_rate": 1.8740862573099417e-05,
+      "loss": 1.9286,
+      "step": 1378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.527131080627441,
+      "learning_rate": 1.8739948830409358e-05,
+      "loss": 2.0258,
+      "step": 1379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8698856830596924,
+      "learning_rate": 1.8739035087719298e-05,
+      "loss": 2.0399,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9598069190979004,
+      "learning_rate": 1.8738121345029242e-05,
+      "loss": 2.1711,
+      "step": 1381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.044051170349121,
+      "learning_rate": 1.8737207602339183e-05,
+      "loss": 2.2146,
+      "step": 1382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.362630367279053,
+      "learning_rate": 1.8736293859649123e-05,
+      "loss": 2.4813,
+      "step": 1383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.033080816268921,
+      "learning_rate": 1.8735380116959067e-05,
+      "loss": 1.6956,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8003792762756348,
+      "learning_rate": 1.8734466374269008e-05,
+      "loss": 2.2517,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.79931902885437,
+      "learning_rate": 1.8733552631578948e-05,
+      "loss": 1.8621,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6010518074035645,
+      "learning_rate": 1.873263888888889e-05,
+      "loss": 1.9657,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.913949966430664,
+      "learning_rate": 1.8731725146198833e-05,
+      "loss": 1.9614,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.632100820541382,
+      "learning_rate": 1.8730811403508773e-05,
+      "loss": 2.034,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.66853666305542,
+      "learning_rate": 1.8729897660818714e-05,
+      "loss": 2.1001,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4953885078430176,
+      "learning_rate": 1.8728983918128658e-05,
+      "loss": 1.7012,
+      "step": 1391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5763914585113525,
+      "learning_rate": 1.8728070175438598e-05,
+      "loss": 1.7348,
+      "step": 1392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6867687702178955,
+      "learning_rate": 1.872715643274854e-05,
+      "loss": 1.9252,
+      "step": 1393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8585073947906494,
+      "learning_rate": 1.872624269005848e-05,
+      "loss": 1.9551,
+      "step": 1394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1783711910247803,
+      "learning_rate": 1.8725328947368423e-05,
+      "loss": 2.0151,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.802135944366455,
+      "learning_rate": 1.8724415204678364e-05,
+      "loss": 1.8847,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.581038236618042,
+      "learning_rate": 1.8723501461988304e-05,
+      "loss": 2.1644,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.035362958908081,
+      "learning_rate": 1.8722587719298248e-05,
+      "loss": 1.6453,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8902416229248047,
+      "learning_rate": 1.872167397660819e-05,
+      "loss": 1.7428,
+      "step": 1399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8307220935821533,
+      "learning_rate": 1.872076023391813e-05,
+      "loss": 2.2889,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.117844104766846,
+      "learning_rate": 1.871984649122807e-05,
+      "loss": 2.339,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.540463447570801,
+      "learning_rate": 1.8718932748538014e-05,
+      "loss": 2.3102,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4786925315856934,
+      "learning_rate": 1.8718019005847954e-05,
+      "loss": 1.9373,
+      "step": 1403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2613110542297363,
+      "learning_rate": 1.8717105263157895e-05,
+      "loss": 1.4523,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6704256534576416,
+      "learning_rate": 1.8716191520467835e-05,
+      "loss": 2.1022,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.209961891174316,
+      "learning_rate": 1.871527777777778e-05,
+      "loss": 2.1473,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5861165523529053,
+      "learning_rate": 1.871436403508772e-05,
+      "loss": 2.0684,
+      "step": 1407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.328214168548584,
+      "learning_rate": 1.871345029239766e-05,
+      "loss": 2.2815,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.065077781677246,
+      "learning_rate": 1.8712536549707604e-05,
+      "loss": 2.1052,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.031284809112549,
+      "learning_rate": 1.8711622807017545e-05,
+      "loss": 1.9818,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.303991794586182,
+      "learning_rate": 1.871070906432749e-05,
+      "loss": 2.1797,
+      "step": 1411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.951753854751587,
+      "learning_rate": 1.870979532163743e-05,
+      "loss": 2.1192,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7352969646453857,
+      "learning_rate": 1.870888157894737e-05,
+      "loss": 2.1928,
+      "step": 1413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3083927631378174,
+      "learning_rate": 1.8707967836257313e-05,
+      "loss": 1.8527,
+      "step": 1414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.185163497924805,
+      "learning_rate": 1.8707054093567254e-05,
+      "loss": 2.6293,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3683650493621826,
+      "learning_rate": 1.8706140350877194e-05,
+      "loss": 2.3839,
+      "step": 1416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4715874195098877,
+      "learning_rate": 1.870522660818714e-05,
+      "loss": 1.7523,
+      "step": 1417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.837944984436035,
+      "learning_rate": 1.870431286549708e-05,
+      "loss": 1.5644,
+      "step": 1418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6606557369232178,
+      "learning_rate": 1.870339912280702e-05,
+      "loss": 2.0937,
+      "step": 1419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.244344234466553,
+      "learning_rate": 1.870248538011696e-05,
+      "loss": 2.1634,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.080252170562744,
+      "learning_rate": 1.8701571637426904e-05,
+      "loss": 2.2721,
+      "step": 1421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7673652172088623,
+      "learning_rate": 1.8700657894736844e-05,
+      "loss": 2.2224,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1705164909362793,
+      "learning_rate": 1.8699744152046785e-05,
+      "loss": 1.9648,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7971885204315186,
+      "learning_rate": 1.869883040935673e-05,
+      "loss": 1.6794,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7603812217712402,
+      "learning_rate": 1.869791666666667e-05,
+      "loss": 2.072,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4197194576263428,
+      "learning_rate": 1.869700292397661e-05,
+      "loss": 1.7777,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.94803524017334,
+      "learning_rate": 1.869608918128655e-05,
+      "loss": 2.4459,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.200811862945557,
+      "learning_rate": 1.8695175438596494e-05,
+      "loss": 2.2819,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5536630153656006,
+      "learning_rate": 1.8694261695906435e-05,
+      "loss": 2.1669,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.663318634033203,
+      "learning_rate": 1.8693347953216375e-05,
+      "loss": 1.9445,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.107954740524292,
+      "learning_rate": 1.869243421052632e-05,
+      "loss": 1.9131,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.926631212234497,
+      "learning_rate": 1.869152046783626e-05,
+      "loss": 1.7328,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.369140148162842,
+      "learning_rate": 1.86906067251462e-05,
+      "loss": 2.1425,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8906097412109375,
+      "learning_rate": 1.868969298245614e-05,
+      "loss": 2.0782,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.476832389831543,
+      "learning_rate": 1.8688779239766085e-05,
+      "loss": 2.111,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.861602544784546,
+      "learning_rate": 1.8687865497076025e-05,
+      "loss": 2.2483,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.954944610595703,
+      "learning_rate": 1.8686951754385966e-05,
+      "loss": 2.0743,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.809142589569092,
+      "learning_rate": 1.868603801169591e-05,
+      "loss": 2.1933,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.947739362716675,
+      "learning_rate": 1.868512426900585e-05,
+      "loss": 1.5234,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8107151985168457,
+      "learning_rate": 1.868421052631579e-05,
+      "loss": 2.2422,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.35613489151001,
+      "learning_rate": 1.868329678362573e-05,
+      "loss": 2.0473,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5232110023498535,
+      "learning_rate": 1.8682383040935675e-05,
+      "loss": 2.5432,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.334928274154663,
+      "learning_rate": 1.8681469298245616e-05,
+      "loss": 1.8269,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6575372219085693,
+      "learning_rate": 1.8680555555555556e-05,
+      "loss": 1.7533,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.85381817817688,
+      "learning_rate": 1.86796418128655e-05,
+      "loss": 1.8748,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3969528675079346,
+      "learning_rate": 1.867872807017544e-05,
+      "loss": 1.886,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8160958290100098,
+      "learning_rate": 1.867781432748538e-05,
+      "loss": 2.1281,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3484976291656494,
+      "learning_rate": 1.8676900584795322e-05,
+      "loss": 1.8425,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.14200758934021,
+      "learning_rate": 1.8675986842105266e-05,
+      "loss": 1.6394,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.130239963531494,
+      "learning_rate": 1.8675073099415206e-05,
+      "loss": 1.8495,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.887519121170044,
+      "learning_rate": 1.8674159356725147e-05,
+      "loss": 1.9669,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.098743438720703,
+      "learning_rate": 1.867324561403509e-05,
+      "loss": 1.9042,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.849604845046997,
+      "learning_rate": 1.867233187134503e-05,
+      "loss": 1.9597,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.132065773010254,
+      "learning_rate": 1.8671418128654972e-05,
+      "loss": 2.1823,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.814635992050171,
+      "learning_rate": 1.8670504385964912e-05,
+      "loss": 2.372,
+      "step": 1455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.673767328262329,
+      "learning_rate": 1.8669590643274856e-05,
+      "loss": 1.9754,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.749140501022339,
+      "learning_rate": 1.8668676900584797e-05,
+      "loss": 2.2556,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9897944927215576,
+      "learning_rate": 1.8667763157894737e-05,
+      "loss": 2.1915,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3330495357513428,
+      "learning_rate": 1.866684941520468e-05,
+      "loss": 1.9286,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.699124336242676,
+      "learning_rate": 1.8665935672514622e-05,
+      "loss": 1.8281,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3129191398620605,
+      "learning_rate": 1.8665021929824562e-05,
+      "loss": 2.0984,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.175876140594482,
+      "learning_rate": 1.8664108187134503e-05,
+      "loss": 2.1127,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.283079624176025,
+      "learning_rate": 1.8663194444444447e-05,
+      "loss": 2.0589,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.559326410293579,
+      "learning_rate": 1.8662280701754387e-05,
+      "loss": 1.7644,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8133556842803955,
+      "learning_rate": 1.8661366959064328e-05,
+      "loss": 2.0164,
+      "step": 1465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4182305335998535,
+      "learning_rate": 1.866045321637427e-05,
+      "loss": 1.8684,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.798362731933594,
+      "learning_rate": 1.8659539473684212e-05,
+      "loss": 1.9229,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.605001926422119,
+      "learning_rate": 1.8658625730994153e-05,
+      "loss": 2.0899,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7254388332366943,
+      "learning_rate": 1.8657711988304093e-05,
+      "loss": 2.1664,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.645408868789673,
+      "learning_rate": 1.8656798245614037e-05,
+      "loss": 1.9712,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8195531368255615,
+      "learning_rate": 1.8655884502923978e-05,
+      "loss": 2.4114,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.633073091506958,
+      "learning_rate": 1.8654970760233918e-05,
+      "loss": 2.5698,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3353521823883057,
+      "learning_rate": 1.8654057017543862e-05,
+      "loss": 2.0268,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7253425121307373,
+      "learning_rate": 1.8653143274853803e-05,
+      "loss": 2.1954,
+      "step": 1474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5395305156707764,
+      "learning_rate": 1.8652229532163743e-05,
+      "loss": 2.094,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.435170650482178,
+      "learning_rate": 1.8651315789473684e-05,
+      "loss": 2.218,
+      "step": 1476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6572515964508057,
+      "learning_rate": 1.8650402046783628e-05,
+      "loss": 2.1088,
+      "step": 1477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1918649673461914,
+      "learning_rate": 1.8649488304093568e-05,
+      "loss": 1.8245,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.837812662124634,
+      "learning_rate": 1.864857456140351e-05,
+      "loss": 1.9632,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6902565956115723,
+      "learning_rate": 1.8647660818713453e-05,
+      "loss": 2.0682,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.525224685668945,
+      "learning_rate": 1.8646747076023393e-05,
+      "loss": 2.1152,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.472782135009766,
+      "learning_rate": 1.8645833333333334e-05,
+      "loss": 2.1382,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.550694465637207,
+      "learning_rate": 1.8644919590643274e-05,
+      "loss": 2.7198,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.221376419067383,
+      "learning_rate": 1.8644005847953218e-05,
+      "loss": 2.2775,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.733400583267212,
+      "learning_rate": 1.864309210526316e-05,
+      "loss": 1.7113,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4384796619415283,
+      "learning_rate": 1.86421783625731e-05,
+      "loss": 1.9805,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.899308681488037,
+      "learning_rate": 1.8641264619883043e-05,
+      "loss": 1.6511,
+      "step": 1487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.80015230178833,
+      "learning_rate": 1.8640350877192984e-05,
+      "loss": 1.9431,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3498013019561768,
+      "learning_rate": 1.8639437134502924e-05,
+      "loss": 2.0251,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.47094988822937,
+      "learning_rate": 1.8638523391812865e-05,
+      "loss": 1.8213,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5363476276397705,
+      "learning_rate": 1.863760964912281e-05,
+      "loss": 1.7241,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.648860931396484,
+      "learning_rate": 1.863669590643275e-05,
+      "loss": 2.0562,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.102911472320557,
+      "learning_rate": 1.863578216374269e-05,
+      "loss": 2.4753,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6873791217803955,
+      "learning_rate": 1.8634868421052634e-05,
+      "loss": 2.1788,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1106767654418945,
+      "learning_rate": 1.8633954678362574e-05,
+      "loss": 1.5125,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8232109546661377,
+      "learning_rate": 1.8633040935672515e-05,
+      "loss": 1.9309,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.657003879547119,
+      "learning_rate": 1.863212719298246e-05,
+      "loss": 1.8372,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.03097677230835,
+      "learning_rate": 1.86312134502924e-05,
+      "loss": 1.7741,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0657799243927,
+      "learning_rate": 1.8630299707602343e-05,
+      "loss": 1.8872,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.219655990600586,
+      "learning_rate": 1.8629385964912283e-05,
+      "loss": 2.048,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.018691062927246,
+      "learning_rate": 1.8628472222222224e-05,
+      "loss": 2.013,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.497490882873535,
+      "learning_rate": 1.8627558479532165e-05,
+      "loss": 1.9468,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.905618190765381,
+      "learning_rate": 1.862664473684211e-05,
+      "loss": 2.2187,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2937467098236084,
+      "learning_rate": 1.862573099415205e-05,
+      "loss": 1.9406,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.546682357788086,
+      "learning_rate": 1.862481725146199e-05,
+      "loss": 1.9997,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.243808269500732,
+      "learning_rate": 1.8623903508771933e-05,
+      "loss": 2.2201,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.328382968902588,
+      "learning_rate": 1.8622989766081874e-05,
+      "loss": 2.083,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.691594362258911,
+      "learning_rate": 1.8622076023391814e-05,
+      "loss": 1.9839,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.117975234985352,
+      "learning_rate": 1.8621162280701755e-05,
+      "loss": 2.4642,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8392446041107178,
+      "learning_rate": 1.86202485380117e-05,
+      "loss": 2.1082,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.450137615203857,
+      "learning_rate": 1.861933479532164e-05,
+      "loss": 1.7886,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.577496528625488,
+      "learning_rate": 1.861842105263158e-05,
+      "loss": 2.3593,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9700655937194824,
+      "learning_rate": 1.8617507309941524e-05,
+      "loss": 1.7542,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4655048847198486,
+      "learning_rate": 1.8616593567251464e-05,
+      "loss": 1.7352,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.206162214279175,
+      "learning_rate": 1.8615679824561405e-05,
+      "loss": 2.1697,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7644948959350586,
+      "learning_rate": 1.8614766081871345e-05,
+      "loss": 1.7601,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.324897050857544,
+      "learning_rate": 1.861385233918129e-05,
+      "loss": 1.6816,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9944465160369873,
+      "learning_rate": 1.861293859649123e-05,
+      "loss": 2.1956,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3850364685058594,
+      "learning_rate": 1.861202485380117e-05,
+      "loss": 1.9782,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.897259473800659,
+      "learning_rate": 1.8611111111111114e-05,
+      "loss": 1.8554,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5077879428863525,
+      "learning_rate": 1.8610197368421055e-05,
+      "loss": 1.8417,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7920398712158203,
+      "learning_rate": 1.8609283625730995e-05,
+      "loss": 1.9359,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.482002258300781,
+      "learning_rate": 1.8608369883040936e-05,
+      "loss": 2.4543,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1171932220458984,
+      "learning_rate": 1.860745614035088e-05,
+      "loss": 1.7838,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.462120294570923,
+      "learning_rate": 1.860654239766082e-05,
+      "loss": 1.9559,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.234951972961426,
+      "learning_rate": 1.860562865497076e-05,
+      "loss": 1.8813,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.398074150085449,
+      "learning_rate": 1.8604714912280705e-05,
+      "loss": 1.8597,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.203364849090576,
+      "learning_rate": 1.8603801169590645e-05,
+      "loss": 1.5409,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8564419746398926,
+      "learning_rate": 1.8602887426900586e-05,
+      "loss": 1.8805,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5041351318359375,
+      "learning_rate": 1.8601973684210526e-05,
+      "loss": 1.9341,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.393110275268555,
+      "learning_rate": 1.860105994152047e-05,
+      "loss": 2.3226,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.288942813873291,
+      "learning_rate": 1.860014619883041e-05,
+      "loss": 1.6551,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.388185739517212,
+      "learning_rate": 1.859923245614035e-05,
+      "loss": 2.0536,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5301692485809326,
+      "learning_rate": 1.8598318713450295e-05,
+      "loss": 1.82,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.310465812683105,
+      "learning_rate": 1.8597404970760236e-05,
+      "loss": 2.2797,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.067268133163452,
+      "learning_rate": 1.8596491228070176e-05,
+      "loss": 1.7983,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3038384914398193,
+      "learning_rate": 1.8595577485380117e-05,
+      "loss": 1.6884,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4317800998687744,
+      "learning_rate": 1.859466374269006e-05,
+      "loss": 1.6892,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.674373149871826,
+      "learning_rate": 1.859375e-05,
+      "loss": 2.2711,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2561233043670654,
+      "learning_rate": 1.8592836257309942e-05,
+      "loss": 1.8763,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.423853874206543,
+      "learning_rate": 1.8591922514619886e-05,
+      "loss": 2.1379,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9273695945739746,
+      "learning_rate": 1.8591008771929826e-05,
+      "loss": 1.6214,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.235342979431152,
+      "learning_rate": 1.8590095029239767e-05,
+      "loss": 1.7766,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.529451370239258,
+      "learning_rate": 1.8589181286549707e-05,
+      "loss": 2.0851,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.963736057281494,
+      "learning_rate": 1.858826754385965e-05,
+      "loss": 1.7865,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3949553966522217,
+      "learning_rate": 1.8587353801169592e-05,
+      "loss": 1.875,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.674971103668213,
+      "learning_rate": 1.8586440058479532e-05,
+      "loss": 1.8063,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0290422439575195,
+      "learning_rate": 1.8585526315789476e-05,
+      "loss": 1.7697,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.633639335632324,
+      "learning_rate": 1.8584612573099417e-05,
+      "loss": 2.118,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.092318534851074,
+      "learning_rate": 1.8583698830409357e-05,
+      "loss": 1.9077,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1006321907043457,
+      "learning_rate": 1.8582785087719298e-05,
+      "loss": 1.6726,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.20955228805542,
+      "learning_rate": 1.8581871345029242e-05,
+      "loss": 2.0768,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.510369300842285,
+      "learning_rate": 1.8580957602339182e-05,
+      "loss": 1.6361,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.817087173461914,
+      "learning_rate": 1.8580043859649123e-05,
+      "loss": 1.8843,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2481460571289062,
+      "learning_rate": 1.8579130116959067e-05,
+      "loss": 2.0103,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8246212005615234,
+      "learning_rate": 1.8578216374269007e-05,
+      "loss": 1.9397,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2211763858795166,
+      "learning_rate": 1.8577302631578948e-05,
+      "loss": 1.5788,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.448509216308594,
+      "learning_rate": 1.8576388888888888e-05,
+      "loss": 2.2134,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.286195278167725,
+      "learning_rate": 1.8575475146198832e-05,
+      "loss": 2.2805,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2683908939361572,
+      "learning_rate": 1.8574561403508773e-05,
+      "loss": 1.7245,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.488153457641602,
+      "learning_rate": 1.8573647660818713e-05,
+      "loss": 2.8235,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.018533229827881,
+      "learning_rate": 1.8572733918128657e-05,
+      "loss": 2.2204,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.879943609237671,
+      "learning_rate": 1.8571820175438598e-05,
+      "loss": 1.8226,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.471980094909668,
+      "learning_rate": 1.8570906432748538e-05,
+      "loss": 1.6661,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.671337127685547,
+      "learning_rate": 1.856999269005848e-05,
+      "loss": 1.9171,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3071558475494385,
+      "learning_rate": 1.8569078947368423e-05,
+      "loss": 1.6983,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6124284267425537,
+      "learning_rate": 1.8568165204678363e-05,
+      "loss": 1.9869,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2777440547943115,
+      "learning_rate": 1.8567251461988304e-05,
+      "loss": 1.7096,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5701491832733154,
+      "learning_rate": 1.8566337719298248e-05,
+      "loss": 1.6457,
+      "step": 1569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3860995769500732,
+      "learning_rate": 1.8565423976608188e-05,
+      "loss": 1.9472,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1221699714660645,
+      "learning_rate": 1.856451023391813e-05,
+      "loss": 1.83,
+      "step": 1571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.186338424682617,
+      "learning_rate": 1.856359649122807e-05,
+      "loss": 2.5383,
+      "step": 1572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.768634796142578,
+      "learning_rate": 1.8562682748538013e-05,
+      "loss": 1.7701,
+      "step": 1573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4330368041992188,
+      "learning_rate": 1.8561769005847954e-05,
+      "loss": 1.864,
+      "step": 1574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8041837215423584,
+      "learning_rate": 1.8560855263157894e-05,
+      "loss": 2.0254,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1046338081359863,
+      "learning_rate": 1.8559941520467838e-05,
+      "loss": 1.5731,
+      "step": 1576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8530139923095703,
+      "learning_rate": 1.855902777777778e-05,
+      "loss": 1.9944,
+      "step": 1577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.660963773727417,
+      "learning_rate": 1.855811403508772e-05,
+      "loss": 1.9483,
+      "step": 1578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.786369800567627,
+      "learning_rate": 1.855720029239766e-05,
+      "loss": 1.9778,
+      "step": 1579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1199517250061035,
+      "learning_rate": 1.8556286549707604e-05,
+      "loss": 1.7882,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3387629985809326,
+      "learning_rate": 1.8555372807017544e-05,
+      "loss": 1.5126,
+      "step": 1581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8365657329559326,
+      "learning_rate": 1.8554459064327488e-05,
+      "loss": 1.9241,
+      "step": 1582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.412703037261963,
+      "learning_rate": 1.855354532163743e-05,
+      "loss": 1.997,
+      "step": 1583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.606473684310913,
+      "learning_rate": 1.8552631578947373e-05,
+      "loss": 2.1094,
+      "step": 1584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.214771270751953,
+      "learning_rate": 1.8551717836257313e-05,
+      "loss": 1.8776,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.092606544494629,
+      "learning_rate": 1.8550804093567254e-05,
+      "loss": 2.4554,
+      "step": 1586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2128243446350098,
+      "learning_rate": 1.8549890350877194e-05,
+      "loss": 1.7138,
+      "step": 1587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1826910972595215,
+      "learning_rate": 1.8548976608187138e-05,
+      "loss": 2.0662,
+      "step": 1588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.341468095779419,
+      "learning_rate": 1.854806286549708e-05,
+      "loss": 2.0345,
+      "step": 1589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3512041568756104,
+      "learning_rate": 1.854714912280702e-05,
+      "loss": 1.6501,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7027430534362793,
+      "learning_rate": 1.8546235380116963e-05,
+      "loss": 2.2633,
+      "step": 1591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.601274013519287,
+      "learning_rate": 1.8545321637426903e-05,
+      "loss": 2.213,
+      "step": 1592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6338326930999756,
+      "learning_rate": 1.8544407894736844e-05,
+      "loss": 2.1961,
+      "step": 1593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7787904739379883,
+      "learning_rate": 1.8543494152046785e-05,
+      "loss": 1.6046,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9766156673431396,
+      "learning_rate": 1.854258040935673e-05,
+      "loss": 1.614,
+      "step": 1595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4926328659057617,
+      "learning_rate": 1.854166666666667e-05,
+      "loss": 2.0544,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1616859436035156,
+      "learning_rate": 1.854075292397661e-05,
+      "loss": 1.6159,
+      "step": 1597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.491739511489868,
+      "learning_rate": 1.8539839181286553e-05,
+      "loss": 1.6426,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.534250736236572,
+      "learning_rate": 1.8538925438596494e-05,
+      "loss": 1.8567,
+      "step": 1599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.058607816696167,
+      "learning_rate": 1.8538011695906434e-05,
+      "loss": 1.4512,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9621734619140625,
+      "learning_rate": 1.8537097953216375e-05,
+      "loss": 2.2893,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1274020671844482,
+      "learning_rate": 1.853618421052632e-05,
+      "loss": 1.7229,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2935850620269775,
+      "learning_rate": 1.853527046783626e-05,
+      "loss": 1.7854,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8116228580474854,
+      "learning_rate": 1.85343567251462e-05,
+      "loss": 2.1388,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.05990743637085,
+      "learning_rate": 1.8533442982456144e-05,
+      "loss": 2.3231,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9033098220825195,
+      "learning_rate": 1.8532529239766084e-05,
+      "loss": 1.8833,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8317549228668213,
+      "learning_rate": 1.8531615497076025e-05,
+      "loss": 2.121,
+      "step": 1607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4989137649536133,
+      "learning_rate": 1.8530701754385965e-05,
+      "loss": 1.723,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6533212661743164,
+      "learning_rate": 1.852978801169591e-05,
+      "loss": 2.0752,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.891177177429199,
+      "learning_rate": 1.852887426900585e-05,
+      "loss": 2.1646,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5202980041503906,
+      "learning_rate": 1.852796052631579e-05,
+      "loss": 1.5981,
+      "step": 1611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.630619525909424,
+      "learning_rate": 1.8527046783625734e-05,
+      "loss": 1.8167,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2419795989990234,
+      "learning_rate": 1.8526133040935675e-05,
+      "loss": 1.9018,
+      "step": 1613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.977834463119507,
+      "learning_rate": 1.8525219298245615e-05,
+      "loss": 1.6616,
+      "step": 1614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0528266429901123,
+      "learning_rate": 1.8524305555555556e-05,
+      "loss": 1.9431,
+      "step": 1615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.104396343231201,
+      "learning_rate": 1.85233918128655e-05,
+      "loss": 2.1453,
+      "step": 1616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6492059230804443,
+      "learning_rate": 1.852247807017544e-05,
+      "loss": 1.8865,
+      "step": 1617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.947937488555908,
+      "learning_rate": 1.852156432748538e-05,
+      "loss": 1.7773,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.163331508636475,
+      "learning_rate": 1.8520650584795325e-05,
+      "loss": 1.8797,
+      "step": 1619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.726156234741211,
+      "learning_rate": 1.8519736842105265e-05,
+      "loss": 1.9791,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2833356857299805,
+      "learning_rate": 1.8518823099415206e-05,
+      "loss": 1.876,
+      "step": 1621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.675936222076416,
+      "learning_rate": 1.8517909356725146e-05,
+      "loss": 2.2239,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7171261310577393,
+      "learning_rate": 1.851699561403509e-05,
+      "loss": 1.3717,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.14286470413208,
+      "learning_rate": 1.851608187134503e-05,
+      "loss": 1.7227,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.245605230331421,
+      "learning_rate": 1.851516812865497e-05,
+      "loss": 1.6145,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5272109508514404,
+      "learning_rate": 1.8514254385964915e-05,
+      "loss": 1.7194,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.432515621185303,
+      "learning_rate": 1.8513340643274856e-05,
+      "loss": 2.4243,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.700162410736084,
+      "learning_rate": 1.8512426900584796e-05,
+      "loss": 2.0743,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.576237678527832,
+      "learning_rate": 1.8511513157894737e-05,
+      "loss": 1.9539,
+      "step": 1629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7110352516174316,
+      "learning_rate": 1.851059941520468e-05,
+      "loss": 2.0453,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.143970966339111,
+      "learning_rate": 1.850968567251462e-05,
+      "loss": 2.1825,
+      "step": 1631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.49237322807312,
+      "learning_rate": 1.8508771929824562e-05,
+      "loss": 2.0259,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0059525966644287,
+      "learning_rate": 1.8507858187134506e-05,
+      "loss": 1.759,
+      "step": 1633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7702651023864746,
+      "learning_rate": 1.8506944444444446e-05,
+      "loss": 1.9149,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.476936340332031,
+      "learning_rate": 1.8506030701754387e-05,
+      "loss": 1.9537,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.045367479324341,
+      "learning_rate": 1.8505116959064327e-05,
+      "loss": 1.6521,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.58015775680542,
+      "learning_rate": 1.850420321637427e-05,
+      "loss": 2.1456,
+      "step": 1637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6031103134155273,
+      "learning_rate": 1.8503289473684212e-05,
+      "loss": 1.9646,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4145662784576416,
+      "learning_rate": 1.8502375730994152e-05,
+      "loss": 1.7404,
+      "step": 1639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.245190143585205,
+      "learning_rate": 1.8501461988304096e-05,
+      "loss": 2.1479,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5101277828216553,
+      "learning_rate": 1.8500548245614037e-05,
+      "loss": 2.2812,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.074889898300171,
+      "learning_rate": 1.8499634502923977e-05,
+      "loss": 1.7688,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.803267478942871,
+      "learning_rate": 1.8498720760233918e-05,
+      "loss": 2.232,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5760626792907715,
+      "learning_rate": 1.8497807017543862e-05,
+      "loss": 2.1477,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.618774652481079,
+      "learning_rate": 1.8496893274853802e-05,
+      "loss": 1.7257,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3541722297668457,
+      "learning_rate": 1.8495979532163743e-05,
+      "loss": 1.7655,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.278564453125,
+      "learning_rate": 1.8495065789473687e-05,
+      "loss": 1.9919,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1264960765838623,
+      "learning_rate": 1.8494152046783627e-05,
+      "loss": 1.8374,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6521315574645996,
+      "learning_rate": 1.8493238304093568e-05,
+      "loss": 1.8101,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4213805198669434,
+      "learning_rate": 1.8492324561403508e-05,
+      "loss": 1.6897,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.043991804122925,
+      "learning_rate": 1.8491410818713452e-05,
+      "loss": 1.5382,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.755777597427368,
+      "learning_rate": 1.8490497076023393e-05,
+      "loss": 2.2396,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1796765327453613,
+      "learning_rate": 1.8489583333333333e-05,
+      "loss": 1.726,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.379009485244751,
+      "learning_rate": 1.8488669590643277e-05,
+      "loss": 1.9676,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.385437488555908,
+      "learning_rate": 1.8487755847953218e-05,
+      "loss": 1.9594,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4825546741485596,
+      "learning_rate": 1.8486842105263158e-05,
+      "loss": 1.9431,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.210363864898682,
+      "learning_rate": 1.84859283625731e-05,
+      "loss": 2.5964,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0847954750061035,
+      "learning_rate": 1.8485014619883043e-05,
+      "loss": 2.3684,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.263486385345459,
+      "learning_rate": 1.8484100877192983e-05,
+      "loss": 1.8675,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8864102363586426,
+      "learning_rate": 1.8483187134502924e-05,
+      "loss": 1.8969,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6319169998168945,
+      "learning_rate": 1.8482273391812868e-05,
+      "loss": 2.456,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4051952362060547,
+      "learning_rate": 1.8481359649122808e-05,
+      "loss": 1.6654,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.752854108810425,
+      "learning_rate": 1.848044590643275e-05,
+      "loss": 2.2912,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7268948554992676,
+      "learning_rate": 1.847953216374269e-05,
+      "loss": 2.2414,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0970985889434814,
+      "learning_rate": 1.8478618421052633e-05,
+      "loss": 1.6418,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1906328201293945,
+      "learning_rate": 1.8477704678362574e-05,
+      "loss": 2.0425,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5523180961608887,
+      "learning_rate": 1.8476790935672514e-05,
+      "loss": 2.2154,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.668957471847534,
+      "learning_rate": 1.8475877192982458e-05,
+      "loss": 2.4299,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.957003593444824,
+      "learning_rate": 1.84749634502924e-05,
+      "loss": 1.7825,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2317276000976562,
+      "learning_rate": 1.8474049707602343e-05,
+      "loss": 1.591,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.439785957336426,
+      "learning_rate": 1.8473135964912283e-05,
+      "loss": 1.9171,
+      "step": 1671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6993348598480225,
+      "learning_rate": 1.8472222222222224e-05,
+      "loss": 2.1677,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.962912082672119,
+      "learning_rate": 1.8471308479532168e-05,
+      "loss": 1.6644,
+      "step": 1673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.286872386932373,
+      "learning_rate": 1.8470394736842108e-05,
+      "loss": 2.3408,
+      "step": 1674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1766834259033203,
+      "learning_rate": 1.846948099415205e-05,
+      "loss": 1.5816,
+      "step": 1675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6634113788604736,
+      "learning_rate": 1.846856725146199e-05,
+      "loss": 2.0285,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.005817890167236,
+      "learning_rate": 1.8467653508771933e-05,
+      "loss": 1.9706,
+      "step": 1677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4942197799682617,
+      "learning_rate": 1.8466739766081874e-05,
+      "loss": 2.0524,
+      "step": 1678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3664586544036865,
+      "learning_rate": 1.8465826023391814e-05,
+      "loss": 1.5467,
+      "step": 1679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9279770851135254,
+      "learning_rate": 1.8464912280701758e-05,
+      "loss": 1.8266,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.474382162094116,
+      "learning_rate": 1.84639985380117e-05,
+      "loss": 1.721,
+      "step": 1681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3359880447387695,
+      "learning_rate": 1.846308479532164e-05,
+      "loss": 1.9278,
+      "step": 1682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.095881462097168,
+      "learning_rate": 1.846217105263158e-05,
+      "loss": 1.6548,
+      "step": 1683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.949612855911255,
+      "learning_rate": 1.8461257309941523e-05,
+      "loss": 1.5333,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.416017532348633,
+      "learning_rate": 1.8460343567251464e-05,
+      "loss": 1.7741,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.455709457397461,
+      "learning_rate": 1.8459429824561405e-05,
+      "loss": 1.7668,
+      "step": 1686
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.10662317276001,
+      "learning_rate": 1.845851608187135e-05,
+      "loss": 2.0351,
+      "step": 1687
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.629863977432251,
+      "learning_rate": 1.845760233918129e-05,
+      "loss": 1.7967,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1737310886383057,
+      "learning_rate": 1.845668859649123e-05,
+      "loss": 1.7993,
+      "step": 1689
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5820236206054688,
+      "learning_rate": 1.845577485380117e-05,
+      "loss": 2.0296,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7759814262390137,
+      "learning_rate": 1.8454861111111114e-05,
+      "loss": 1.689,
+      "step": 1691
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5932183265686035,
+      "learning_rate": 1.8453947368421054e-05,
+      "loss": 1.9475,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7904670238494873,
+      "learning_rate": 1.8453033625730995e-05,
+      "loss": 1.9338,
+      "step": 1693
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.377399206161499,
+      "learning_rate": 1.845211988304094e-05,
+      "loss": 1.72,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.015493869781494,
+      "learning_rate": 1.845120614035088e-05,
+      "loss": 1.8874,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.226555109024048,
+      "learning_rate": 1.845029239766082e-05,
+      "loss": 2.0257,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4996888637542725,
+      "learning_rate": 1.844937865497076e-05,
+      "loss": 1.7545,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.685644626617432,
+      "learning_rate": 1.8448464912280704e-05,
+      "loss": 2.3489,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.655076026916504,
+      "learning_rate": 1.8447551169590645e-05,
+      "loss": 1.9995,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.926061630249023,
+      "learning_rate": 1.8446637426900585e-05,
+      "loss": 2.3882,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9418179988861084,
+      "learning_rate": 1.844572368421053e-05,
+      "loss": 2.0465,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.563292503356934,
+      "learning_rate": 1.844480994152047e-05,
+      "loss": 2.5639,
+      "step": 1702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.351402759552002,
+      "learning_rate": 1.844389619883041e-05,
+      "loss": 1.7954,
+      "step": 1703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6497585773468018,
+      "learning_rate": 1.844298245614035e-05,
+      "loss": 2.1045,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.455808401107788,
+      "learning_rate": 1.8442068713450295e-05,
+      "loss": 1.6714,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5250425338745117,
+      "learning_rate": 1.8441154970760235e-05,
+      "loss": 1.9137,
+      "step": 1706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.650874614715576,
+      "learning_rate": 1.8440241228070176e-05,
+      "loss": 2.0122,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.868781089782715,
+      "learning_rate": 1.843932748538012e-05,
+      "loss": 2.3332,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1715385913848877,
+      "learning_rate": 1.843841374269006e-05,
+      "loss": 1.6533,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7753031253814697,
+      "learning_rate": 1.84375e-05,
+      "loss": 2.3099,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.705357074737549,
+      "learning_rate": 1.843658625730994e-05,
+      "loss": 2.0563,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.290741443634033,
+      "learning_rate": 1.8435672514619885e-05,
+      "loss": 1.9245,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.495229721069336,
+      "learning_rate": 1.8434758771929826e-05,
+      "loss": 2.0168,
+      "step": 1713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.722038507461548,
+      "learning_rate": 1.8433845029239766e-05,
+      "loss": 2.3206,
+      "step": 1714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4892752170562744,
+      "learning_rate": 1.843293128654971e-05,
+      "loss": 1.9011,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.065434217453003,
+      "learning_rate": 1.843201754385965e-05,
+      "loss": 2.0299,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2634711265563965,
+      "learning_rate": 1.843110380116959e-05,
+      "loss": 1.9636,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.213098049163818,
+      "learning_rate": 1.8430190058479532e-05,
+      "loss": 2.0692,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.141603946685791,
+      "learning_rate": 1.8429276315789476e-05,
+      "loss": 2.3638,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.912722587585449,
+      "learning_rate": 1.8428362573099416e-05,
+      "loss": 1.6319,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5037763118743896,
+      "learning_rate": 1.8427448830409357e-05,
+      "loss": 1.8374,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.222700595855713,
+      "learning_rate": 1.84265350877193e-05,
+      "loss": 2.1151,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.111246109008789,
+      "learning_rate": 1.842562134502924e-05,
+      "loss": 1.7883,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.674837589263916,
+      "learning_rate": 1.8424707602339182e-05,
+      "loss": 2.1397,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9646565914154053,
+      "learning_rate": 1.8423793859649122e-05,
+      "loss": 1.8107,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.535853862762451,
+      "learning_rate": 1.8422880116959066e-05,
+      "loss": 1.7194,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.488327980041504,
+      "learning_rate": 1.8421966374269007e-05,
+      "loss": 1.8687,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.054926872253418,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 1.9795,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0503785610198975,
+      "learning_rate": 1.842013888888889e-05,
+      "loss": 1.6356,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.319324016571045,
+      "learning_rate": 1.8419225146198832e-05,
+      "loss": 2.106,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3023343086242676,
+      "learning_rate": 1.8418311403508772e-05,
+      "loss": 1.7887,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8559067249298096,
+      "learning_rate": 1.8417397660818713e-05,
+      "loss": 2.2621,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.627439498901367,
+      "learning_rate": 1.8416483918128657e-05,
+      "loss": 2.106,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.1305060386657715,
+      "learning_rate": 1.8415570175438597e-05,
+      "loss": 2.3403,
+      "step": 1734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9016950130462646,
+      "learning_rate": 1.8414656432748538e-05,
+      "loss": 1.4594,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.566157579421997,
+      "learning_rate": 1.8413742690058482e-05,
+      "loss": 1.6412,
+      "step": 1736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.811051368713379,
+      "learning_rate": 1.8412828947368422e-05,
+      "loss": 1.9835,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.321310997009277,
+      "learning_rate": 1.8411915204678363e-05,
+      "loss": 1.5783,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9080028533935547,
+      "learning_rate": 1.8411001461988303e-05,
+      "loss": 1.8713,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.09164571762085,
+      "learning_rate": 1.8410087719298247e-05,
+      "loss": 2.0294,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.289452075958252,
+      "learning_rate": 1.8409173976608188e-05,
+      "loss": 1.8593,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5328369140625,
+      "learning_rate": 1.8408260233918128e-05,
+      "loss": 1.9208,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9847447872161865,
+      "learning_rate": 1.8407346491228072e-05,
+      "loss": 1.5895,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.704875946044922,
+      "learning_rate": 1.8406432748538013e-05,
+      "loss": 2.0245,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.20236873626709,
+      "learning_rate": 1.8405519005847953e-05,
+      "loss": 2.0521,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.484451770782471,
+      "learning_rate": 1.8404605263157894e-05,
+      "loss": 2.5047,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7198503017425537,
+      "learning_rate": 1.8403691520467838e-05,
+      "loss": 2.0322,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5911002159118652,
+      "learning_rate": 1.8402777777777778e-05,
+      "loss": 2.2075,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.586730718612671,
+      "learning_rate": 1.840186403508772e-05,
+      "loss": 2.0111,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9045169353485107,
+      "learning_rate": 1.8400950292397663e-05,
+      "loss": 2.3517,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.652836322784424,
+      "learning_rate": 1.8400036549707603e-05,
+      "loss": 1.7227,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.855511426925659,
+      "learning_rate": 1.8399122807017544e-05,
+      "loss": 1.8241,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8338561058044434,
+      "learning_rate": 1.8398209064327488e-05,
+      "loss": 2.0627,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.16660475730896,
+      "learning_rate": 1.8397295321637428e-05,
+      "loss": 1.6881,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7037370204925537,
+      "learning_rate": 1.8396381578947372e-05,
+      "loss": 1.829,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8102211952209473,
+      "learning_rate": 1.8395467836257313e-05,
+      "loss": 1.6489,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9428257942199707,
+      "learning_rate": 1.8394554093567253e-05,
+      "loss": 1.4048,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.16597318649292,
+      "learning_rate": 1.8393640350877197e-05,
+      "loss": 2.2953,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.614978313446045,
+      "learning_rate": 1.8392726608187138e-05,
+      "loss": 1.8827,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5799577236175537,
+      "learning_rate": 1.8391812865497078e-05,
+      "loss": 1.7522,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.284323692321777,
+      "learning_rate": 1.839089912280702e-05,
+      "loss": 1.6198,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9395580291748047,
+      "learning_rate": 1.8389985380116963e-05,
+      "loss": 1.2902,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.761319160461426,
+      "learning_rate": 1.8389071637426903e-05,
+      "loss": 1.9004,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7188215255737305,
+      "learning_rate": 1.8388157894736844e-05,
+      "loss": 2.412,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6227805614471436,
+      "learning_rate": 1.8387244152046788e-05,
+      "loss": 1.9048,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5463767051696777,
+      "learning_rate": 1.8386330409356728e-05,
+      "loss": 1.7417,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5776710510253906,
+      "learning_rate": 1.838541666666667e-05,
+      "loss": 1.9234,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9643616676330566,
+      "learning_rate": 1.838450292397661e-05,
+      "loss": 2.1627,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.674381732940674,
+      "learning_rate": 1.8383589181286553e-05,
+      "loss": 1.9665,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1651604175567627,
+      "learning_rate": 1.8382675438596494e-05,
+      "loss": 1.5564,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4558470249176025,
+      "learning_rate": 1.8381761695906434e-05,
+      "loss": 1.5885,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.127411127090454,
+      "learning_rate": 1.8380847953216378e-05,
+      "loss": 1.7781,
+      "step": 1772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.382510662078857,
+      "learning_rate": 1.837993421052632e-05,
+      "loss": 2.1019,
+      "step": 1773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.112536668777466,
+      "learning_rate": 1.837902046783626e-05,
+      "loss": 1.7575,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.674985408782959,
+      "learning_rate": 1.83781067251462e-05,
+      "loss": 1.4545,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.000480890274048,
+      "learning_rate": 1.8377192982456143e-05,
+      "loss": 1.6762,
+      "step": 1776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7127575874328613,
+      "learning_rate": 1.8376279239766084e-05,
+      "loss": 2.4338,
+      "step": 1777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0600452423095703,
+      "learning_rate": 1.8375365497076025e-05,
+      "loss": 1.751,
+      "step": 1778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5195436477661133,
+      "learning_rate": 1.837445175438597e-05,
+      "loss": 1.6432,
+      "step": 1779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.082019329071045,
+      "learning_rate": 1.837353801169591e-05,
+      "loss": 1.8251,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.317196846008301,
+      "learning_rate": 1.837262426900585e-05,
+      "loss": 1.7903,
+      "step": 1781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.657648086547852,
+      "learning_rate": 1.837171052631579e-05,
+      "loss": 2.183,
+      "step": 1782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1200246810913086,
+      "learning_rate": 1.8370796783625734e-05,
+      "loss": 1.8352,
+      "step": 1783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.188619613647461,
+      "learning_rate": 1.8369883040935674e-05,
+      "loss": 1.5145,
+      "step": 1784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.196319580078125,
+      "learning_rate": 1.8368969298245615e-05,
+      "loss": 1.8977,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.442833662033081,
+      "learning_rate": 1.836805555555556e-05,
+      "loss": 1.9042,
+      "step": 1786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.32389760017395,
+      "learning_rate": 1.83671418128655e-05,
+      "loss": 1.7672,
+      "step": 1787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1154050827026367,
+      "learning_rate": 1.836622807017544e-05,
+      "loss": 1.6967,
+      "step": 1788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8209776878356934,
+      "learning_rate": 1.836531432748538e-05,
+      "loss": 1.9085,
+      "step": 1789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6859819889068604,
+      "learning_rate": 1.8364400584795324e-05,
+      "loss": 1.9246,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.994459867477417,
+      "learning_rate": 1.8363486842105265e-05,
+      "loss": 1.5721,
+      "step": 1791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.328281879425049,
+      "learning_rate": 1.8362573099415205e-05,
+      "loss": 2.5134,
+      "step": 1792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.940948963165283,
+      "learning_rate": 1.836165935672515e-05,
+      "loss": 1.5392,
+      "step": 1793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6611690521240234,
+      "learning_rate": 1.836074561403509e-05,
+      "loss": 2.1874,
+      "step": 1794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.059964179992676,
+      "learning_rate": 1.835983187134503e-05,
+      "loss": 2.1172,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4647140502929688,
+      "learning_rate": 1.835891812865497e-05,
+      "loss": 2.3302,
+      "step": 1796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.890833616256714,
+      "learning_rate": 1.8358004385964915e-05,
+      "loss": 1.7021,
+      "step": 1797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9342193603515625,
+      "learning_rate": 1.8357090643274855e-05,
+      "loss": 2.4211,
+      "step": 1798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8897509574890137,
+      "learning_rate": 1.8356176900584796e-05,
+      "loss": 2.2722,
+      "step": 1799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.520047903060913,
+      "learning_rate": 1.835526315789474e-05,
+      "loss": 1.8587,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.670983076095581,
+      "learning_rate": 1.835434941520468e-05,
+      "loss": 2.3948,
+      "step": 1801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.060164213180542,
+      "learning_rate": 1.835343567251462e-05,
+      "loss": 1.599,
+      "step": 1802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1441264152526855,
+      "learning_rate": 1.835252192982456e-05,
+      "loss": 1.6928,
+      "step": 1803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.12995982170105,
+      "learning_rate": 1.8351608187134505e-05,
+      "loss": 1.9041,
+      "step": 1804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.327519416809082,
+      "learning_rate": 1.8350694444444446e-05,
+      "loss": 1.9504,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4131534099578857,
+      "learning_rate": 1.8349780701754386e-05,
+      "loss": 1.857,
+      "step": 1806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.58610200881958,
+      "learning_rate": 1.834886695906433e-05,
+      "loss": 2.0714,
+      "step": 1807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.729541540145874,
+      "learning_rate": 1.834795321637427e-05,
+      "loss": 2.0768,
+      "step": 1808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2278072834014893,
+      "learning_rate": 1.834703947368421e-05,
+      "loss": 1.8123,
+      "step": 1809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.390305042266846,
+      "learning_rate": 1.8346125730994152e-05,
+      "loss": 2.2401,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.927510976791382,
+      "learning_rate": 1.8345211988304096e-05,
+      "loss": 2.1918,
+      "step": 1811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.130406141281128,
+      "learning_rate": 1.8344298245614036e-05,
+      "loss": 1.573,
+      "step": 1812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.39042592048645,
+      "learning_rate": 1.8343384502923977e-05,
+      "loss": 1.7538,
+      "step": 1813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.212217807769775,
+      "learning_rate": 1.834247076023392e-05,
+      "loss": 2.1864,
+      "step": 1814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.243086576461792,
+      "learning_rate": 1.834155701754386e-05,
+      "loss": 1.8698,
+      "step": 1815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5527305603027344,
+      "learning_rate": 1.8340643274853802e-05,
+      "loss": 1.8873,
+      "step": 1816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3310720920562744,
+      "learning_rate": 1.8339729532163742e-05,
+      "loss": 2.1693,
+      "step": 1817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.293595790863037,
+      "learning_rate": 1.8338815789473686e-05,
+      "loss": 2.2445,
+      "step": 1818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.006136655807495,
+      "learning_rate": 1.8337902046783627e-05,
+      "loss": 1.6295,
+      "step": 1819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6906516551971436,
+      "learning_rate": 1.8336988304093567e-05,
+      "loss": 2.296,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0506906509399414,
+      "learning_rate": 1.833607456140351e-05,
+      "loss": 1.6433,
+      "step": 1821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3916234970092773,
+      "learning_rate": 1.8335160818713452e-05,
+      "loss": 1.7828,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.640214443206787,
+      "learning_rate": 1.8334247076023392e-05,
+      "loss": 1.7232,
+      "step": 1823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8052468299865723,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.9525,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.948742389678955,
+      "learning_rate": 1.8332419590643277e-05,
+      "loss": 1.8846,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3020496368408203,
+      "learning_rate": 1.8331505847953217e-05,
+      "loss": 1.517,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4786479473114014,
+      "learning_rate": 1.8330592105263158e-05,
+      "loss": 1.7264,
+      "step": 1827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.283276081085205,
+      "learning_rate": 1.8329678362573102e-05,
+      "loss": 1.6925,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.048499584197998,
+      "learning_rate": 1.8328764619883042e-05,
+      "loss": 1.8402,
+      "step": 1829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7560999393463135,
+      "learning_rate": 1.8327850877192983e-05,
+      "loss": 1.5218,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.032163143157959,
+      "learning_rate": 1.8326937134502923e-05,
+      "loss": 2.475,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.264061689376831,
+      "learning_rate": 1.8326023391812867e-05,
+      "loss": 1.6314,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9800376892089844,
+      "learning_rate": 1.8325109649122808e-05,
+      "loss": 1.9615,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5794012546539307,
+      "learning_rate": 1.8324195906432748e-05,
+      "loss": 2.0127,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2589032649993896,
+      "learning_rate": 1.8323282163742692e-05,
+      "loss": 1.8756,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.616459369659424,
+      "learning_rate": 1.8322368421052633e-05,
+      "loss": 1.7447,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3854124546051025,
+      "learning_rate": 1.8321454678362573e-05,
+      "loss": 1.2088,
+      "step": 1837
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.369399070739746,
+      "learning_rate": 1.8320540935672514e-05,
+      "loss": 1.724,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.032334327697754,
+      "learning_rate": 1.8319627192982458e-05,
+      "loss": 2.1453,
+      "step": 1839
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.152447700500488,
+      "learning_rate": 1.83187134502924e-05,
+      "loss": 2.3791,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6801071166992188,
+      "learning_rate": 1.8317799707602342e-05,
+      "loss": 2.217,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7878775596618652,
+      "learning_rate": 1.8316885964912283e-05,
+      "loss": 1.6119,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.105273723602295,
+      "learning_rate": 1.8315972222222223e-05,
+      "loss": 2.0188,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2480010986328125,
+      "learning_rate": 1.8315058479532167e-05,
+      "loss": 2.1287,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5021939277648926,
+      "learning_rate": 1.8314144736842108e-05,
+      "loss": 1.8702,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4196457862854004,
+      "learning_rate": 1.8313230994152048e-05,
+      "loss": 1.9878,
+      "step": 1846
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0976057052612305,
+      "learning_rate": 1.8312317251461992e-05,
+      "loss": 2.0608,
+      "step": 1847
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6494688987731934,
+      "learning_rate": 1.8311403508771933e-05,
+      "loss": 1.7032,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.804206371307373,
+      "learning_rate": 1.8310489766081873e-05,
+      "loss": 1.9094,
+      "step": 1849
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.642030715942383,
+      "learning_rate": 1.8309576023391814e-05,
+      "loss": 2.1893,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.510610342025757,
+      "learning_rate": 1.8308662280701758e-05,
+      "loss": 2.0122,
+      "step": 1851
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1659767627716064,
+      "learning_rate": 1.8307748538011698e-05,
+      "loss": 1.462,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9443767070770264,
+      "learning_rate": 1.830683479532164e-05,
+      "loss": 2.059,
+      "step": 1853
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.838723659515381,
+      "learning_rate": 1.8305921052631583e-05,
+      "loss": 1.853,
+      "step": 1854
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.174806594848633,
+      "learning_rate": 1.8305007309941523e-05,
+      "loss": 2.3259,
+      "step": 1855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.836319446563721,
+      "learning_rate": 1.8304093567251464e-05,
+      "loss": 1.4047,
+      "step": 1856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.361579656600952,
+      "learning_rate": 1.8303179824561404e-05,
+      "loss": 1.9344,
+      "step": 1857
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3526034355163574,
+      "learning_rate": 1.8302266081871348e-05,
+      "loss": 1.8576,
+      "step": 1858
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.187843084335327,
+      "learning_rate": 1.830135233918129e-05,
+      "loss": 1.8861,
+      "step": 1859
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3893802165985107,
+      "learning_rate": 1.830043859649123e-05,
+      "loss": 1.8995,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5583014488220215,
+      "learning_rate": 1.8299524853801173e-05,
+      "loss": 2.2199,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.788254976272583,
+      "learning_rate": 1.8298611111111114e-05,
+      "loss": 2.1738,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.517437219619751,
+      "learning_rate": 1.8297697368421054e-05,
+      "loss": 1.929,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.709906816482544,
+      "learning_rate": 1.8296783625730995e-05,
+      "loss": 1.3499,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.517996788024902,
+      "learning_rate": 1.829586988304094e-05,
+      "loss": 1.7026,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.351405382156372,
+      "learning_rate": 1.829495614035088e-05,
+      "loss": 2.026,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.408881902694702,
+      "learning_rate": 1.829404239766082e-05,
+      "loss": 2.2088,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.91823673248291,
+      "learning_rate": 1.8293128654970763e-05,
+      "loss": 1.4715,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5493323802948,
+      "learning_rate": 1.8292214912280704e-05,
+      "loss": 2.1571,
+      "step": 1869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8254177570343018,
+      "learning_rate": 1.8291301169590645e-05,
+      "loss": 2.0432,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0372233390808105,
+      "learning_rate": 1.8290387426900585e-05,
+      "loss": 2.1182,
+      "step": 1871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.396136999130249,
+      "learning_rate": 1.828947368421053e-05,
+      "loss": 1.9543,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6252193450927734,
+      "learning_rate": 1.828855994152047e-05,
+      "loss": 1.9172,
+      "step": 1873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1227221488952637,
+      "learning_rate": 1.828764619883041e-05,
+      "loss": 1.5382,
+      "step": 1874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.47740364074707,
+      "learning_rate": 1.8286732456140354e-05,
+      "loss": 1.9819,
+      "step": 1875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1399285793304443,
+      "learning_rate": 1.8285818713450294e-05,
+      "loss": 1.5185,
+      "step": 1876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9958529472351074,
+      "learning_rate": 1.8284904970760235e-05,
+      "loss": 2.1159,
+      "step": 1877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.025847911834717,
+      "learning_rate": 1.8283991228070176e-05,
+      "loss": 1.7027,
+      "step": 1878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1816723346710205,
+      "learning_rate": 1.828307748538012e-05,
+      "loss": 1.8971,
+      "step": 1879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.210463047027588,
+      "learning_rate": 1.828216374269006e-05,
+      "loss": 2.2306,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1570310592651367,
+      "learning_rate": 1.828125e-05,
+      "loss": 1.5204,
+      "step": 1881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5929319858551025,
+      "learning_rate": 1.8280336257309944e-05,
+      "loss": 1.9332,
+      "step": 1882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.447859287261963,
+      "learning_rate": 1.8279422514619885e-05,
+      "loss": 1.8947,
+      "step": 1883
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6880478858947754,
+      "learning_rate": 1.8278508771929825e-05,
+      "loss": 1.9949,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.172726631164551,
+      "learning_rate": 1.8277595029239766e-05,
+      "loss": 1.7888,
+      "step": 1885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.036844730377197,
+      "learning_rate": 1.827668128654971e-05,
+      "loss": 2.2627,
+      "step": 1886
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.773205280303955,
+      "learning_rate": 1.827576754385965e-05,
+      "loss": 1.8915,
+      "step": 1887
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.437415361404419,
+      "learning_rate": 1.827485380116959e-05,
+      "loss": 1.6795,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9283761978149414,
+      "learning_rate": 1.8273940058479535e-05,
+      "loss": 2.0413,
+      "step": 1889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1075315475463867,
+      "learning_rate": 1.8273026315789475e-05,
+      "loss": 1.7225,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.536686420440674,
+      "learning_rate": 1.8272112573099416e-05,
+      "loss": 2.3603,
+      "step": 1891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6313230991363525,
+      "learning_rate": 1.8271198830409356e-05,
+      "loss": 1.8872,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.054316997528076,
+      "learning_rate": 1.82702850877193e-05,
+      "loss": 1.9567,
+      "step": 1893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.002024173736572,
+      "learning_rate": 1.826937134502924e-05,
+      "loss": 2.3786,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5932488441467285,
+      "learning_rate": 1.826845760233918e-05,
+      "loss": 1.7387,
+      "step": 1895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4062421321868896,
+      "learning_rate": 1.8267543859649125e-05,
+      "loss": 1.5572,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2116637229919434,
+      "learning_rate": 1.8266630116959066e-05,
+      "loss": 1.7227,
+      "step": 1897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.69071102142334,
+      "learning_rate": 1.8265716374269006e-05,
+      "loss": 2.0196,
+      "step": 1898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1463406085968018,
+      "learning_rate": 1.8264802631578947e-05,
+      "loss": 1.8065,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.049501895904541,
+      "learning_rate": 1.826388888888889e-05,
+      "loss": 2.1623,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.472548007965088,
+      "learning_rate": 1.826297514619883e-05,
+      "loss": 1.9604,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0247340202331543,
+      "learning_rate": 1.8262061403508772e-05,
+      "loss": 1.5571,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4072420597076416,
+      "learning_rate": 1.8261147660818716e-05,
+      "loss": 1.8069,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8117482662200928,
+      "learning_rate": 1.8260233918128656e-05,
+      "loss": 1.5805,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.479018449783325,
+      "learning_rate": 1.8259320175438597e-05,
+      "loss": 2.0255,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.918611764907837,
+      "learning_rate": 1.8258406432748537e-05,
+      "loss": 1.9637,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8283333778381348,
+      "learning_rate": 1.825749269005848e-05,
+      "loss": 2.1607,
+      "step": 1907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.01801061630249,
+      "learning_rate": 1.8256578947368422e-05,
+      "loss": 1.799,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0352671146392822,
+      "learning_rate": 1.8255665204678362e-05,
+      "loss": 1.6059,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4117588996887207,
+      "learning_rate": 1.8254751461988306e-05,
+      "loss": 1.8793,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.209097385406494,
+      "learning_rate": 1.8253837719298247e-05,
+      "loss": 1.957,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3097193241119385,
+      "learning_rate": 1.8252923976608187e-05,
+      "loss": 1.6944,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5208492279052734,
+      "learning_rate": 1.8252010233918128e-05,
+      "loss": 1.7929,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2870545387268066,
+      "learning_rate": 1.8251096491228072e-05,
+      "loss": 1.4448,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2803120613098145,
+      "learning_rate": 1.8250182748538012e-05,
+      "loss": 1.6459,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3655614852905273,
+      "learning_rate": 1.8249269005847953e-05,
+      "loss": 1.9047,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.812443733215332,
+      "learning_rate": 1.8248355263157897e-05,
+      "loss": 2.5132,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.435551643371582,
+      "learning_rate": 1.8247441520467837e-05,
+      "loss": 2.2109,
+      "step": 1918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.431724786758423,
+      "learning_rate": 1.8246527777777778e-05,
+      "loss": 2.4091,
+      "step": 1919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.351396083831787,
+      "learning_rate": 1.824561403508772e-05,
+      "loss": 2.0125,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9377341270446777,
+      "learning_rate": 1.8244700292397662e-05,
+      "loss": 1.6122,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7304329872131348,
+      "learning_rate": 1.8243786549707603e-05,
+      "loss": 2.4887,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1129660606384277,
+      "learning_rate": 1.8242872807017543e-05,
+      "loss": 1.5804,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4208664894104004,
+      "learning_rate": 1.8241959064327487e-05,
+      "loss": 2.0678,
+      "step": 1924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.91495418548584,
+      "learning_rate": 1.824104532163743e-05,
+      "loss": 1.4255,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.087061405181885,
+      "learning_rate": 1.824013157894737e-05,
+      "loss": 2.5492,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4325919151306152,
+      "learning_rate": 1.8239217836257312e-05,
+      "loss": 2.0616,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.828425407409668,
+      "learning_rate": 1.8238304093567253e-05,
+      "loss": 2.3309,
+      "step": 1928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9485995769500732,
+      "learning_rate": 1.8237390350877197e-05,
+      "loss": 1.8557,
+      "step": 1929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.080631732940674,
+      "learning_rate": 1.8236476608187137e-05,
+      "loss": 2.0827,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.218719959259033,
+      "learning_rate": 1.8235562865497078e-05,
+      "loss": 2.0608,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4744231700897217,
+      "learning_rate": 1.823464912280702e-05,
+      "loss": 2.1538,
+      "step": 1932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.242288112640381,
+      "learning_rate": 1.8233735380116962e-05,
+      "loss": 1.7917,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4467742443084717,
+      "learning_rate": 1.8232821637426903e-05,
+      "loss": 1.6499,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.431375503540039,
+      "learning_rate": 1.8231907894736843e-05,
+      "loss": 1.8738,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4313247203826904,
+      "learning_rate": 1.8230994152046787e-05,
+      "loss": 2.1458,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8259377479553223,
+      "learning_rate": 1.8230080409356728e-05,
+      "loss": 1.7878,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3479163646698,
+      "learning_rate": 1.8229166666666668e-05,
+      "loss": 2.0132,
+      "step": 1938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6269724369049072,
+      "learning_rate": 1.822825292397661e-05,
+      "loss": 1.9593,
+      "step": 1939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2148993015289307,
+      "learning_rate": 1.8227339181286553e-05,
+      "loss": 1.7861,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3095452785491943,
+      "learning_rate": 1.8226425438596493e-05,
+      "loss": 2.0262,
+      "step": 1941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.912097215652466,
+      "learning_rate": 1.8225511695906434e-05,
+      "loss": 1.6517,
+      "step": 1942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.538542747497559,
+      "learning_rate": 1.8224597953216378e-05,
+      "loss": 2.6841,
+      "step": 1943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.167004346847534,
+      "learning_rate": 1.8223684210526318e-05,
+      "loss": 1.7986,
+      "step": 1944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2547237873077393,
+      "learning_rate": 1.822277046783626e-05,
+      "loss": 1.948,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.228688955307007,
+      "learning_rate": 1.82218567251462e-05,
+      "loss": 1.8298,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8060877323150635,
+      "learning_rate": 1.8220942982456143e-05,
+      "loss": 1.7195,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1928298473358154,
+      "learning_rate": 1.8220029239766084e-05,
+      "loss": 1.8693,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5389294624328613,
+      "learning_rate": 1.8219115497076024e-05,
+      "loss": 2.0782,
+      "step": 1949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.009763717651367,
+      "learning_rate": 1.8218201754385968e-05,
+      "loss": 2.3659,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0168306827545166,
+      "learning_rate": 1.821728801169591e-05,
+      "loss": 1.8227,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8919334411621094,
+      "learning_rate": 1.821637426900585e-05,
+      "loss": 1.8868,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.101327896118164,
+      "learning_rate": 1.821546052631579e-05,
+      "loss": 1.7988,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.979253053665161,
+      "learning_rate": 1.8214546783625734e-05,
+      "loss": 1.7414,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2464518547058105,
+      "learning_rate": 1.8213633040935674e-05,
+      "loss": 1.7935,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.688585042953491,
+      "learning_rate": 1.8212719298245615e-05,
+      "loss": 2.0067,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9477105140686035,
+      "learning_rate": 1.821180555555556e-05,
+      "loss": 1.5999,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.227452516555786,
+      "learning_rate": 1.82108918128655e-05,
+      "loss": 1.8236,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2213656902313232,
+      "learning_rate": 1.820997807017544e-05,
+      "loss": 2.0216,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.973494052886963,
+      "learning_rate": 1.820906432748538e-05,
+      "loss": 2.3778,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7013590335845947,
+      "learning_rate": 1.8208150584795324e-05,
+      "loss": 1.4642,
+      "step": 1961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2238264083862305,
+      "learning_rate": 1.8207236842105265e-05,
+      "loss": 1.9788,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.447213649749756,
+      "learning_rate": 1.8206323099415205e-05,
+      "loss": 1.6904,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5143280029296875,
+      "learning_rate": 1.820540935672515e-05,
+      "loss": 1.9461,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7186965942382812,
+      "learning_rate": 1.820449561403509e-05,
+      "loss": 2.346,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.635211944580078,
+      "learning_rate": 1.820358187134503e-05,
+      "loss": 1.7913,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.660200595855713,
+      "learning_rate": 1.820266812865497e-05,
+      "loss": 2.3584,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9546704292297363,
+      "learning_rate": 1.8201754385964914e-05,
+      "loss": 1.794,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9620308876037598,
+      "learning_rate": 1.8200840643274855e-05,
+      "loss": 1.7514,
+      "step": 1969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4334609508514404,
+      "learning_rate": 1.8199926900584796e-05,
+      "loss": 1.7603,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.004573106765747,
+      "learning_rate": 1.819901315789474e-05,
+      "loss": 1.5973,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.954944133758545,
+      "learning_rate": 1.819809941520468e-05,
+      "loss": 1.4305,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0216832160949707,
+      "learning_rate": 1.819718567251462e-05,
+      "loss": 1.6496,
+      "step": 1973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5562212467193604,
+      "learning_rate": 1.819627192982456e-05,
+      "loss": 1.5369,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.0657219886779785,
+      "learning_rate": 1.8195358187134505e-05,
+      "loss": 2.0626,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.685276508331299,
+      "learning_rate": 1.8194444444444445e-05,
+      "loss": 1.9119,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7344675064086914,
+      "learning_rate": 1.8193530701754386e-05,
+      "loss": 2.0321,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7810914516448975,
+      "learning_rate": 1.819261695906433e-05,
+      "loss": 2.1696,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.544874906539917,
+      "learning_rate": 1.819170321637427e-05,
+      "loss": 1.8681,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5148165225982666,
+      "learning_rate": 1.819078947368421e-05,
+      "loss": 2.0694,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9129011631011963,
+      "learning_rate": 1.818987573099415e-05,
+      "loss": 2.0556,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.328404188156128,
+      "learning_rate": 1.8188961988304095e-05,
+      "loss": 1.7535,
+      "step": 1982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.307253122329712,
+      "learning_rate": 1.8188048245614036e-05,
+      "loss": 1.7591,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5081539154052734,
+      "learning_rate": 1.8187134502923976e-05,
+      "loss": 1.9676,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.192087173461914,
+      "learning_rate": 1.818622076023392e-05,
+      "loss": 1.8265,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.433798789978027,
+      "learning_rate": 1.818530701754386e-05,
+      "loss": 2.3615,
+      "step": 1986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8296542167663574,
+      "learning_rate": 1.81843932748538e-05,
+      "loss": 1.7061,
+      "step": 1987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2883100509643555,
+      "learning_rate": 1.8183479532163742e-05,
+      "loss": 1.8354,
+      "step": 1988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.243037223815918,
+      "learning_rate": 1.8182565789473686e-05,
+      "loss": 1.8553,
+      "step": 1989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2032816410064697,
+      "learning_rate": 1.8181652046783626e-05,
+      "loss": 2.0789,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.468019723892212,
+      "learning_rate": 1.8180738304093567e-05,
+      "loss": 1.886,
+      "step": 1991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.889874219894409,
+      "learning_rate": 1.817982456140351e-05,
+      "loss": 1.874,
+      "step": 1992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.414418935775757,
+      "learning_rate": 1.817891081871345e-05,
+      "loss": 1.8184,
+      "step": 1993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.437539577484131,
+      "learning_rate": 1.8177997076023392e-05,
+      "loss": 1.9804,
+      "step": 1994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.911975145339966,
+      "learning_rate": 1.8177083333333332e-05,
+      "loss": 2.0128,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8718955516815186,
+      "learning_rate": 1.8176169590643276e-05,
+      "loss": 1.5476,
+      "step": 1996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4116361141204834,
+      "learning_rate": 1.8175255847953217e-05,
+      "loss": 1.7442,
+      "step": 1997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2369606494903564,
+      "learning_rate": 1.8174342105263157e-05,
+      "loss": 1.921,
+      "step": 1998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2603020668029785,
+      "learning_rate": 1.81734283625731e-05,
+      "loss": 2.2006,
+      "step": 1999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5946478843688965,
+      "learning_rate": 1.8172514619883042e-05,
+      "loss": 1.957,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5848019123077393,
+      "learning_rate": 1.8171600877192982e-05,
+      "loss": 2.2122,
+      "step": 2001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.64174747467041,
+      "learning_rate": 1.8170687134502923e-05,
+      "loss": 2.0747,
+      "step": 2002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.937408924102783,
+      "learning_rate": 1.8169773391812867e-05,
+      "loss": 2.1756,
+      "step": 2003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8716814517974854,
+      "learning_rate": 1.8168859649122807e-05,
+      "loss": 1.4715,
+      "step": 2004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5827131271362305,
+      "learning_rate": 1.8167945906432748e-05,
+      "loss": 1.9633,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.589057683944702,
+      "learning_rate": 1.8167032163742692e-05,
+      "loss": 1.758,
+      "step": 2006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.128985404968262,
+      "learning_rate": 1.8166118421052632e-05,
+      "loss": 2.2208,
+      "step": 2007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4942383766174316,
+      "learning_rate": 1.8165204678362573e-05,
+      "loss": 1.525,
+      "step": 2008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.07134747505188,
+      "learning_rate": 1.8164290935672513e-05,
+      "loss": 1.6079,
+      "step": 2009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3856606483459473,
+      "learning_rate": 1.8163377192982457e-05,
+      "loss": 1.9344,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.539785146713257,
+      "learning_rate": 1.81624634502924e-05,
+      "loss": 1.9942,
+      "step": 2011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1993935108184814,
+      "learning_rate": 1.8161549707602342e-05,
+      "loss": 1.9129,
+      "step": 2012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.1024394035339355,
+      "learning_rate": 1.8160635964912282e-05,
+      "loss": 2.2647,
+      "step": 2013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2860569953918457,
+      "learning_rate": 1.8159722222222226e-05,
+      "loss": 1.767,
+      "step": 2014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2871129512786865,
+      "learning_rate": 1.8158808479532167e-05,
+      "loss": 1.614,
+      "step": 2015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.961859703063965,
+      "learning_rate": 1.8157894736842107e-05,
+      "loss": 1.4119,
+      "step": 2016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9614312648773193,
+      "learning_rate": 1.8156980994152048e-05,
+      "loss": 1.6746,
+      "step": 2017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3954243659973145,
+      "learning_rate": 1.815606725146199e-05,
+      "loss": 2.2679,
+      "step": 2018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3101847171783447,
+      "learning_rate": 1.8155153508771932e-05,
+      "loss": 1.9965,
+      "step": 2019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9659197330474854,
+      "learning_rate": 1.8154239766081873e-05,
+      "loss": 1.6941,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.864116907119751,
+      "learning_rate": 1.8153326023391817e-05,
+      "loss": 1.3952,
+      "step": 2021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8644649982452393,
+      "learning_rate": 1.8152412280701757e-05,
+      "loss": 2.405,
+      "step": 2022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.678011178970337,
+      "learning_rate": 1.8151498538011698e-05,
+      "loss": 1.8983,
+      "step": 2023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.07637357711792,
+      "learning_rate": 1.8150584795321638e-05,
+      "loss": 1.7565,
+      "step": 2024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.010897159576416,
+      "learning_rate": 1.8149671052631582e-05,
+      "loss": 1.9921,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.177135944366455,
+      "learning_rate": 1.8148757309941523e-05,
+      "loss": 2.1682,
+      "step": 2026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.093393325805664,
+      "learning_rate": 1.8147843567251463e-05,
+      "loss": 1.9859,
+      "step": 2027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3743107318878174,
+      "learning_rate": 1.8146929824561407e-05,
+      "loss": 1.7088,
+      "step": 2028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.160353899002075,
+      "learning_rate": 1.8146016081871348e-05,
+      "loss": 1.9649,
+      "step": 2029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.278964042663574,
+      "learning_rate": 1.8145102339181288e-05,
+      "loss": 1.9539,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.804508686065674,
+      "learning_rate": 1.814418859649123e-05,
+      "loss": 1.9827,
+      "step": 2031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.298215627670288,
+      "learning_rate": 1.8143274853801173e-05,
+      "loss": 1.6548,
+      "step": 2032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.013182163238525,
+      "learning_rate": 1.8142361111111113e-05,
+      "loss": 1.695,
+      "step": 2033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.09699821472168,
+      "learning_rate": 1.8141447368421054e-05,
+      "loss": 1.8616,
+      "step": 2034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1409244537353516,
+      "learning_rate": 1.8140533625730998e-05,
+      "loss": 1.7778,
+      "step": 2035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.269198417663574,
+      "learning_rate": 1.8139619883040938e-05,
+      "loss": 1.862,
+      "step": 2036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.265542030334473,
+      "learning_rate": 1.813870614035088e-05,
+      "loss": 2.4922,
+      "step": 2037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0573606491088867,
+      "learning_rate": 1.813779239766082e-05,
+      "loss": 1.5883,
+      "step": 2038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.757375955581665,
+      "learning_rate": 1.8136878654970763e-05,
+      "loss": 1.6954,
+      "step": 2039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6055080890655518,
+      "learning_rate": 1.8135964912280704e-05,
+      "loss": 2.1369,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4633493423461914,
+      "learning_rate": 1.8135051169590644e-05,
+      "loss": 1.6633,
+      "step": 2041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.438472270965576,
+      "learning_rate": 1.8134137426900588e-05,
+      "loss": 1.8929,
+      "step": 2042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9308791160583496,
+      "learning_rate": 1.813322368421053e-05,
+      "loss": 1.9991,
+      "step": 2043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.08495569229126,
+      "learning_rate": 1.813230994152047e-05,
+      "loss": 2.3253,
+      "step": 2044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.572796106338501,
+      "learning_rate": 1.813139619883041e-05,
+      "loss": 1.9125,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2925798892974854,
+      "learning_rate": 1.8130482456140354e-05,
+      "loss": 2.0534,
+      "step": 2046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.428433656692505,
+      "learning_rate": 1.8129568713450294e-05,
+      "loss": 1.7581,
+      "step": 2047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6667637825012207,
+      "learning_rate": 1.8128654970760235e-05,
+      "loss": 2.1908,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.354578971862793,
+      "learning_rate": 1.812774122807018e-05,
+      "loss": 1.8119,
+      "step": 2049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4564716815948486,
+      "learning_rate": 1.812682748538012e-05,
+      "loss": 1.6798,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.357335090637207,
+      "learning_rate": 1.812591374269006e-05,
+      "loss": 1.9825,
+      "step": 2051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1471498012542725,
+      "learning_rate": 1.8125e-05,
+      "loss": 1.8402,
+      "step": 2052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2423760890960693,
+      "learning_rate": 1.8124086257309944e-05,
+      "loss": 1.8053,
+      "step": 2053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.57812762260437,
+      "learning_rate": 1.8123172514619885e-05,
+      "loss": 1.9401,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1274876594543457,
+      "learning_rate": 1.8122258771929825e-05,
+      "loss": 1.7371,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2948968410491943,
+      "learning_rate": 1.812134502923977e-05,
+      "loss": 1.8026,
+      "step": 2056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.351862907409668,
+      "learning_rate": 1.812043128654971e-05,
+      "loss": 2.0818,
+      "step": 2057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.475735902786255,
+      "learning_rate": 1.811951754385965e-05,
+      "loss": 1.8783,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.391279697418213,
+      "learning_rate": 1.811860380116959e-05,
+      "loss": 2.3664,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1230597496032715,
+      "learning_rate": 1.8117690058479534e-05,
+      "loss": 1.6864,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6966395378112793,
+      "learning_rate": 1.8116776315789475e-05,
+      "loss": 1.7744,
+      "step": 2061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.000626802444458,
+      "learning_rate": 1.8115862573099416e-05,
+      "loss": 1.6543,
+      "step": 2062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.062527656555176,
+      "learning_rate": 1.811494883040936e-05,
+      "loss": 2.0453,
+      "step": 2063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.162020444869995,
+      "learning_rate": 1.81140350877193e-05,
+      "loss": 2.1291,
+      "step": 2064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.499754905700684,
+      "learning_rate": 1.811312134502924e-05,
+      "loss": 2.3664,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.809937000274658,
+      "learning_rate": 1.811220760233918e-05,
+      "loss": 1.729,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2632813453674316,
+      "learning_rate": 1.8111293859649125e-05,
+      "loss": 1.9014,
+      "step": 2067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.201141357421875,
+      "learning_rate": 1.8110380116959065e-05,
+      "loss": 2.0315,
+      "step": 2068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.521798849105835,
+      "learning_rate": 1.8109466374269006e-05,
+      "loss": 2.0916,
+      "step": 2069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.139560699462891,
+      "learning_rate": 1.810855263157895e-05,
+      "loss": 2.174,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7207539081573486,
+      "learning_rate": 1.810763888888889e-05,
+      "loss": 2.3953,
+      "step": 2071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9657275676727295,
+      "learning_rate": 1.810672514619883e-05,
+      "loss": 2.1659,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.167140007019043,
+      "learning_rate": 1.810581140350877e-05,
+      "loss": 1.6842,
+      "step": 2073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.432276964187622,
+      "learning_rate": 1.8104897660818715e-05,
+      "loss": 1.9357,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.389986276626587,
+      "learning_rate": 1.8103983918128656e-05,
+      "loss": 1.6917,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5281982421875,
+      "learning_rate": 1.8103070175438596e-05,
+      "loss": 1.7766,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3048224449157715,
+      "learning_rate": 1.810215643274854e-05,
+      "loss": 1.7454,
+      "step": 2077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0129153728485107,
+      "learning_rate": 1.810124269005848e-05,
+      "loss": 1.746,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.588357925415039,
+      "learning_rate": 1.810032894736842e-05,
+      "loss": 1.7889,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7597131729125977,
+      "learning_rate": 1.8099415204678362e-05,
+      "loss": 2.0736,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.231980800628662,
+      "learning_rate": 1.8098501461988306e-05,
+      "loss": 1.6867,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.167822360992432,
+      "learning_rate": 1.8097587719298246e-05,
+      "loss": 2.2702,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5865747928619385,
+      "learning_rate": 1.8096673976608187e-05,
+      "loss": 1.6577,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.984569787979126,
+      "learning_rate": 1.809576023391813e-05,
+      "loss": 1.7949,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0898995399475098,
+      "learning_rate": 1.809484649122807e-05,
+      "loss": 1.7937,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4503262042999268,
+      "learning_rate": 1.8093932748538012e-05,
+      "loss": 1.9849,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.247025489807129,
+      "learning_rate": 1.8093019005847952e-05,
+      "loss": 2.0838,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.574066162109375,
+      "learning_rate": 1.8092105263157896e-05,
+      "loss": 2.3224,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8395378589630127,
+      "learning_rate": 1.8091191520467837e-05,
+      "loss": 2.1772,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2717485427856445,
+      "learning_rate": 1.8090277777777777e-05,
+      "loss": 2.0672,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.446027994155884,
+      "learning_rate": 1.808936403508772e-05,
+      "loss": 1.8261,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.318159818649292,
+      "learning_rate": 1.8088450292397662e-05,
+      "loss": 2.1656,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.25111722946167,
+      "learning_rate": 1.8087536549707602e-05,
+      "loss": 1.8804,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.569958448410034,
+      "learning_rate": 1.8086622807017543e-05,
+      "loss": 1.7627,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.626230239868164,
+      "learning_rate": 1.8085709064327487e-05,
+      "loss": 2.2485,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.805541753768921,
+      "learning_rate": 1.808479532163743e-05,
+      "loss": 1.698,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1647183895111084,
+      "learning_rate": 1.808388157894737e-05,
+      "loss": 1.8226,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7645413875579834,
+      "learning_rate": 1.8082967836257312e-05,
+      "loss": 2.0123,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4411580562591553,
+      "learning_rate": 1.8082054093567252e-05,
+      "loss": 2.0114,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.189635753631592,
+      "learning_rate": 1.8081140350877196e-05,
+      "loss": 1.7267,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.357818841934204,
+      "learning_rate": 1.8080226608187137e-05,
+      "loss": 2.0079,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.723302364349365,
+      "learning_rate": 1.8079312865497077e-05,
+      "loss": 2.2996,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1690189838409424,
+      "learning_rate": 1.807839912280702e-05,
+      "loss": 1.752,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4056668281555176,
+      "learning_rate": 1.8077485380116962e-05,
+      "loss": 2.2923,
+      "step": 2104
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.477675914764404,
+      "learning_rate": 1.8076571637426902e-05,
+      "loss": 2.0825,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0901877880096436,
+      "learning_rate": 1.8075657894736843e-05,
+      "loss": 1.7338,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5680758953094482,
+      "learning_rate": 1.8074744152046787e-05,
+      "loss": 1.8566,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6349403858184814,
+      "learning_rate": 1.8073830409356727e-05,
+      "loss": 1.2511,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7140657901763916,
+      "learning_rate": 1.8072916666666668e-05,
+      "loss": 1.5887,
+      "step": 2109
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1029257774353027,
+      "learning_rate": 1.807200292397661e-05,
+      "loss": 1.802,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.957409143447876,
+      "learning_rate": 1.8071089181286552e-05,
+      "loss": 1.6215,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0663161277770996,
+      "learning_rate": 1.8070175438596493e-05,
+      "loss": 2.0985,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.425666332244873,
+      "learning_rate": 1.8069261695906433e-05,
+      "loss": 1.8971,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2021384239196777,
+      "learning_rate": 1.8068347953216377e-05,
+      "loss": 1.5304,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5556704998016357,
+      "learning_rate": 1.8067434210526318e-05,
+      "loss": 1.5924,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6896555423736572,
+      "learning_rate": 1.8066520467836258e-05,
+      "loss": 1.5767,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3203611373901367,
+      "learning_rate": 1.8065606725146202e-05,
+      "loss": 1.7807,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1829023361206055,
+      "learning_rate": 1.8064692982456143e-05,
+      "loss": 2.0559,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7126009464263916,
+      "learning_rate": 1.8063779239766083e-05,
+      "loss": 2.06,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1580255031585693,
+      "learning_rate": 1.8062865497076024e-05,
+      "loss": 1.7791,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.750633478164673,
+      "learning_rate": 1.8061951754385968e-05,
+      "loss": 2.4869,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.034003973007202,
+      "learning_rate": 1.8061038011695908e-05,
+      "loss": 1.8258,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6053695678710938,
+      "learning_rate": 1.806012426900585e-05,
+      "loss": 1.7815,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9505457878112793,
+      "learning_rate": 1.8059210526315793e-05,
+      "loss": 1.7131,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.137901782989502,
+      "learning_rate": 1.8058296783625733e-05,
+      "loss": 1.7739,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5192418098449707,
+      "learning_rate": 1.8057383040935674e-05,
+      "loss": 1.7863,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.574002265930176,
+      "learning_rate": 1.8056469298245614e-05,
+      "loss": 1.9256,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.498568534851074,
+      "learning_rate": 1.8055555555555558e-05,
+      "loss": 1.8461,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.123870372772217,
+      "learning_rate": 1.80546418128655e-05,
+      "loss": 2.6597,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8153724670410156,
+      "learning_rate": 1.805372807017544e-05,
+      "loss": 2.4003,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.427057981491089,
+      "learning_rate": 1.8052814327485383e-05,
+      "loss": 1.6941,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.267096519470215,
+      "learning_rate": 1.8051900584795324e-05,
+      "loss": 1.821,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0790984630584717,
+      "learning_rate": 1.8050986842105264e-05,
+      "loss": 1.5791,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9507229328155518,
+      "learning_rate": 1.8050073099415205e-05,
+      "loss": 1.9549,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8236894607543945,
+      "learning_rate": 1.804915935672515e-05,
+      "loss": 2.1829,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.887631893157959,
+      "learning_rate": 1.804824561403509e-05,
+      "loss": 1.3312,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0846941471099854,
+      "learning_rate": 1.804733187134503e-05,
+      "loss": 1.6662,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.63918399810791,
+      "learning_rate": 1.8046418128654974e-05,
+      "loss": 2.2087,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4134626388549805,
+      "learning_rate": 1.8045504385964914e-05,
+      "loss": 1.8701,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6184191703796387,
+      "learning_rate": 1.8044590643274855e-05,
+      "loss": 1.7564,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7917537689208984,
+      "learning_rate": 1.8043676900584795e-05,
+      "loss": 1.8889,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2653183937072754,
+      "learning_rate": 1.804276315789474e-05,
+      "loss": 1.5296,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.105008125305176,
+      "learning_rate": 1.804184941520468e-05,
+      "loss": 1.7745,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6585659980773926,
+      "learning_rate": 1.804093567251462e-05,
+      "loss": 1.565,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.041123390197754,
+      "learning_rate": 1.8040021929824564e-05,
+      "loss": 1.7299,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9343531131744385,
+      "learning_rate": 1.8039108187134505e-05,
+      "loss": 1.947,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.990222930908203,
+      "learning_rate": 1.8038194444444445e-05,
+      "loss": 1.5958,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.701829433441162,
+      "learning_rate": 1.8037280701754386e-05,
+      "loss": 1.8407,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.59606671333313,
+      "learning_rate": 1.803636695906433e-05,
+      "loss": 2.2901,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5222301483154297,
+      "learning_rate": 1.803545321637427e-05,
+      "loss": 2.1532,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.456449031829834,
+      "learning_rate": 1.803453947368421e-05,
+      "loss": 1.9254,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.652958393096924,
+      "learning_rate": 1.8033625730994154e-05,
+      "loss": 2.049,
+      "step": 2152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.584595203399658,
+      "learning_rate": 1.8032711988304095e-05,
+      "loss": 1.6593,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7136735916137695,
+      "learning_rate": 1.8031798245614036e-05,
+      "loss": 1.9861,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.447970151901245,
+      "learning_rate": 1.8030884502923976e-05,
+      "loss": 2.1971,
+      "step": 2155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.628357172012329,
+      "learning_rate": 1.802997076023392e-05,
+      "loss": 2.2139,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4020373821258545,
+      "learning_rate": 1.802905701754386e-05,
+      "loss": 1.661,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.477456569671631,
+      "learning_rate": 1.80281432748538e-05,
+      "loss": 1.6767,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3410632610321045,
+      "learning_rate": 1.8027229532163745e-05,
+      "loss": 2.0181,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3598990440368652,
+      "learning_rate": 1.8026315789473685e-05,
+      "loss": 1.9353,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7070059776306152,
+      "learning_rate": 1.8025402046783626e-05,
+      "loss": 2.0097,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1975717544555664,
+      "learning_rate": 1.8024488304093567e-05,
+      "loss": 1.9351,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.518405914306641,
+      "learning_rate": 1.802357456140351e-05,
+      "loss": 2.2216,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8574843406677246,
+      "learning_rate": 1.802266081871345e-05,
+      "loss": 2.1459,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.762469530105591,
+      "learning_rate": 1.802174707602339e-05,
+      "loss": 1.7286,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6747632026672363,
+      "learning_rate": 1.8020833333333335e-05,
+      "loss": 1.7311,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.912165880203247,
+      "learning_rate": 1.8019919590643276e-05,
+      "loss": 1.9339,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3870303630828857,
+      "learning_rate": 1.8019005847953216e-05,
+      "loss": 2.0383,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5635287761688232,
+      "learning_rate": 1.8018092105263157e-05,
+      "loss": 1.6455,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2788496017456055,
+      "learning_rate": 1.80171783625731e-05,
+      "loss": 2.0638,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4206464290618896,
+      "learning_rate": 1.801626461988304e-05,
+      "loss": 1.8544,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3869762420654297,
+      "learning_rate": 1.8015350877192982e-05,
+      "loss": 1.9984,
+      "step": 2172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2768359184265137,
+      "learning_rate": 1.8014437134502926e-05,
+      "loss": 1.9857,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.627713680267334,
+      "learning_rate": 1.8013523391812866e-05,
+      "loss": 2.1967,
+      "step": 2174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.930541515350342,
+      "learning_rate": 1.8012609649122807e-05,
+      "loss": 1.4914,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.21821928024292,
+      "learning_rate": 1.8011695906432747e-05,
+      "loss": 1.994,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.107012987136841,
+      "learning_rate": 1.801078216374269e-05,
+      "loss": 1.7592,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.557236909866333,
+      "learning_rate": 1.8009868421052632e-05,
+      "loss": 1.8525,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.912508964538574,
+      "learning_rate": 1.8008954678362572e-05,
+      "loss": 1.501,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.063901424407959,
+      "learning_rate": 1.8008040935672516e-05,
+      "loss": 2.0748,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6648545265197754,
+      "learning_rate": 1.800712719298246e-05,
+      "loss": 1.9663,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.165778160095215,
+      "learning_rate": 1.80062134502924e-05,
+      "loss": 1.9643,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3138515949249268,
+      "learning_rate": 1.800529970760234e-05,
+      "loss": 1.9569,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2071304321289062,
+      "learning_rate": 1.8004385964912282e-05,
+      "loss": 1.6806,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8169562816619873,
+      "learning_rate": 1.8003472222222226e-05,
+      "loss": 1.7624,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.323399066925049,
+      "learning_rate": 1.8002558479532166e-05,
+      "loss": 2.2843,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6078145503997803,
+      "learning_rate": 1.8001644736842107e-05,
+      "loss": 1.6807,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6310999393463135,
+      "learning_rate": 1.800073099415205e-05,
+      "loss": 2.0202,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.400463819503784,
+      "learning_rate": 1.799981725146199e-05,
+      "loss": 1.6951,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1,
+      "eval_train_bleu": 0.20040223261702134,
+      "eval_train_loss": 2.2013635635375977,
+      "eval_train_runtime": 36.0783,
+      "eval_train_samples_per_second": 27.745,
+      "eval_train_steps_per_second": 0.887,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.896940231323242,
+      "learning_rate": 1.7998903508771932e-05,
+      "loss": 1.4576,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.024843692779541,
+      "learning_rate": 1.7997989766081872e-05,
+      "loss": 1.591,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5784919261932373,
+      "learning_rate": 1.7997076023391816e-05,
+      "loss": 1.8327,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9958157539367676,
+      "learning_rate": 1.7996162280701757e-05,
+      "loss": 1.7636,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.454035758972168,
+      "learning_rate": 1.7995248538011697e-05,
+      "loss": 1.8524,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3105313777923584,
+      "learning_rate": 1.799433479532164e-05,
+      "loss": 1.8778,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.529165506362915,
+      "learning_rate": 1.7993421052631582e-05,
+      "loss": 1.6712,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3776278495788574,
+      "learning_rate": 1.7992507309941522e-05,
+      "loss": 1.7975,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8141698837280273,
+      "learning_rate": 1.7991593567251463e-05,
+      "loss": 2.0343,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.645447254180908,
+      "learning_rate": 1.7990679824561407e-05,
+      "loss": 2.0349,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5651843547821045,
+      "learning_rate": 1.7989766081871347e-05,
+      "loss": 1.6647,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.936692237854004,
+      "learning_rate": 1.7988852339181288e-05,
+      "loss": 1.6987,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0643887519836426,
+      "learning_rate": 1.798793859649123e-05,
+      "loss": 1.7055,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1042447090148926,
+      "learning_rate": 1.7987024853801172e-05,
+      "loss": 1.6699,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6604104042053223,
+      "learning_rate": 1.7986111111111113e-05,
+      "loss": 2.0666,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5316827297210693,
+      "learning_rate": 1.7985197368421053e-05,
+      "loss": 2.2844,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2793097496032715,
+      "learning_rate": 1.7984283625730997e-05,
+      "loss": 1.8113,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.002910614013672,
+      "learning_rate": 1.7983369883040938e-05,
+      "loss": 2.1977,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.566758632659912,
+      "learning_rate": 1.7982456140350878e-05,
+      "loss": 1.6261,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2675113677978516,
+      "learning_rate": 1.7981542397660822e-05,
+      "loss": 1.9234,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.129540205001831,
+      "learning_rate": 1.7980628654970763e-05,
+      "loss": 1.7515,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0663070678710938,
+      "learning_rate": 1.7979714912280703e-05,
+      "loss": 1.6505,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.812297821044922,
+      "learning_rate": 1.7978801169590644e-05,
+      "loss": 1.603,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.388770580291748,
+      "learning_rate": 1.7977887426900588e-05,
+      "loss": 1.8502,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4318509101867676,
+      "learning_rate": 1.7976973684210528e-05,
+      "loss": 1.7981,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5754497051239014,
+      "learning_rate": 1.797605994152047e-05,
+      "loss": 2.1517,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8046722412109375,
+      "learning_rate": 1.7975146198830413e-05,
+      "loss": 1.7016,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.190574884414673,
+      "learning_rate": 1.7974232456140353e-05,
+      "loss": 1.8903,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1773598194122314,
+      "learning_rate": 1.7973318713450294e-05,
+      "loss": 2.006,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.000472545623779,
+      "learning_rate": 1.7972404970760234e-05,
+      "loss": 2.1137,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7436635494232178,
+      "learning_rate": 1.7971491228070178e-05,
+      "loss": 1.9359,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.67657732963562,
+      "learning_rate": 1.797057748538012e-05,
+      "loss": 2.069,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.730975389480591,
+      "learning_rate": 1.796966374269006e-05,
+      "loss": 2.0625,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4449877738952637,
+      "learning_rate": 1.7968750000000003e-05,
+      "loss": 2.1737,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.301199197769165,
+      "learning_rate": 1.7967836257309944e-05,
+      "loss": 1.7961,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.467402935028076,
+      "learning_rate": 1.7966922514619884e-05,
+      "loss": 2.2226,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.880094051361084,
+      "learning_rate": 1.7966008771929825e-05,
+      "loss": 1.8242,
+      "step": 2226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.515824556350708,
+      "learning_rate": 1.796509502923977e-05,
+      "loss": 2.1103,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.96211838722229,
+      "learning_rate": 1.796418128654971e-05,
+      "loss": 1.6154,
+      "step": 2228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.895592212677002,
+      "learning_rate": 1.796326754385965e-05,
+      "loss": 1.6827,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7575156688690186,
+      "learning_rate": 1.7962353801169594e-05,
+      "loss": 1.6243,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.797656774520874,
+      "learning_rate": 1.7961440058479534e-05,
+      "loss": 1.6277,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.291386365890503,
+      "learning_rate": 1.7960526315789475e-05,
+      "loss": 1.5634,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0939242839813232,
+      "learning_rate": 1.7959612573099415e-05,
+      "loss": 1.858,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.219286918640137,
+      "learning_rate": 1.795869883040936e-05,
+      "loss": 2.6458,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0894763469696045,
+      "learning_rate": 1.79577850877193e-05,
+      "loss": 1.7271,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8695735931396484,
+      "learning_rate": 1.795687134502924e-05,
+      "loss": 1.678,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9431324005126953,
+      "learning_rate": 1.7955957602339184e-05,
+      "loss": 1.6719,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.225442409515381,
+      "learning_rate": 1.7955043859649125e-05,
+      "loss": 1.7283,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5707480907440186,
+      "learning_rate": 1.7954130116959065e-05,
+      "loss": 1.7978,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.125620126724243,
+      "learning_rate": 1.7953216374269006e-05,
+      "loss": 1.6201,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.701726198196411,
+      "learning_rate": 1.795230263157895e-05,
+      "loss": 1.3951,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.952181100845337,
+      "learning_rate": 1.795138888888889e-05,
+      "loss": 2.2757,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.847029447555542,
+      "learning_rate": 1.795047514619883e-05,
+      "loss": 1.8023,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.068589448928833,
+      "learning_rate": 1.7949561403508774e-05,
+      "loss": 1.8867,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4590446949005127,
+      "learning_rate": 1.7948647660818715e-05,
+      "loss": 2.2695,
+      "step": 2245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7529642581939697,
+      "learning_rate": 1.7947733918128656e-05,
+      "loss": 2.3109,
+      "step": 2246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.889817953109741,
+      "learning_rate": 1.7946820175438596e-05,
+      "loss": 1.554,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.110274076461792,
+      "learning_rate": 1.794590643274854e-05,
+      "loss": 1.5008,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.907294988632202,
+      "learning_rate": 1.794499269005848e-05,
+      "loss": 2.2863,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6280674934387207,
+      "learning_rate": 1.794407894736842e-05,
+      "loss": 1.8228,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.869234323501587,
+      "learning_rate": 1.7943165204678365e-05,
+      "loss": 2.2489,
+      "step": 2251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6233625411987305,
+      "learning_rate": 1.7942251461988305e-05,
+      "loss": 2.1511,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5029568672180176,
+      "learning_rate": 1.7941337719298246e-05,
+      "loss": 2.1085,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.341686248779297,
+      "learning_rate": 1.7940423976608187e-05,
+      "loss": 1.9528,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2351226806640625,
+      "learning_rate": 1.793951023391813e-05,
+      "loss": 1.6294,
+      "step": 2255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.426051616668701,
+      "learning_rate": 1.793859649122807e-05,
+      "loss": 1.8012,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3357810974121094,
+      "learning_rate": 1.793768274853801e-05,
+      "loss": 1.4728,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7167327404022217,
+      "learning_rate": 1.7936769005847955e-05,
+      "loss": 1.5806,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7683660984039307,
+      "learning_rate": 1.7935855263157896e-05,
+      "loss": 2.1872,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0851430892944336,
+      "learning_rate": 1.7934941520467836e-05,
+      "loss": 1.6793,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.339564561843872,
+      "learning_rate": 1.7934027777777777e-05,
+      "loss": 1.9167,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6976985931396484,
+      "learning_rate": 1.793311403508772e-05,
+      "loss": 2.3947,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.109163999557495,
+      "learning_rate": 1.793220029239766e-05,
+      "loss": 1.9161,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.454575538635254,
+      "learning_rate": 1.7931286549707602e-05,
+      "loss": 2.0202,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1243107318878174,
+      "learning_rate": 1.7930372807017546e-05,
+      "loss": 2.0297,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6229147911071777,
+      "learning_rate": 1.7929459064327486e-05,
+      "loss": 1.9731,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.182385206222534,
+      "learning_rate": 1.792854532163743e-05,
+      "loss": 1.4756,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.71923303604126,
+      "learning_rate": 1.792763157894737e-05,
+      "loss": 3.403,
+      "step": 2268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.875469923019409,
+      "learning_rate": 1.792671783625731e-05,
+      "loss": 1.5347,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0966522693634033,
+      "learning_rate": 1.7925804093567255e-05,
+      "loss": 1.6939,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.510272979736328,
+      "learning_rate": 1.7924890350877196e-05,
+      "loss": 2.5928,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.559112787246704,
+      "learning_rate": 1.7923976608187136e-05,
+      "loss": 2.2333,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.263666868209839,
+      "learning_rate": 1.7923062865497077e-05,
+      "loss": 1.6635,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3673319816589355,
+      "learning_rate": 1.792214912280702e-05,
+      "loss": 2.1475,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.331494092941284,
+      "learning_rate": 1.792123538011696e-05,
+      "loss": 1.6882,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.273167133331299,
+      "learning_rate": 1.7920321637426902e-05,
+      "loss": 2.1065,
+      "step": 2276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3105154037475586,
+      "learning_rate": 1.7919407894736846e-05,
+      "loss": 1.6395,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.103750228881836,
+      "learning_rate": 1.7918494152046786e-05,
+      "loss": 1.719,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.659158945083618,
+      "learning_rate": 1.7917580409356727e-05,
+      "loss": 2.0391,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8889119625091553,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 1.6678,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.537752866744995,
+      "learning_rate": 1.791575292397661e-05,
+      "loss": 2.3094,
+      "step": 2281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.620361804962158,
+      "learning_rate": 1.7914839181286552e-05,
+      "loss": 1.7721,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.136227607727051,
+      "learning_rate": 1.7913925438596492e-05,
+      "loss": 1.8639,
+      "step": 2283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2283413410186768,
+      "learning_rate": 1.7913011695906436e-05,
+      "loss": 1.5196,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5882749557495117,
+      "learning_rate": 1.7912097953216377e-05,
+      "loss": 1.8727,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.070293664932251,
+      "learning_rate": 1.7911184210526317e-05,
+      "loss": 1.8913,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.552963972091675,
+      "learning_rate": 1.7910270467836258e-05,
+      "loss": 2.1219,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.57529354095459,
+      "learning_rate": 1.7909356725146202e-05,
+      "loss": 2.3572,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4926300048828125,
+      "learning_rate": 1.7908442982456142e-05,
+      "loss": 1.7022,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8120806217193604,
+      "learning_rate": 1.7907529239766083e-05,
+      "loss": 1.3804,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2859914302825928,
+      "learning_rate": 1.7906615497076027e-05,
+      "loss": 1.5583,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2261505126953125,
+      "learning_rate": 1.7905701754385967e-05,
+      "loss": 1.9313,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6070122718811035,
+      "learning_rate": 1.7904788011695908e-05,
+      "loss": 1.664,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.84452748298645,
+      "learning_rate": 1.7903874269005848e-05,
+      "loss": 1.7435,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3773410320281982,
+      "learning_rate": 1.7902960526315792e-05,
+      "loss": 1.8629,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.227962017059326,
+      "learning_rate": 1.7902046783625733e-05,
+      "loss": 2.0399,
+      "step": 2296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9964663982391357,
+      "learning_rate": 1.7901133040935673e-05,
+      "loss": 1.8329,
+      "step": 2297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.860121726989746,
+      "learning_rate": 1.7900219298245617e-05,
+      "loss": 1.822,
+      "step": 2298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.525698184967041,
+      "learning_rate": 1.7899305555555558e-05,
+      "loss": 2.0671,
+      "step": 2299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.70863151550293,
+      "learning_rate": 1.7898391812865498e-05,
+      "loss": 1.852,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8824548721313477,
+      "learning_rate": 1.789747807017544e-05,
+      "loss": 1.9119,
+      "step": 2301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4362401962280273,
+      "learning_rate": 1.7896564327485383e-05,
+      "loss": 1.7324,
+      "step": 2302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.854076385498047,
+      "learning_rate": 1.7895650584795323e-05,
+      "loss": 1.7124,
+      "step": 2303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.558182716369629,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 2.0316,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3386390209198,
+      "learning_rate": 1.7893823099415208e-05,
+      "loss": 1.9051,
+      "step": 2305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8609201908111572,
+      "learning_rate": 1.7892909356725148e-05,
+      "loss": 1.7784,
+      "step": 2306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1231882572174072,
+      "learning_rate": 1.789199561403509e-05,
+      "loss": 1.544,
+      "step": 2307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2545220851898193,
+      "learning_rate": 1.789108187134503e-05,
+      "loss": 1.5001,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.3557939529418945,
+      "learning_rate": 1.7890168128654973e-05,
+      "loss": 1.8074,
+      "step": 2309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9398810863494873,
+      "learning_rate": 1.7889254385964914e-05,
+      "loss": 1.5633,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2168221473693848,
+      "learning_rate": 1.7888340643274854e-05,
+      "loss": 1.8266,
+      "step": 2311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4434423446655273,
+      "learning_rate": 1.7887426900584798e-05,
+      "loss": 1.7614,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9485630989074707,
+      "learning_rate": 1.788651315789474e-05,
+      "loss": 1.4056,
+      "step": 2313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.715372085571289,
+      "learning_rate": 1.788559941520468e-05,
+      "loss": 2.1883,
+      "step": 2314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8429224491119385,
+      "learning_rate": 1.788468567251462e-05,
+      "loss": 1.578,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.764182090759277,
+      "learning_rate": 1.7883771929824564e-05,
+      "loss": 1.8051,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5248730182647705,
+      "learning_rate": 1.7882858187134504e-05,
+      "loss": 2.016,
+      "step": 2317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.851124286651611,
+      "learning_rate": 1.7881944444444445e-05,
+      "loss": 2.3656,
+      "step": 2318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1804676055908203,
+      "learning_rate": 1.788103070175439e-05,
+      "loss": 1.5764,
+      "step": 2319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.996840715408325,
+      "learning_rate": 1.788011695906433e-05,
+      "loss": 2.1776,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.686502456665039,
+      "learning_rate": 1.787920321637427e-05,
+      "loss": 1.8564,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.17362117767334,
+      "learning_rate": 1.787828947368421e-05,
+      "loss": 2.1801,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8362650871276855,
+      "learning_rate": 1.7877375730994154e-05,
+      "loss": 1.9461,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.671250343322754,
+      "learning_rate": 1.7876461988304095e-05,
+      "loss": 1.847,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.861727714538574,
+      "learning_rate": 1.7875548245614035e-05,
+      "loss": 2.4643,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6756491661071777,
+      "learning_rate": 1.787463450292398e-05,
+      "loss": 1.9415,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.019631862640381,
+      "learning_rate": 1.787372076023392e-05,
+      "loss": 2.2532,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.876211166381836,
+      "learning_rate": 1.787280701754386e-05,
+      "loss": 1.9985,
+      "step": 2328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.426421642303467,
+      "learning_rate": 1.78718932748538e-05,
+      "loss": 1.7424,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8930227756500244,
+      "learning_rate": 1.7870979532163745e-05,
+      "loss": 1.8303,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0747647285461426,
+      "learning_rate": 1.7870065789473685e-05,
+      "loss": 1.7356,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2399239540100098,
+      "learning_rate": 1.7869152046783626e-05,
+      "loss": 2.0041,
+      "step": 2332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.628474473953247,
+      "learning_rate": 1.786823830409357e-05,
+      "loss": 2.16,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.903700590133667,
+      "learning_rate": 1.786732456140351e-05,
+      "loss": 2.0321,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.907522201538086,
+      "learning_rate": 1.786641081871345e-05,
+      "loss": 1.8336,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5416452884674072,
+      "learning_rate": 1.786549707602339e-05,
+      "loss": 1.4573,
+      "step": 2336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7252275943756104,
+      "learning_rate": 1.7864583333333335e-05,
+      "loss": 1.685,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.330787181854248,
+      "learning_rate": 1.7863669590643276e-05,
+      "loss": 2.07,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9057250022888184,
+      "learning_rate": 1.7862755847953216e-05,
+      "loss": 1.5436,
+      "step": 2339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.487884283065796,
+      "learning_rate": 1.786184210526316e-05,
+      "loss": 1.737,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2506470680236816,
+      "learning_rate": 1.78609283625731e-05,
+      "loss": 1.9888,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5076446533203125,
+      "learning_rate": 1.786001461988304e-05,
+      "loss": 1.893,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6509950160980225,
+      "learning_rate": 1.785910087719298e-05,
+      "loss": 2.1198,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.915966510772705,
+      "learning_rate": 1.7858187134502925e-05,
+      "loss": 1.7262,
+      "step": 2344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.654109001159668,
+      "learning_rate": 1.7857273391812866e-05,
+      "loss": 1.7917,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9846012592315674,
+      "learning_rate": 1.7856359649122807e-05,
+      "loss": 1.6749,
+      "step": 2346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.970515012741089,
+      "learning_rate": 1.785544590643275e-05,
+      "loss": 1.9099,
+      "step": 2347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8728551864624023,
+      "learning_rate": 1.785453216374269e-05,
+      "loss": 1.9451,
+      "step": 2348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.199223756790161,
+      "learning_rate": 1.785361842105263e-05,
+      "loss": 1.8095,
+      "step": 2349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.063656806945801,
+      "learning_rate": 1.7852704678362572e-05,
+      "loss": 2.5279,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3635268211364746,
+      "learning_rate": 1.7851790935672516e-05,
+      "loss": 1.7423,
+      "step": 2351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.423642158508301,
+      "learning_rate": 1.785087719298246e-05,
+      "loss": 1.79,
+      "step": 2352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.293857574462891,
+      "learning_rate": 1.78499634502924e-05,
+      "loss": 2.2814,
+      "step": 2353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.233362913131714,
+      "learning_rate": 1.784904970760234e-05,
+      "loss": 1.6706,
+      "step": 2354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6586198806762695,
+      "learning_rate": 1.7848135964912285e-05,
+      "loss": 1.906,
+      "step": 2355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.410034656524658,
+      "learning_rate": 1.7847222222222225e-05,
+      "loss": 1.2453,
+      "step": 2356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2973268032073975,
+      "learning_rate": 1.7846308479532166e-05,
+      "loss": 1.9077,
+      "step": 2357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1365628242492676,
+      "learning_rate": 1.7845394736842106e-05,
+      "loss": 1.828,
+      "step": 2358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8512721061706543,
+      "learning_rate": 1.784448099415205e-05,
+      "loss": 1.8019,
+      "step": 2359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.971132516860962,
+      "learning_rate": 1.784356725146199e-05,
+      "loss": 2.0319,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.815791130065918,
+      "learning_rate": 1.784265350877193e-05,
+      "loss": 1.9881,
+      "step": 2361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.043490409851074,
+      "learning_rate": 1.7841739766081875e-05,
+      "loss": 1.8063,
+      "step": 2362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.809960126876831,
+      "learning_rate": 1.7840826023391816e-05,
+      "loss": 1.4454,
+      "step": 2363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.022782802581787,
+      "learning_rate": 1.7839912280701756e-05,
+      "loss": 2.0933,
+      "step": 2364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.729248046875,
+      "learning_rate": 1.7838998538011697e-05,
+      "loss": 1.5323,
+      "step": 2365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3517749309539795,
+      "learning_rate": 1.783808479532164e-05,
+      "loss": 2.005,
+      "step": 2366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.686091899871826,
+      "learning_rate": 1.783717105263158e-05,
+      "loss": 1.4643,
+      "step": 2367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.190613031387329,
+      "learning_rate": 1.7836257309941522e-05,
+      "loss": 1.9548,
+      "step": 2368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6203315258026123,
+      "learning_rate": 1.7835343567251466e-05,
+      "loss": 1.964,
+      "step": 2369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.587595224380493,
+      "learning_rate": 1.7834429824561406e-05,
+      "loss": 2.0348,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6614291667938232,
+      "learning_rate": 1.7833516081871347e-05,
+      "loss": 1.325,
+      "step": 2371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.321658134460449,
+      "learning_rate": 1.7832602339181287e-05,
+      "loss": 1.9,
+      "step": 2372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.117687225341797,
+      "learning_rate": 1.783168859649123e-05,
+      "loss": 1.8108,
+      "step": 2373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8733928203582764,
+      "learning_rate": 1.7830774853801172e-05,
+      "loss": 1.6763,
+      "step": 2374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.051464080810547,
+      "learning_rate": 1.7829861111111112e-05,
+      "loss": 1.7327,
+      "step": 2375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9967527389526367,
+      "learning_rate": 1.7828947368421056e-05,
+      "loss": 1.7224,
+      "step": 2376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3001904487609863,
+      "learning_rate": 1.7828033625730997e-05,
+      "loss": 1.9397,
+      "step": 2377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.076704978942871,
+      "learning_rate": 1.7827119883040937e-05,
+      "loss": 1.9348,
+      "step": 2378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.362337827682495,
+      "learning_rate": 1.7826206140350878e-05,
+      "loss": 1.699,
+      "step": 2379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8294661045074463,
+      "learning_rate": 1.7825292397660822e-05,
+      "loss": 2.099,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.977383613586426,
+      "learning_rate": 1.7824378654970762e-05,
+      "loss": 1.4073,
+      "step": 2381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5110273361206055,
+      "learning_rate": 1.7823464912280703e-05,
+      "loss": 2.1726,
+      "step": 2382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.899852991104126,
+      "learning_rate": 1.7822551169590647e-05,
+      "loss": 2.2463,
+      "step": 2383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1141066551208496,
+      "learning_rate": 1.7821637426900587e-05,
+      "loss": 1.9588,
+      "step": 2384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.820843458175659,
+      "learning_rate": 1.7820723684210528e-05,
+      "loss": 1.9682,
+      "step": 2385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.109349489212036,
+      "learning_rate": 1.7819809941520468e-05,
+      "loss": 1.6143,
+      "step": 2386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.483941078186035,
+      "learning_rate": 1.7818896198830412e-05,
+      "loss": 1.7744,
+      "step": 2387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2389659881591797,
+      "learning_rate": 1.7817982456140353e-05,
+      "loss": 2.213,
+      "step": 2388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.242635726928711,
+      "learning_rate": 1.7817068713450293e-05,
+      "loss": 1.888,
+      "step": 2389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3740878105163574,
+      "learning_rate": 1.7816154970760234e-05,
+      "loss": 1.6163,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1719558238983154,
+      "learning_rate": 1.7815241228070178e-05,
+      "loss": 2.2058,
+      "step": 2391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9160637855529785,
+      "learning_rate": 1.7814327485380118e-05,
+      "loss": 1.6694,
+      "step": 2392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.457160234451294,
+      "learning_rate": 1.781341374269006e-05,
+      "loss": 2.0631,
+      "step": 2393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4308652877807617,
+      "learning_rate": 1.7812500000000003e-05,
+      "loss": 1.6151,
+      "step": 2394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.227898597717285,
+      "learning_rate": 1.7811586257309943e-05,
+      "loss": 1.756,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8685824871063232,
+      "learning_rate": 1.7810672514619884e-05,
+      "loss": 1.7421,
+      "step": 2396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8125977516174316,
+      "learning_rate": 1.7809758771929824e-05,
+      "loss": 2.1166,
+      "step": 2397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.797717332839966,
+      "learning_rate": 1.7808845029239768e-05,
+      "loss": 1.6342,
+      "step": 2398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3438637256622314,
+      "learning_rate": 1.780793128654971e-05,
+      "loss": 1.6719,
+      "step": 2399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.657040596008301,
+      "learning_rate": 1.780701754385965e-05,
+      "loss": 2.2346,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3567748069763184,
+      "learning_rate": 1.7806103801169593e-05,
+      "loss": 2.055,
+      "step": 2401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.859867811203003,
+      "learning_rate": 1.7805190058479534e-05,
+      "loss": 1.6524,
+      "step": 2402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9819998741149902,
+      "learning_rate": 1.7804276315789474e-05,
+      "loss": 2.2261,
+      "step": 2403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.182161331176758,
+      "learning_rate": 1.7803362573099415e-05,
+      "loss": 1.7559,
+      "step": 2404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.148635387420654,
+      "learning_rate": 1.780244883040936e-05,
+      "loss": 1.8169,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2523040771484375,
+      "learning_rate": 1.78015350877193e-05,
+      "loss": 2.2627,
+      "step": 2406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9946088790893555,
+      "learning_rate": 1.780062134502924e-05,
+      "loss": 1.458,
+      "step": 2407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7758891582489014,
+      "learning_rate": 1.7799707602339184e-05,
+      "loss": 1.6778,
+      "step": 2408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8212978839874268,
+      "learning_rate": 1.7798793859649124e-05,
+      "loss": 1.6447,
+      "step": 2409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9351916313171387,
+      "learning_rate": 1.7797880116959065e-05,
+      "loss": 2.2207,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3434054851531982,
+      "learning_rate": 1.7796966374269005e-05,
+      "loss": 1.5812,
+      "step": 2411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2062432765960693,
+      "learning_rate": 1.779605263157895e-05,
+      "loss": 1.6832,
+      "step": 2412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9394233226776123,
+      "learning_rate": 1.779513888888889e-05,
+      "loss": 1.3983,
+      "step": 2413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1705496311187744,
+      "learning_rate": 1.779422514619883e-05,
+      "loss": 1.6646,
+      "step": 2414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1191835403442383,
+      "learning_rate": 1.7793311403508774e-05,
+      "loss": 1.9505,
+      "step": 2415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3613741397857666,
+      "learning_rate": 1.7792397660818715e-05,
+      "loss": 1.7908,
+      "step": 2416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9623677730560303,
+      "learning_rate": 1.7791483918128655e-05,
+      "loss": 1.7233,
+      "step": 2417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1861462593078613,
+      "learning_rate": 1.7790570175438596e-05,
+      "loss": 1.5856,
+      "step": 2418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2053236961364746,
+      "learning_rate": 1.778965643274854e-05,
+      "loss": 1.6619,
+      "step": 2419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4938628673553467,
+      "learning_rate": 1.778874269005848e-05,
+      "loss": 2.0855,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.187936305999756,
+      "learning_rate": 1.778782894736842e-05,
+      "loss": 1.8607,
+      "step": 2421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.13988995552063,
+      "learning_rate": 1.7786915204678365e-05,
+      "loss": 1.6362,
+      "step": 2422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2416765689849854,
+      "learning_rate": 1.7786001461988305e-05,
+      "loss": 2.1994,
+      "step": 2423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.635049819946289,
+      "learning_rate": 1.7785087719298246e-05,
+      "loss": 1.8921,
+      "step": 2424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.003246784210205,
+      "learning_rate": 1.7784173976608186e-05,
+      "loss": 2.3755,
+      "step": 2425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2322239875793457,
+      "learning_rate": 1.778326023391813e-05,
+      "loss": 1.6822,
+      "step": 2426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.660566806793213,
+      "learning_rate": 1.778234649122807e-05,
+      "loss": 2.1335,
+      "step": 2427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7204439640045166,
+      "learning_rate": 1.778143274853801e-05,
+      "loss": 2.3545,
+      "step": 2428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.273789644241333,
+      "learning_rate": 1.7780519005847955e-05,
+      "loss": 1.9155,
+      "step": 2429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8292853832244873,
+      "learning_rate": 1.7779605263157896e-05,
+      "loss": 1.5808,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.690399646759033,
+      "learning_rate": 1.7778691520467836e-05,
+      "loss": 2.0081,
+      "step": 2431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.110264778137207,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.9578,
+      "step": 2432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.656806230545044,
+      "learning_rate": 1.777686403508772e-05,
+      "loss": 1.9568,
+      "step": 2433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.110957145690918,
+      "learning_rate": 1.777595029239766e-05,
+      "loss": 1.7185,
+      "step": 2434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.796110153198242,
+      "learning_rate": 1.77750365497076e-05,
+      "loss": 1.4837,
+      "step": 2435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7498247623443604,
+      "learning_rate": 1.7774122807017545e-05,
+      "loss": 2.3063,
+      "step": 2436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8897757530212402,
+      "learning_rate": 1.777320906432749e-05,
+      "loss": 1.5414,
+      "step": 2437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2702829837799072,
+      "learning_rate": 1.777229532163743e-05,
+      "loss": 1.8237,
+      "step": 2438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.659505844116211,
+      "learning_rate": 1.777138157894737e-05,
+      "loss": 1.9374,
+      "step": 2439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.69710636138916,
+      "learning_rate": 1.777046783625731e-05,
+      "loss": 1.8046,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1809439659118652,
+      "learning_rate": 1.7769554093567255e-05,
+      "loss": 1.6415,
+      "step": 2441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.298020124435425,
+      "learning_rate": 1.7768640350877195e-05,
+      "loss": 1.9247,
+      "step": 2442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.148529529571533,
+      "learning_rate": 1.7767726608187136e-05,
+      "loss": 1.7487,
+      "step": 2443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0881502628326416,
+      "learning_rate": 1.776681286549708e-05,
+      "loss": 1.9851,
+      "step": 2444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1332478523254395,
+      "learning_rate": 1.776589912280702e-05,
+      "loss": 1.9092,
+      "step": 2445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.423842906951904,
+      "learning_rate": 1.776498538011696e-05,
+      "loss": 2.4655,
+      "step": 2446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7712831497192383,
+      "learning_rate": 1.77640716374269e-05,
+      "loss": 1.5077,
+      "step": 2447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.668642520904541,
+      "learning_rate": 1.7763157894736845e-05,
+      "loss": 1.979,
+      "step": 2448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.129690647125244,
+      "learning_rate": 1.7762244152046786e-05,
+      "loss": 1.5362,
+      "step": 2449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4209721088409424,
+      "learning_rate": 1.7761330409356726e-05,
+      "loss": 2.0141,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6672394275665283,
+      "learning_rate": 1.776041666666667e-05,
+      "loss": 2.1755,
+      "step": 2451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.396003007888794,
+      "learning_rate": 1.775950292397661e-05,
+      "loss": 2.1817,
+      "step": 2452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.615095615386963,
+      "learning_rate": 1.775858918128655e-05,
+      "loss": 2.0937,
+      "step": 2453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.285179376602173,
+      "learning_rate": 1.7757675438596492e-05,
+      "loss": 1.7576,
+      "step": 2454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.439733505249023,
+      "learning_rate": 1.7756761695906436e-05,
+      "loss": 1.9993,
+      "step": 2455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.841787576675415,
+      "learning_rate": 1.7755847953216376e-05,
+      "loss": 1.7152,
+      "step": 2456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3807313442230225,
+      "learning_rate": 1.7754934210526317e-05,
+      "loss": 1.9602,
+      "step": 2457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 15.588988304138184,
+      "learning_rate": 1.775402046783626e-05,
+      "loss": 1.7785,
+      "step": 2458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6133458614349365,
+      "learning_rate": 1.77531067251462e-05,
+      "loss": 1.7413,
+      "step": 2459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8370044231414795,
+      "learning_rate": 1.7752192982456142e-05,
+      "loss": 2.0012,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.14031457901001,
+      "learning_rate": 1.7751279239766082e-05,
+      "loss": 2.7057,
+      "step": 2461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.920795202255249,
+      "learning_rate": 1.7750365497076026e-05,
+      "loss": 2.2183,
+      "step": 2462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8905868530273438,
+      "learning_rate": 1.7749451754385967e-05,
+      "loss": 1.8798,
+      "step": 2463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.663860559463501,
+      "learning_rate": 1.7748538011695907e-05,
+      "loss": 1.8974,
+      "step": 2464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6330041885375977,
+      "learning_rate": 1.774762426900585e-05,
+      "loss": 1.938,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.488787889480591,
+      "learning_rate": 1.7746710526315792e-05,
+      "loss": 1.654,
+      "step": 2466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.986536979675293,
+      "learning_rate": 1.7745796783625732e-05,
+      "loss": 1.5002,
+      "step": 2467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4828004837036133,
+      "learning_rate": 1.7744883040935673e-05,
+      "loss": 1.7099,
+      "step": 2468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.253345489501953,
+      "learning_rate": 1.7743969298245617e-05,
+      "loss": 1.8214,
+      "step": 2469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.651587963104248,
+      "learning_rate": 1.7743055555555557e-05,
+      "loss": 1.5084,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.258056879043579,
+      "learning_rate": 1.7742141812865498e-05,
+      "loss": 2.0904,
+      "step": 2471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.24908709526062,
+      "learning_rate": 1.7741228070175442e-05,
+      "loss": 1.8834,
+      "step": 2472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.943521738052368,
+      "learning_rate": 1.7740314327485382e-05,
+      "loss": 2.5679,
+      "step": 2473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2826457023620605,
+      "learning_rate": 1.7739400584795323e-05,
+      "loss": 2.0217,
+      "step": 2474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.218604326248169,
+      "learning_rate": 1.7738486842105263e-05,
+      "loss": 1.9485,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9525697231292725,
+      "learning_rate": 1.7737573099415207e-05,
+      "loss": 1.748,
+      "step": 2476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.373831272125244,
+      "learning_rate": 1.7736659356725148e-05,
+      "loss": 2.0475,
+      "step": 2477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4142258167266846,
+      "learning_rate": 1.7735745614035088e-05,
+      "loss": 1.9093,
+      "step": 2478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9244401454925537,
+      "learning_rate": 1.7734831871345032e-05,
+      "loss": 2.0358,
+      "step": 2479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.287858486175537,
+      "learning_rate": 1.7733918128654973e-05,
+      "loss": 1.8869,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6727755069732666,
+      "learning_rate": 1.7733004385964913e-05,
+      "loss": 2.2507,
+      "step": 2481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.078418731689453,
+      "learning_rate": 1.7732090643274854e-05,
+      "loss": 1.8569,
+      "step": 2482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.53843355178833,
+      "learning_rate": 1.7731176900584798e-05,
+      "loss": 1.8247,
+      "step": 2483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.171365261077881,
+      "learning_rate": 1.7730263157894738e-05,
+      "loss": 1.4781,
+      "step": 2484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.686250925064087,
+      "learning_rate": 1.772934941520468e-05,
+      "loss": 2.1902,
+      "step": 2485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9632999897003174,
+      "learning_rate": 1.7728435672514623e-05,
+      "loss": 1.7515,
+      "step": 2486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1866185665130615,
+      "learning_rate": 1.7727521929824563e-05,
+      "loss": 1.6523,
+      "step": 2487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.140347480773926,
+      "learning_rate": 1.7726608187134504e-05,
+      "loss": 1.7858,
+      "step": 2488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9320597648620605,
+      "learning_rate": 1.7725694444444444e-05,
+      "loss": 2.0978,
+      "step": 2489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.227847099304199,
+      "learning_rate": 1.7724780701754388e-05,
+      "loss": 2.0103,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1693599224090576,
+      "learning_rate": 1.772386695906433e-05,
+      "loss": 1.7931,
+      "step": 2491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.108831882476807,
+      "learning_rate": 1.772295321637427e-05,
+      "loss": 2.0867,
+      "step": 2492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9438750743865967,
+      "learning_rate": 1.7722039473684213e-05,
+      "loss": 1.5702,
+      "step": 2493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.496263027191162,
+      "learning_rate": 1.7721125730994154e-05,
+      "loss": 1.9808,
+      "step": 2494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8716137409210205,
+      "learning_rate": 1.7720211988304094e-05,
+      "loss": 2.2646,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1060988903045654,
+      "learning_rate": 1.7719298245614035e-05,
+      "loss": 1.8668,
+      "step": 2496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.102616786956787,
+      "learning_rate": 1.771838450292398e-05,
+      "loss": 1.7776,
+      "step": 2497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2536864280700684,
+      "learning_rate": 1.771747076023392e-05,
+      "loss": 1.7034,
+      "step": 2498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.938295364379883,
+      "learning_rate": 1.771655701754386e-05,
+      "loss": 1.8544,
+      "step": 2499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6080124378204346,
+      "learning_rate": 1.7715643274853804e-05,
+      "loss": 1.6628,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3720264434814453,
+      "learning_rate": 1.7714729532163744e-05,
+      "loss": 1.8266,
+      "step": 2501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1985368728637695,
+      "learning_rate": 1.7713815789473685e-05,
+      "loss": 1.9388,
+      "step": 2502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.172682285308838,
+      "learning_rate": 1.7712902046783625e-05,
+      "loss": 1.5583,
+      "step": 2503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6328351497650146,
+      "learning_rate": 1.771198830409357e-05,
+      "loss": 1.748,
+      "step": 2504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.074177265167236,
+      "learning_rate": 1.771107456140351e-05,
+      "loss": 1.7345,
+      "step": 2505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.356972694396973,
+      "learning_rate": 1.771016081871345e-05,
+      "loss": 2.0661,
+      "step": 2506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.099945068359375,
+      "learning_rate": 1.7709247076023394e-05,
+      "loss": 1.6968,
+      "step": 2507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.123664617538452,
+      "learning_rate": 1.7708333333333335e-05,
+      "loss": 1.6163,
+      "step": 2508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1622982025146484,
+      "learning_rate": 1.7707419590643275e-05,
+      "loss": 1.5319,
+      "step": 2509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.347419738769531,
+      "learning_rate": 1.7706505847953216e-05,
+      "loss": 1.9133,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.937251567840576,
+      "learning_rate": 1.770559210526316e-05,
+      "loss": 1.649,
+      "step": 2511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3048036098480225,
+      "learning_rate": 1.77046783625731e-05,
+      "loss": 1.7842,
+      "step": 2512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1962666511535645,
+      "learning_rate": 1.770376461988304e-05,
+      "loss": 1.9331,
+      "step": 2513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.805332660675049,
+      "learning_rate": 1.7702850877192985e-05,
+      "loss": 1.4856,
+      "step": 2514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.320110559463501,
+      "learning_rate": 1.7701937134502925e-05,
+      "loss": 2.0014,
+      "step": 2515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.527543306350708,
+      "learning_rate": 1.7701023391812866e-05,
+      "loss": 2.1343,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.728086233139038,
+      "learning_rate": 1.7700109649122806e-05,
+      "loss": 1.9774,
+      "step": 2517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.320300817489624,
+      "learning_rate": 1.769919590643275e-05,
+      "loss": 2.1403,
+      "step": 2518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6999411582946777,
+      "learning_rate": 1.769828216374269e-05,
+      "loss": 1.4637,
+      "step": 2519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.308513879776001,
+      "learning_rate": 1.769736842105263e-05,
+      "loss": 1.7654,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.715714454650879,
+      "learning_rate": 1.7696454678362575e-05,
+      "loss": 1.8086,
+      "step": 2521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6389687061309814,
+      "learning_rate": 1.7695540935672516e-05,
+      "loss": 1.969,
+      "step": 2522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4280290603637695,
+      "learning_rate": 1.769462719298246e-05,
+      "loss": 1.7239,
+      "step": 2523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2645742893218994,
+      "learning_rate": 1.76937134502924e-05,
+      "loss": 1.9963,
+      "step": 2524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7885549068450928,
+      "learning_rate": 1.769279970760234e-05,
+      "loss": 2.2389,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3980395793914795,
+      "learning_rate": 1.7691885964912284e-05,
+      "loss": 1.9224,
+      "step": 2526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3409054279327393,
+      "learning_rate": 1.7690972222222225e-05,
+      "loss": 1.7076,
+      "step": 2527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3322064876556396,
+      "learning_rate": 1.7690058479532165e-05,
+      "loss": 1.6789,
+      "step": 2528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.975412368774414,
+      "learning_rate": 1.7689144736842106e-05,
+      "loss": 2.2328,
+      "step": 2529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3601391315460205,
+      "learning_rate": 1.768823099415205e-05,
+      "loss": 1.7409,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.162658214569092,
+      "learning_rate": 1.768731725146199e-05,
+      "loss": 1.8087,
+      "step": 2531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.562309980392456,
+      "learning_rate": 1.768640350877193e-05,
+      "loss": 2.0603,
+      "step": 2532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.365509510040283,
+      "learning_rate": 1.7685489766081875e-05,
+      "loss": 1.9735,
+      "step": 2533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.341247081756592,
+      "learning_rate": 1.7684576023391815e-05,
+      "loss": 1.9605,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0109715461730957,
+      "learning_rate": 1.7683662280701756e-05,
+      "loss": 1.7824,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.859222173690796,
+      "learning_rate": 1.7682748538011696e-05,
+      "loss": 1.7917,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.816470146179199,
+      "learning_rate": 1.768183479532164e-05,
+      "loss": 1.5619,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.756808280944824,
+      "learning_rate": 1.768092105263158e-05,
+      "loss": 1.5997,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9187819957733154,
+      "learning_rate": 1.768000730994152e-05,
+      "loss": 1.7694,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.778564691543579,
+      "learning_rate": 1.7679093567251465e-05,
+      "loss": 2.2735,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.49438214302063,
+      "learning_rate": 1.7678179824561406e-05,
+      "loss": 1.6636,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.198237419128418,
+      "learning_rate": 1.7677266081871346e-05,
+      "loss": 1.8504,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.099489212036133,
+      "learning_rate": 1.7676352339181287e-05,
+      "loss": 1.7734,
+      "step": 2543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5576727390289307,
+      "learning_rate": 1.767543859649123e-05,
+      "loss": 1.9548,
+      "step": 2544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9767770767211914,
+      "learning_rate": 1.767452485380117e-05,
+      "loss": 1.8577,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.053501605987549,
+      "learning_rate": 1.7673611111111112e-05,
+      "loss": 1.8705,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2926199436187744,
+      "learning_rate": 1.7672697368421056e-05,
+      "loss": 1.6272,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7900047302246094,
+      "learning_rate": 1.7671783625730996e-05,
+      "loss": 1.5897,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.217372179031372,
+      "learning_rate": 1.7670869883040937e-05,
+      "loss": 1.7644,
+      "step": 2549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6220571994781494,
+      "learning_rate": 1.7669956140350877e-05,
+      "loss": 1.6587,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8490047454833984,
+      "learning_rate": 1.766904239766082e-05,
+      "loss": 2.0048,
+      "step": 2551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.792701244354248,
+      "learning_rate": 1.7668128654970762e-05,
+      "loss": 1.5616,
+      "step": 2552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.392512083053589,
+      "learning_rate": 1.7667214912280702e-05,
+      "loss": 2.2184,
+      "step": 2553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.367055892944336,
+      "learning_rate": 1.7666301169590646e-05,
+      "loss": 1.9799,
+      "step": 2554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3201348781585693,
+      "learning_rate": 1.7665387426900587e-05,
+      "loss": 2.1357,
+      "step": 2555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.766364336013794,
+      "learning_rate": 1.7664473684210527e-05,
+      "loss": 2.4266,
+      "step": 2556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6506998538970947,
+      "learning_rate": 1.7663559941520468e-05,
+      "loss": 1.6483,
+      "step": 2557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5887629985809326,
+      "learning_rate": 1.7662646198830412e-05,
+      "loss": 1.9546,
+      "step": 2558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.186765193939209,
+      "learning_rate": 1.7661732456140352e-05,
+      "loss": 1.7595,
+      "step": 2559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.737819194793701,
+      "learning_rate": 1.7660818713450293e-05,
+      "loss": 1.7496,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6634674072265625,
+      "learning_rate": 1.7659904970760237e-05,
+      "loss": 1.6612,
+      "step": 2561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.507413148880005,
+      "learning_rate": 1.7658991228070177e-05,
+      "loss": 1.826,
+      "step": 2562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2640275955200195,
+      "learning_rate": 1.7658077485380118e-05,
+      "loss": 1.8626,
+      "step": 2563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.089165449142456,
+      "learning_rate": 1.765716374269006e-05,
+      "loss": 1.7364,
+      "step": 2564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8243815898895264,
+      "learning_rate": 1.7656250000000002e-05,
+      "loss": 1.7305,
+      "step": 2565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.6238322257995605,
+      "learning_rate": 1.7655336257309943e-05,
+      "loss": 1.8922,
+      "step": 2566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1045920848846436,
+      "learning_rate": 1.7654422514619883e-05,
+      "loss": 1.7253,
+      "step": 2567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4028708934783936,
+      "learning_rate": 1.7653508771929827e-05,
+      "loss": 1.906,
+      "step": 2568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.269893169403076,
+      "learning_rate": 1.7652595029239768e-05,
+      "loss": 1.8823,
+      "step": 2569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6626060009002686,
+      "learning_rate": 1.7651681286549708e-05,
+      "loss": 1.8511,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.01861572265625,
+      "learning_rate": 1.765076754385965e-05,
+      "loss": 1.6768,
+      "step": 2571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.110417366027832,
+      "learning_rate": 1.7649853801169593e-05,
+      "loss": 1.6205,
+      "step": 2572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4030921459198,
+      "learning_rate": 1.7648940058479533e-05,
+      "loss": 2.0884,
+      "step": 2573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7047038078308105,
+      "learning_rate": 1.7648026315789474e-05,
+      "loss": 1.5013,
+      "step": 2574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.797086715698242,
+      "learning_rate": 1.7647112573099418e-05,
+      "loss": 2.1096,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7811689376831055,
+      "learning_rate": 1.7646198830409358e-05,
+      "loss": 1.5707,
+      "step": 2576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2413690090179443,
+      "learning_rate": 1.76452850877193e-05,
+      "loss": 1.5759,
+      "step": 2577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6799159049987793,
+      "learning_rate": 1.764437134502924e-05,
+      "loss": 1.5861,
+      "step": 2578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.497640371322632,
+      "learning_rate": 1.7643457602339183e-05,
+      "loss": 1.9963,
+      "step": 2579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6872119903564453,
+      "learning_rate": 1.7642543859649124e-05,
+      "loss": 2.4788,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.765531063079834,
+      "learning_rate": 1.7641630116959064e-05,
+      "loss": 2.0905,
+      "step": 2581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.972140312194824,
+      "learning_rate": 1.7640716374269008e-05,
+      "loss": 1.6893,
+      "step": 2582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7508933544158936,
+      "learning_rate": 1.763980263157895e-05,
+      "loss": 1.7448,
+      "step": 2583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1332216262817383,
+      "learning_rate": 1.763888888888889e-05,
+      "loss": 1.5675,
+      "step": 2584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.028195858001709,
+      "learning_rate": 1.763797514619883e-05,
+      "loss": 1.5624,
+      "step": 2585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3250010013580322,
+      "learning_rate": 1.7637061403508774e-05,
+      "loss": 2.0964,
+      "step": 2586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.040179491043091,
+      "learning_rate": 1.7636147660818714e-05,
+      "loss": 1.3877,
+      "step": 2587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.130342960357666,
+      "learning_rate": 1.7635233918128655e-05,
+      "loss": 2.1395,
+      "step": 2588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9040095806121826,
+      "learning_rate": 1.76343201754386e-05,
+      "loss": 1.7148,
+      "step": 2589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0882396697998047,
+      "learning_rate": 1.763340643274854e-05,
+      "loss": 1.6657,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.058547019958496,
+      "learning_rate": 1.763249269005848e-05,
+      "loss": 2.2598,
+      "step": 2591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7472012042999268,
+      "learning_rate": 1.763157894736842e-05,
+      "loss": 2.4621,
+      "step": 2592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 42.92102813720703,
+      "learning_rate": 1.7630665204678364e-05,
+      "loss": 1.7276,
+      "step": 2593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.075150966644287,
+      "learning_rate": 1.7629751461988305e-05,
+      "loss": 1.7207,
+      "step": 2594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2305173873901367,
+      "learning_rate": 1.7628837719298245e-05,
+      "loss": 1.9656,
+      "step": 2595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.252394676208496,
+      "learning_rate": 1.762792397660819e-05,
+      "loss": 1.8679,
+      "step": 2596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0278453826904297,
+      "learning_rate": 1.762701023391813e-05,
+      "loss": 1.6596,
+      "step": 2597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4868216514587402,
+      "learning_rate": 1.762609649122807e-05,
+      "loss": 2.0818,
+      "step": 2598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7522847652435303,
+      "learning_rate": 1.762518274853801e-05,
+      "loss": 1.5806,
+      "step": 2599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1654655933380127,
+      "learning_rate": 1.7624269005847955e-05,
+      "loss": 1.679,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.041954517364502,
+      "learning_rate": 1.7623355263157895e-05,
+      "loss": 1.5982,
+      "step": 2601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.004634380340576,
+      "learning_rate": 1.7622441520467836e-05,
+      "loss": 2.2036,
+      "step": 2602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.954258918762207,
+      "learning_rate": 1.762152777777778e-05,
+      "loss": 1.9468,
+      "step": 2603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9265456199645996,
+      "learning_rate": 1.762061403508772e-05,
+      "loss": 1.4509,
+      "step": 2604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8688764572143555,
+      "learning_rate": 1.761970029239766e-05,
+      "loss": 2.0064,
+      "step": 2605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.825951099395752,
+      "learning_rate": 1.76187865497076e-05,
+      "loss": 2.1188,
+      "step": 2606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.843437671661377,
+      "learning_rate": 1.7617872807017545e-05,
+      "loss": 1.4828,
+      "step": 2607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4597690105438232,
+      "learning_rate": 1.761695906432749e-05,
+      "loss": 1.6794,
+      "step": 2608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8998076915740967,
+      "learning_rate": 1.761604532163743e-05,
+      "loss": 1.7383,
+      "step": 2609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9264492988586426,
+      "learning_rate": 1.761513157894737e-05,
+      "loss": 1.9383,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7734973430633545,
+      "learning_rate": 1.7614217836257314e-05,
+      "loss": 1.9747,
+      "step": 2611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.998638868331909,
+      "learning_rate": 1.7613304093567254e-05,
+      "loss": 1.5301,
+      "step": 2612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.19736647605896,
+      "learning_rate": 1.7612390350877195e-05,
+      "loss": 1.7225,
+      "step": 2613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5032949447631836,
+      "learning_rate": 1.7611476608187136e-05,
+      "loss": 1.9889,
+      "step": 2614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1157565116882324,
+      "learning_rate": 1.761056286549708e-05,
+      "loss": 1.5737,
+      "step": 2615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.174581289291382,
+      "learning_rate": 1.760964912280702e-05,
+      "loss": 1.6409,
+      "step": 2616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.203972339630127,
+      "learning_rate": 1.760873538011696e-05,
+      "loss": 1.8462,
+      "step": 2617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.014742374420166,
+      "learning_rate": 1.7607821637426904e-05,
+      "loss": 1.9598,
+      "step": 2618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.208650588989258,
+      "learning_rate": 1.7606907894736845e-05,
+      "loss": 1.8112,
+      "step": 2619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0170700550079346,
+      "learning_rate": 1.7605994152046785e-05,
+      "loss": 1.5835,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3328733444213867,
+      "learning_rate": 1.7605080409356726e-05,
+      "loss": 2.0494,
+      "step": 2621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2106130123138428,
+      "learning_rate": 1.760416666666667e-05,
+      "loss": 1.8039,
+      "step": 2622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.595116376876831,
+      "learning_rate": 1.760325292397661e-05,
+      "loss": 1.5074,
+      "step": 2623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6715009212493896,
+      "learning_rate": 1.760233918128655e-05,
+      "loss": 1.8332,
+      "step": 2624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5680320262908936,
+      "learning_rate": 1.7601425438596495e-05,
+      "loss": 2.0694,
+      "step": 2625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.766078472137451,
+      "learning_rate": 1.7600511695906435e-05,
+      "loss": 2.2247,
+      "step": 2626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1669960021972656,
+      "learning_rate": 1.7599597953216376e-05,
+      "loss": 2.0089,
+      "step": 2627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.489938259124756,
+      "learning_rate": 1.7598684210526316e-05,
+      "loss": 2.0158,
+      "step": 2628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.950974464416504,
+      "learning_rate": 1.759777046783626e-05,
+      "loss": 1.6065,
+      "step": 2629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.360391855239868,
+      "learning_rate": 1.75968567251462e-05,
+      "loss": 1.9176,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9054269790649414,
+      "learning_rate": 1.759594298245614e-05,
+      "loss": 1.629,
+      "step": 2631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2417447566986084,
+      "learning_rate": 1.7595029239766085e-05,
+      "loss": 1.5541,
+      "step": 2632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9715445041656494,
+      "learning_rate": 1.7594115497076026e-05,
+      "loss": 1.4698,
+      "step": 2633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4055097103118896,
+      "learning_rate": 1.7593201754385966e-05,
+      "loss": 1.8297,
+      "step": 2634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0649402141571045,
+      "learning_rate": 1.7592288011695907e-05,
+      "loss": 1.7905,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.418107986450195,
+      "learning_rate": 1.759137426900585e-05,
+      "loss": 2.5291,
+      "step": 2636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2787411212921143,
+      "learning_rate": 1.759046052631579e-05,
+      "loss": 1.8066,
+      "step": 2637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7016682624816895,
+      "learning_rate": 1.7589546783625732e-05,
+      "loss": 1.8186,
+      "step": 2638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3142807483673096,
+      "learning_rate": 1.7588633040935676e-05,
+      "loss": 1.6066,
+      "step": 2639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.43341064453125,
+      "learning_rate": 1.7587719298245616e-05,
+      "loss": 2.1043,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0850393772125244,
+      "learning_rate": 1.7586805555555557e-05,
+      "loss": 1.8706,
+      "step": 2641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.28659725189209,
+      "learning_rate": 1.7585891812865497e-05,
+      "loss": 1.854,
+      "step": 2642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.046062707901001,
+      "learning_rate": 1.758497807017544e-05,
+      "loss": 1.8397,
+      "step": 2643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2618393898010254,
+      "learning_rate": 1.7584064327485382e-05,
+      "loss": 1.8657,
+      "step": 2644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1359922885894775,
+      "learning_rate": 1.7583150584795322e-05,
+      "loss": 1.5204,
+      "step": 2645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.203941822052002,
+      "learning_rate": 1.7582236842105266e-05,
+      "loss": 1.8732,
+      "step": 2646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.056548833847046,
+      "learning_rate": 1.7581323099415207e-05,
+      "loss": 1.9256,
+      "step": 2647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8398377895355225,
+      "learning_rate": 1.7580409356725147e-05,
+      "loss": 1.5886,
+      "step": 2648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1486713886260986,
+      "learning_rate": 1.7579495614035088e-05,
+      "loss": 1.7385,
+      "step": 2649
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2430686950683594,
+      "learning_rate": 1.7578581871345032e-05,
+      "loss": 1.7785,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.945873975753784,
+      "learning_rate": 1.7577668128654972e-05,
+      "loss": 1.7781,
+      "step": 2651
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2178711891174316,
+      "learning_rate": 1.7576754385964913e-05,
+      "loss": 1.7537,
+      "step": 2652
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6025025844573975,
+      "learning_rate": 1.7575840643274857e-05,
+      "loss": 1.894,
+      "step": 2653
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2475366592407227,
+      "learning_rate": 1.7574926900584797e-05,
+      "loss": 1.8454,
+      "step": 2654
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.405622720718384,
+      "learning_rate": 1.7574013157894738e-05,
+      "loss": 1.9411,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5418508052825928,
+      "learning_rate": 1.757309941520468e-05,
+      "loss": 1.5471,
+      "step": 2656
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2616589069366455,
+      "learning_rate": 1.7572185672514622e-05,
+      "loss": 2.0967,
+      "step": 2657
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.575958490371704,
+      "learning_rate": 1.7571271929824563e-05,
+      "loss": 2.069,
+      "step": 2658
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.60434889793396,
+      "learning_rate": 1.7570358187134503e-05,
+      "loss": 1.6571,
+      "step": 2659
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.445313930511475,
+      "learning_rate": 1.7569444444444447e-05,
+      "loss": 1.8302,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2443346977233887,
+      "learning_rate": 1.7568530701754388e-05,
+      "loss": 1.8175,
+      "step": 2661
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9155046939849854,
+      "learning_rate": 1.7567616959064328e-05,
+      "loss": 1.5436,
+      "step": 2662
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.819413423538208,
+      "learning_rate": 1.756670321637427e-05,
+      "loss": 1.4897,
+      "step": 2663
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.08935809135437,
+      "learning_rate": 1.7565789473684213e-05,
+      "loss": 1.5952,
+      "step": 2664
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.004251003265381,
+      "learning_rate": 1.7564875730994153e-05,
+      "loss": 1.7061,
+      "step": 2665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4102730751037598,
+      "learning_rate": 1.7563961988304094e-05,
+      "loss": 1.4954,
+      "step": 2666
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4931230545043945,
+      "learning_rate": 1.7563048245614038e-05,
+      "loss": 1.9129,
+      "step": 2667
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.140371799468994,
+      "learning_rate": 1.7562134502923978e-05,
+      "loss": 1.7277,
+      "step": 2668
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1007096767425537,
+      "learning_rate": 1.756122076023392e-05,
+      "loss": 1.6194,
+      "step": 2669
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.00425124168396,
+      "learning_rate": 1.756030701754386e-05,
+      "loss": 1.5406,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1255645751953125,
+      "learning_rate": 1.7559393274853803e-05,
+      "loss": 1.5809,
+      "step": 2671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2986912727355957,
+      "learning_rate": 1.7558479532163744e-05,
+      "loss": 2.1516,
+      "step": 2672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.113476276397705,
+      "learning_rate": 1.7557565789473684e-05,
+      "loss": 1.6768,
+      "step": 2673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0607144832611084,
+      "learning_rate": 1.7556652046783628e-05,
+      "loss": 1.6611,
+      "step": 2674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.965559959411621,
+      "learning_rate": 1.755573830409357e-05,
+      "loss": 1.7393,
+      "step": 2675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9469072818756104,
+      "learning_rate": 1.755482456140351e-05,
+      "loss": 1.8711,
+      "step": 2676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9817116260528564,
+      "learning_rate": 1.755391081871345e-05,
+      "loss": 1.9064,
+      "step": 2677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.544843673706055,
+      "learning_rate": 1.7552997076023394e-05,
+      "loss": 2.4715,
+      "step": 2678
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9745423793792725,
+      "learning_rate": 1.7552083333333334e-05,
+      "loss": 1.6784,
+      "step": 2679
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8499302864074707,
+      "learning_rate": 1.7551169590643275e-05,
+      "loss": 2.1293,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.327683448791504,
+      "learning_rate": 1.755025584795322e-05,
+      "loss": 1.8752,
+      "step": 2681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4364256858825684,
+      "learning_rate": 1.754934210526316e-05,
+      "loss": 1.7475,
+      "step": 2682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3887782096862793,
+      "learning_rate": 1.75484283625731e-05,
+      "loss": 1.7783,
+      "step": 2683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.319347381591797,
+      "learning_rate": 1.754751461988304e-05,
+      "loss": 1.6396,
+      "step": 2684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.203977346420288,
+      "learning_rate": 1.7546600877192984e-05,
+      "loss": 2.1534,
+      "step": 2685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0969672203063965,
+      "learning_rate": 1.7545687134502925e-05,
+      "loss": 1.8926,
+      "step": 2686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3527681827545166,
+      "learning_rate": 1.7544773391812865e-05,
+      "loss": 1.8252,
+      "step": 2687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7334797382354736,
+      "learning_rate": 1.754385964912281e-05,
+      "loss": 2.1064,
+      "step": 2688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.368467330932617,
+      "learning_rate": 1.754294590643275e-05,
+      "loss": 1.7155,
+      "step": 2689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.0448994636535645,
+      "learning_rate": 1.754203216374269e-05,
+      "loss": 1.9944,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.731544017791748,
+      "learning_rate": 1.754111842105263e-05,
+      "loss": 1.7578,
+      "step": 2691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5563862323760986,
+      "learning_rate": 1.7540204678362575e-05,
+      "loss": 1.8388,
+      "step": 2692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.992419958114624,
+      "learning_rate": 1.7539290935672515e-05,
+      "loss": 1.853,
+      "step": 2693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.874338150024414,
+      "learning_rate": 1.753837719298246e-05,
+      "loss": 2.2614,
+      "step": 2694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.154977321624756,
+      "learning_rate": 1.75374634502924e-05,
+      "loss": 1.9124,
+      "step": 2695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2527341842651367,
+      "learning_rate": 1.753654970760234e-05,
+      "loss": 1.6837,
+      "step": 2696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.123645305633545,
+      "learning_rate": 1.7535635964912284e-05,
+      "loss": 1.6676,
+      "step": 2697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3674087524414062,
+      "learning_rate": 1.7534722222222225e-05,
+      "loss": 1.6818,
+      "step": 2698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.067234039306641,
+      "learning_rate": 1.7533808479532165e-05,
+      "loss": 1.74,
+      "step": 2699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.563063383102417,
+      "learning_rate": 1.753289473684211e-05,
+      "loss": 2.3421,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0476293563842773,
+      "learning_rate": 1.753198099415205e-05,
+      "loss": 1.5248,
+      "step": 2701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0688817501068115,
+      "learning_rate": 1.753106725146199e-05,
+      "loss": 1.5879,
+      "step": 2702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7860617637634277,
+      "learning_rate": 1.753015350877193e-05,
+      "loss": 1.1475,
+      "step": 2703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8655929565429688,
+      "learning_rate": 1.7529239766081874e-05,
+      "loss": 1.7345,
+      "step": 2704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.04779314994812,
+      "learning_rate": 1.7528326023391815e-05,
+      "loss": 1.7918,
+      "step": 2705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.137045383453369,
+      "learning_rate": 1.7527412280701756e-05,
+      "loss": 1.996,
+      "step": 2706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.801429510116577,
+      "learning_rate": 1.75264985380117e-05,
+      "loss": 1.7338,
+      "step": 2707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2554707527160645,
+      "learning_rate": 1.752558479532164e-05,
+      "loss": 1.9938,
+      "step": 2708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.110925197601318,
+      "learning_rate": 1.752467105263158e-05,
+      "loss": 2.1656,
+      "step": 2709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3494184017181396,
+      "learning_rate": 1.752375730994152e-05,
+      "loss": 2.1644,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1680662631988525,
+      "learning_rate": 1.7522843567251465e-05,
+      "loss": 1.8484,
+      "step": 2711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0307493209838867,
+      "learning_rate": 1.7521929824561405e-05,
+      "loss": 1.4211,
+      "step": 2712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5166494846343994,
+      "learning_rate": 1.7521016081871346e-05,
+      "loss": 2.1071,
+      "step": 2713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.297717571258545,
+      "learning_rate": 1.752010233918129e-05,
+      "loss": 1.5744,
+      "step": 2714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.8297553062438965,
+      "learning_rate": 1.751918859649123e-05,
+      "loss": 2.2605,
+      "step": 2715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.026991605758667,
+      "learning_rate": 1.751827485380117e-05,
+      "loss": 1.8487,
+      "step": 2716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9043684005737305,
+      "learning_rate": 1.751736111111111e-05,
+      "loss": 1.8758,
+      "step": 2717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1146836280822754,
+      "learning_rate": 1.7516447368421055e-05,
+      "loss": 1.8971,
+      "step": 2718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6586453914642334,
+      "learning_rate": 1.7515533625730996e-05,
+      "loss": 1.8173,
+      "step": 2719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.303483486175537,
+      "learning_rate": 1.7514619883040936e-05,
+      "loss": 2.055,
+      "step": 2720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1643052101135254,
+      "learning_rate": 1.751370614035088e-05,
+      "loss": 1.614,
+      "step": 2721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.068394184112549,
+      "learning_rate": 1.751279239766082e-05,
+      "loss": 1.9896,
+      "step": 2722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9444820880889893,
+      "learning_rate": 1.751187865497076e-05,
+      "loss": 2.0327,
+      "step": 2723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0023305416107178,
+      "learning_rate": 1.7510964912280702e-05,
+      "loss": 1.4996,
+      "step": 2724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8939263820648193,
+      "learning_rate": 1.7510051169590646e-05,
+      "loss": 1.4191,
+      "step": 2725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6731317043304443,
+      "learning_rate": 1.7509137426900586e-05,
+      "loss": 1.4858,
+      "step": 2726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1144278049468994,
+      "learning_rate": 1.7508223684210527e-05,
+      "loss": 1.7315,
+      "step": 2727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.010509729385376,
+      "learning_rate": 1.750730994152047e-05,
+      "loss": 1.9436,
+      "step": 2728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.248021125793457,
+      "learning_rate": 1.750639619883041e-05,
+      "loss": 1.9086,
+      "step": 2729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8909590244293213,
+      "learning_rate": 1.7505482456140352e-05,
+      "loss": 2.0772,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.376441955566406,
+      "learning_rate": 1.7504568713450292e-05,
+      "loss": 2.007,
+      "step": 2731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.017995834350586,
+      "learning_rate": 1.7503654970760236e-05,
+      "loss": 1.594,
+      "step": 2732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8070907592773438,
+      "learning_rate": 1.7502741228070177e-05,
+      "loss": 1.6495,
+      "step": 2733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6280970573425293,
+      "learning_rate": 1.7501827485380117e-05,
+      "loss": 1.8887,
+      "step": 2734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.99180006980896,
+      "learning_rate": 1.750091374269006e-05,
+      "loss": 2.2291,
+      "step": 2735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7094061374664307,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.5402,
+      "step": 2736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2677829265594482,
+      "learning_rate": 1.7499086257309942e-05,
+      "loss": 2.0829,
+      "step": 2737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.605764389038086,
+      "learning_rate": 1.7498172514619883e-05,
+      "loss": 1.5815,
+      "step": 2738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.550248384475708,
+      "learning_rate": 1.7497258771929827e-05,
+      "loss": 1.5547,
+      "step": 2739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.568324327468872,
+      "learning_rate": 1.7496345029239767e-05,
+      "loss": 1.8413,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.197413206100464,
+      "learning_rate": 1.7495431286549708e-05,
+      "loss": 1.7282,
+      "step": 2741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.263859510421753,
+      "learning_rate": 1.7494517543859652e-05,
+      "loss": 1.8712,
+      "step": 2742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7096967697143555,
+      "learning_rate": 1.7493603801169592e-05,
+      "loss": 2.1104,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7955825328826904,
+      "learning_rate": 1.7492690058479533e-05,
+      "loss": 1.6629,
+      "step": 2744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.119974374771118,
+      "learning_rate": 1.7491776315789473e-05,
+      "loss": 1.8536,
+      "step": 2745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.132538318634033,
+      "learning_rate": 1.7490862573099417e-05,
+      "loss": 1.6473,
+      "step": 2746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7820699214935303,
+      "learning_rate": 1.7489948830409358e-05,
+      "loss": 1.8874,
+      "step": 2747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3913910388946533,
+      "learning_rate": 1.74890350877193e-05,
+      "loss": 1.776,
+      "step": 2748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9759700298309326,
+      "learning_rate": 1.7488121345029242e-05,
+      "loss": 1.7374,
+      "step": 2749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.02977180480957,
+      "learning_rate": 1.7487207602339183e-05,
+      "loss": 2.4269,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0215566158294678,
+      "learning_rate": 1.7486293859649123e-05,
+      "loss": 1.7054,
+      "step": 2751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3705074787139893,
+      "learning_rate": 1.7485380116959064e-05,
+      "loss": 2.1128,
+      "step": 2752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7182304859161377,
+      "learning_rate": 1.7484466374269008e-05,
+      "loss": 1.5381,
+      "step": 2753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.951809883117676,
+      "learning_rate": 1.7483552631578948e-05,
+      "loss": 1.9467,
+      "step": 2754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.710148572921753,
+      "learning_rate": 1.748263888888889e-05,
+      "loss": 1.3786,
+      "step": 2755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2515759468078613,
+      "learning_rate": 1.7481725146198833e-05,
+      "loss": 2.0663,
+      "step": 2756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.358302354812622,
+      "learning_rate": 1.7480811403508773e-05,
+      "loss": 2.1064,
+      "step": 2757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.214130163192749,
+      "learning_rate": 1.7479897660818714e-05,
+      "loss": 1.6787,
+      "step": 2758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.001364231109619,
+      "learning_rate": 1.7478983918128654e-05,
+      "loss": 1.3455,
+      "step": 2759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6205272674560547,
+      "learning_rate": 1.7478070175438598e-05,
+      "loss": 1.453,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.191128730773926,
+      "learning_rate": 1.747715643274854e-05,
+      "loss": 1.6347,
+      "step": 2761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6571638584136963,
+      "learning_rate": 1.747624269005848e-05,
+      "loss": 1.4239,
+      "step": 2762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5636301040649414,
+      "learning_rate": 1.7475328947368423e-05,
+      "loss": 1.8484,
+      "step": 2763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.47452712059021,
+      "learning_rate": 1.7474415204678364e-05,
+      "loss": 2.1081,
+      "step": 2764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.556861639022827,
+      "learning_rate": 1.7473501461988304e-05,
+      "loss": 1.36,
+      "step": 2765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5848517417907715,
+      "learning_rate": 1.7472587719298245e-05,
+      "loss": 1.5622,
+      "step": 2766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.612360715866089,
+      "learning_rate": 1.747167397660819e-05,
+      "loss": 2.0078,
+      "step": 2767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9010181427001953,
+      "learning_rate": 1.747076023391813e-05,
+      "loss": 1.9319,
+      "step": 2768
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1908342838287354,
+      "learning_rate": 1.746984649122807e-05,
+      "loss": 1.8056,
+      "step": 2769
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.290436267852783,
+      "learning_rate": 1.7468932748538014e-05,
+      "loss": 2.1628,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.21773624420166,
+      "learning_rate": 1.7468019005847954e-05,
+      "loss": 1.7953,
+      "step": 2771
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6943414211273193,
+      "learning_rate": 1.7467105263157895e-05,
+      "loss": 1.9273,
+      "step": 2772
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.171126365661621,
+      "learning_rate": 1.7466191520467835e-05,
+      "loss": 2.1947,
+      "step": 2773
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6477692127227783,
+      "learning_rate": 1.746527777777778e-05,
+      "loss": 1.7722,
+      "step": 2774
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0468697547912598,
+      "learning_rate": 1.746436403508772e-05,
+      "loss": 1.5863,
+      "step": 2775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1053287982940674,
+      "learning_rate": 1.746345029239766e-05,
+      "loss": 1.8904,
+      "step": 2776
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1636345386505127,
+      "learning_rate": 1.7462536549707604e-05,
+      "loss": 1.9538,
+      "step": 2777
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9465270042419434,
+      "learning_rate": 1.7461622807017545e-05,
+      "loss": 1.6136,
+      "step": 2778
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.100740432739258,
+      "learning_rate": 1.746070906432749e-05,
+      "loss": 1.4458,
+      "step": 2779
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3358640670776367,
+      "learning_rate": 1.745979532163743e-05,
+      "loss": 1.8355,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.980219602584839,
+      "learning_rate": 1.745888157894737e-05,
+      "loss": 1.7871,
+      "step": 2781
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2573482990264893,
+      "learning_rate": 1.7457967836257314e-05,
+      "loss": 1.6648,
+      "step": 2782
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.324049234390259,
+      "learning_rate": 1.7457054093567254e-05,
+      "loss": 1.8302,
+      "step": 2783
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.443674087524414,
+      "learning_rate": 1.7456140350877195e-05,
+      "loss": 1.9249,
+      "step": 2784
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.857825756072998,
+      "learning_rate": 1.745522660818714e-05,
+      "loss": 2.141,
+      "step": 2785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.690337657928467,
+      "learning_rate": 1.745431286549708e-05,
+      "loss": 2.2009,
+      "step": 2786
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2767059803009033,
+      "learning_rate": 1.745339912280702e-05,
+      "loss": 1.995,
+      "step": 2787
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.188621759414673,
+      "learning_rate": 1.745248538011696e-05,
+      "loss": 1.6233,
+      "step": 2788
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9077799320220947,
+      "learning_rate": 1.7451571637426904e-05,
+      "loss": 1.911,
+      "step": 2789
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.080432891845703,
+      "learning_rate": 1.7450657894736845e-05,
+      "loss": 1.8103,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0953733921051025,
+      "learning_rate": 1.7449744152046785e-05,
+      "loss": 1.9417,
+      "step": 2791
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.637502908706665,
+      "learning_rate": 1.744883040935673e-05,
+      "loss": 1.7268,
+      "step": 2792
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8901262283325195,
+      "learning_rate": 1.744791666666667e-05,
+      "loss": 1.4445,
+      "step": 2793
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4436862468719482,
+      "learning_rate": 1.744700292397661e-05,
+      "loss": 2.1962,
+      "step": 2794
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3226022720336914,
+      "learning_rate": 1.744608918128655e-05,
+      "loss": 1.9896,
+      "step": 2795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6932308673858643,
+      "learning_rate": 1.7445175438596494e-05,
+      "loss": 1.1486,
+      "step": 2796
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.908660888671875,
+      "learning_rate": 1.7444261695906435e-05,
+      "loss": 1.5593,
+      "step": 2797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3897335529327393,
+      "learning_rate": 1.7443347953216376e-05,
+      "loss": 1.6951,
+      "step": 2798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3417017459869385,
+      "learning_rate": 1.744243421052632e-05,
+      "loss": 2.1967,
+      "step": 2799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0153074264526367,
+      "learning_rate": 1.744152046783626e-05,
+      "loss": 1.574,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.790703535079956,
+      "learning_rate": 1.74406067251462e-05,
+      "loss": 1.6896,
+      "step": 2801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.362935543060303,
+      "learning_rate": 1.743969298245614e-05,
+      "loss": 1.6314,
+      "step": 2802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.895829677581787,
+      "learning_rate": 1.7438779239766085e-05,
+      "loss": 1.9368,
+      "step": 2803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.163577079772949,
+      "learning_rate": 1.7437865497076025e-05,
+      "loss": 2.5606,
+      "step": 2804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7188456058502197,
+      "learning_rate": 1.7436951754385966e-05,
+      "loss": 1.542,
+      "step": 2805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3406171798706055,
+      "learning_rate": 1.743603801169591e-05,
+      "loss": 2.301,
+      "step": 2806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.08978009223938,
+      "learning_rate": 1.743512426900585e-05,
+      "loss": 2.0466,
+      "step": 2807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.102182626724243,
+      "learning_rate": 1.743421052631579e-05,
+      "loss": 1.4877,
+      "step": 2808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1672918796539307,
+      "learning_rate": 1.743329678362573e-05,
+      "loss": 1.593,
+      "step": 2809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.173335075378418,
+      "learning_rate": 1.7432383040935675e-05,
+      "loss": 1.6739,
+      "step": 2810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7511203289031982,
+      "learning_rate": 1.7431469298245616e-05,
+      "loss": 1.9308,
+      "step": 2811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.784459352493286,
+      "learning_rate": 1.7430555555555556e-05,
+      "loss": 1.6733,
+      "step": 2812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9201533794403076,
+      "learning_rate": 1.74296418128655e-05,
+      "loss": 1.6503,
+      "step": 2813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.329380512237549,
+      "learning_rate": 1.742872807017544e-05,
+      "loss": 2.0436,
+      "step": 2814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9868924617767334,
+      "learning_rate": 1.742781432748538e-05,
+      "loss": 1.8103,
+      "step": 2815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.121378183364868,
+      "learning_rate": 1.7426900584795322e-05,
+      "loss": 1.7562,
+      "step": 2816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.36906099319458,
+      "learning_rate": 1.7425986842105266e-05,
+      "loss": 1.9286,
+      "step": 2817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.955063819885254,
+      "learning_rate": 1.7425073099415206e-05,
+      "loss": 1.8912,
+      "step": 2818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.482230186462402,
+      "learning_rate": 1.7424159356725147e-05,
+      "loss": 2.195,
+      "step": 2819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4036614894866943,
+      "learning_rate": 1.742324561403509e-05,
+      "loss": 2.1406,
+      "step": 2820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3576483726501465,
+      "learning_rate": 1.742233187134503e-05,
+      "loss": 1.9354,
+      "step": 2821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.666940927505493,
+      "learning_rate": 1.7421418128654972e-05,
+      "loss": 1.9495,
+      "step": 2822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1529271602630615,
+      "learning_rate": 1.7420504385964912e-05,
+      "loss": 1.6787,
+      "step": 2823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1130263805389404,
+      "learning_rate": 1.7419590643274856e-05,
+      "loss": 1.7553,
+      "step": 2824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.01096510887146,
+      "learning_rate": 1.7418676900584797e-05,
+      "loss": 1.7038,
+      "step": 2825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7412686347961426,
+      "learning_rate": 1.7417763157894737e-05,
+      "loss": 2.1289,
+      "step": 2826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.493710517883301,
+      "learning_rate": 1.741684941520468e-05,
+      "loss": 2.4868,
+      "step": 2827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1998536586761475,
+      "learning_rate": 1.7415935672514622e-05,
+      "loss": 1.8634,
+      "step": 2828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.046403169631958,
+      "learning_rate": 1.7415021929824562e-05,
+      "loss": 1.7647,
+      "step": 2829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.846731662750244,
+      "learning_rate": 1.7414108187134503e-05,
+      "loss": 1.5411,
+      "step": 2830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.347651481628418,
+      "learning_rate": 1.7413194444444447e-05,
+      "loss": 2.1722,
+      "step": 2831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.522559404373169,
+      "learning_rate": 1.7412280701754387e-05,
+      "loss": 2.1324,
+      "step": 2832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3962485790252686,
+      "learning_rate": 1.7411366959064328e-05,
+      "loss": 1.9789,
+      "step": 2833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.445873975753784,
+      "learning_rate": 1.7410453216374272e-05,
+      "loss": 1.7255,
+      "step": 2834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.125566005706787,
+      "learning_rate": 1.7409539473684212e-05,
+      "loss": 1.7852,
+      "step": 2835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7061092853546143,
+      "learning_rate": 1.7408625730994153e-05,
+      "loss": 1.5813,
+      "step": 2836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3055241107940674,
+      "learning_rate": 1.7407711988304093e-05,
+      "loss": 1.6796,
+      "step": 2837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.041292905807495,
+      "learning_rate": 1.7406798245614037e-05,
+      "loss": 1.8058,
+      "step": 2838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9388439655303955,
+      "learning_rate": 1.7405884502923978e-05,
+      "loss": 1.5842,
+      "step": 2839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.029515027999878,
+      "learning_rate": 1.740497076023392e-05,
+      "loss": 1.6431,
+      "step": 2840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3714113235473633,
+      "learning_rate": 1.740405701754386e-05,
+      "loss": 1.8172,
+      "step": 2841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.076776504516602,
+      "learning_rate": 1.7403143274853803e-05,
+      "loss": 2.0538,
+      "step": 2842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9452946186065674,
+      "learning_rate": 1.7402229532163743e-05,
+      "loss": 1.824,
+      "step": 2843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1727561950683594,
+      "learning_rate": 1.7401315789473684e-05,
+      "loss": 1.6009,
+      "step": 2844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0262162685394287,
+      "learning_rate": 1.7400402046783628e-05,
+      "loss": 1.9532,
+      "step": 2845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0070836544036865,
+      "learning_rate": 1.7399488304093568e-05,
+      "loss": 1.4837,
+      "step": 2846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2351834774017334,
+      "learning_rate": 1.739857456140351e-05,
+      "loss": 1.4769,
+      "step": 2847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.173797369003296,
+      "learning_rate": 1.739766081871345e-05,
+      "loss": 1.8263,
+      "step": 2848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9470627307891846,
+      "learning_rate": 1.7396747076023393e-05,
+      "loss": 1.5313,
+      "step": 2849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.728346824645996,
+      "learning_rate": 1.7395833333333334e-05,
+      "loss": 1.7169,
+      "step": 2850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.380157470703125,
+      "learning_rate": 1.7394919590643274e-05,
+      "loss": 1.9248,
+      "step": 2851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3627207279205322,
+      "learning_rate": 1.7394005847953218e-05,
+      "loss": 2.1016,
+      "step": 2852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2189319133758545,
+      "learning_rate": 1.739309210526316e-05,
+      "loss": 1.8139,
+      "step": 2853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.282947540283203,
+      "learning_rate": 1.73921783625731e-05,
+      "loss": 1.8619,
+      "step": 2854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8233234882354736,
+      "learning_rate": 1.739126461988304e-05,
+      "loss": 1.6698,
+      "step": 2855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8217079639434814,
+      "learning_rate": 1.7390350877192984e-05,
+      "loss": 1.4001,
+      "step": 2856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1536784172058105,
+      "learning_rate": 1.7389437134502924e-05,
+      "loss": 1.8702,
+      "step": 2857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8851125240325928,
+      "learning_rate": 1.7388523391812865e-05,
+      "loss": 1.5687,
+      "step": 2858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9868459701538086,
+      "learning_rate": 1.738760964912281e-05,
+      "loss": 1.7209,
+      "step": 2859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6645090579986572,
+      "learning_rate": 1.738669590643275e-05,
+      "loss": 1.9252,
+      "step": 2860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9980554580688477,
+      "learning_rate": 1.738578216374269e-05,
+      "loss": 1.7639,
+      "step": 2861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4673423767089844,
+      "learning_rate": 1.738486842105263e-05,
+      "loss": 2.094,
+      "step": 2862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8213558197021484,
+      "learning_rate": 1.7383954678362574e-05,
+      "loss": 2.2457,
+      "step": 2863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.932018995285034,
+      "learning_rate": 1.7383040935672515e-05,
+      "loss": 1.7473,
+      "step": 2864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5067689418792725,
+      "learning_rate": 1.738212719298246e-05,
+      "loss": 1.8407,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0696465969085693,
+      "learning_rate": 1.73812134502924e-05,
+      "loss": 1.8451,
+      "step": 2866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3307621479034424,
+      "learning_rate": 1.7380299707602343e-05,
+      "loss": 1.7982,
+      "step": 2867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6134681701660156,
+      "learning_rate": 1.7379385964912284e-05,
+      "loss": 2.145,
+      "step": 2868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1367788314819336,
+      "learning_rate": 1.7378472222222224e-05,
+      "loss": 1.911,
+      "step": 2869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.044448137283325,
+      "learning_rate": 1.7377558479532165e-05,
+      "loss": 1.8581,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.842559337615967,
+      "learning_rate": 1.737664473684211e-05,
+      "loss": 1.6646,
+      "step": 2871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8505301475524902,
+      "learning_rate": 1.737573099415205e-05,
+      "loss": 1.7427,
+      "step": 2872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.798403024673462,
+      "learning_rate": 1.737481725146199e-05,
+      "loss": 2.2283,
+      "step": 2873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.028724670410156,
+      "learning_rate": 1.7373903508771934e-05,
+      "loss": 2.3684,
+      "step": 2874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.512460947036743,
+      "learning_rate": 1.7372989766081874e-05,
+      "loss": 2.2119,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8201043605804443,
+      "learning_rate": 1.7372076023391815e-05,
+      "loss": 1.7848,
+      "step": 2876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.330790042877197,
+      "learning_rate": 1.7371162280701755e-05,
+      "loss": 1.898,
+      "step": 2877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.88324236869812,
+      "learning_rate": 1.73702485380117e-05,
+      "loss": 1.9344,
+      "step": 2878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.701653242111206,
+      "learning_rate": 1.736933479532164e-05,
+      "loss": 1.8376,
+      "step": 2879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.863701343536377,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 1.4845,
+      "step": 2880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.296753168106079,
+      "learning_rate": 1.7367507309941524e-05,
+      "loss": 1.908,
+      "step": 2881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.938777208328247,
+      "learning_rate": 1.7366593567251465e-05,
+      "loss": 1.8337,
+      "step": 2882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2822816371917725,
+      "learning_rate": 1.7365679824561405e-05,
+      "loss": 1.3346,
+      "step": 2883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9580297470092773,
+      "learning_rate": 1.7364766081871346e-05,
+      "loss": 1.798,
+      "step": 2884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.944118022918701,
+      "learning_rate": 1.736385233918129e-05,
+      "loss": 1.7343,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5400948524475098,
+      "learning_rate": 1.736293859649123e-05,
+      "loss": 2.3175,
+      "step": 2886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6262757778167725,
+      "learning_rate": 1.736202485380117e-05,
+      "loss": 1.5665,
+      "step": 2887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.440164566040039,
+      "learning_rate": 1.7361111111111114e-05,
+      "loss": 1.9663,
+      "step": 2888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5585274696350098,
+      "learning_rate": 1.7360197368421055e-05,
+      "loss": 1.9092,
+      "step": 2889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9374806880950928,
+      "learning_rate": 1.7359283625730996e-05,
+      "loss": 1.8003,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.893019199371338,
+      "learning_rate": 1.7358369883040936e-05,
+      "loss": 1.5476,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0867326259613037,
+      "learning_rate": 1.735745614035088e-05,
+      "loss": 1.5459,
+      "step": 2892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4926297664642334,
+      "learning_rate": 1.735654239766082e-05,
+      "loss": 1.5525,
+      "step": 2893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6851389408111572,
+      "learning_rate": 1.735562865497076e-05,
+      "loss": 1.3371,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.107189178466797,
+      "learning_rate": 1.7354714912280705e-05,
+      "loss": 2.0105,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.242105484008789,
+      "learning_rate": 1.7353801169590645e-05,
+      "loss": 2.0527,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.275820016860962,
+      "learning_rate": 1.7352887426900586e-05,
+      "loss": 1.7886,
+      "step": 2897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4826247692108154,
+      "learning_rate": 1.7351973684210527e-05,
+      "loss": 1.9329,
+      "step": 2898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.807482957839966,
+      "learning_rate": 1.735105994152047e-05,
+      "loss": 1.3632,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3509182929992676,
+      "learning_rate": 1.735014619883041e-05,
+      "loss": 1.991,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0328915119171143,
+      "learning_rate": 1.734923245614035e-05,
+      "loss": 1.6942,
+      "step": 2901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.816396713256836,
+      "learning_rate": 1.7348318713450295e-05,
+      "loss": 1.6543,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.584907054901123,
+      "learning_rate": 1.7347404970760236e-05,
+      "loss": 1.927,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.606428861618042,
+      "learning_rate": 1.7346491228070176e-05,
+      "loss": 1.6188,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.153346300125122,
+      "learning_rate": 1.7345577485380117e-05,
+      "loss": 1.7231,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8031387329101562,
+      "learning_rate": 1.734466374269006e-05,
+      "loss": 1.7868,
+      "step": 2906
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9999256134033203,
+      "learning_rate": 1.734375e-05,
+      "loss": 1.9495,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.4340434074401855,
+      "learning_rate": 1.7342836257309942e-05,
+      "loss": 1.6789,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8429224491119385,
+      "learning_rate": 1.7341922514619886e-05,
+      "loss": 1.5466,
+      "step": 2909
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1824963092803955,
+      "learning_rate": 1.7341008771929826e-05,
+      "loss": 1.9311,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8060317039489746,
+      "learning_rate": 1.7340095029239767e-05,
+      "loss": 1.5071,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3776657581329346,
+      "learning_rate": 1.7339181286549707e-05,
+      "loss": 1.5446,
+      "step": 2912
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.623973846435547,
+      "learning_rate": 1.733826754385965e-05,
+      "loss": 1.4068,
+      "step": 2913
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1358938217163086,
+      "learning_rate": 1.7337353801169592e-05,
+      "loss": 1.8368,
+      "step": 2914
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7046380043029785,
+      "learning_rate": 1.7336440058479532e-05,
+      "loss": 2.264,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.125504970550537,
+      "learning_rate": 1.7335526315789476e-05,
+      "loss": 1.5999,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1534974575042725,
+      "learning_rate": 1.7334612573099417e-05,
+      "loss": 1.8758,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0782456398010254,
+      "learning_rate": 1.7333698830409357e-05,
+      "loss": 1.8989,
+      "step": 2918
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.979672908782959,
+      "learning_rate": 1.7332785087719298e-05,
+      "loss": 1.8565,
+      "step": 2919
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.34360933303833,
+      "learning_rate": 1.7331871345029242e-05,
+      "loss": 1.9453,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9948699474334717,
+      "learning_rate": 1.7330957602339182e-05,
+      "loss": 1.785,
+      "step": 2921
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.89884090423584,
+      "learning_rate": 1.7330043859649123e-05,
+      "loss": 1.6014,
+      "step": 2922
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2928104400634766,
+      "learning_rate": 1.7329130116959067e-05,
+      "loss": 1.7234,
+      "step": 2923
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.211982488632202,
+      "learning_rate": 1.7328216374269007e-05,
+      "loss": 1.7614,
+      "step": 2924
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.527576446533203,
+      "learning_rate": 1.7327302631578948e-05,
+      "loss": 1.8626,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.577430009841919,
+      "learning_rate": 1.732638888888889e-05,
+      "loss": 1.379,
+      "step": 2926
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.221498727798462,
+      "learning_rate": 1.7325475146198832e-05,
+      "loss": 2.0088,
+      "step": 2927
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3508200645446777,
+      "learning_rate": 1.7324561403508773e-05,
+      "loss": 1.9314,
+      "step": 2928
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.658888101577759,
+      "learning_rate": 1.7323647660818713e-05,
+      "loss": 1.9815,
+      "step": 2929
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9310526847839355,
+      "learning_rate": 1.7322733918128657e-05,
+      "loss": 1.4592,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.530092477798462,
+      "learning_rate": 1.7321820175438598e-05,
+      "loss": 1.8353,
+      "step": 2931
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1064515113830566,
+      "learning_rate": 1.732090643274854e-05,
+      "loss": 1.6629,
+      "step": 2932
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5365617275238037,
+      "learning_rate": 1.731999269005848e-05,
+      "loss": 1.9164,
+      "step": 2933
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.481884717941284,
+      "learning_rate": 1.7319078947368423e-05,
+      "loss": 2.1705,
+      "step": 2934
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.631134271621704,
+      "learning_rate": 1.7318165204678363e-05,
+      "loss": 1.6055,
+      "step": 2935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7737274169921875,
+      "learning_rate": 1.7317251461988304e-05,
+      "loss": 1.4547,
+      "step": 2936
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.175727605819702,
+      "learning_rate": 1.7316337719298248e-05,
+      "loss": 1.5419,
+      "step": 2937
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3231282234191895,
+      "learning_rate": 1.7315423976608188e-05,
+      "loss": 1.7989,
+      "step": 2938
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0127830505371094,
+      "learning_rate": 1.731451023391813e-05,
+      "loss": 1.4194,
+      "step": 2939
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2085673809051514,
+      "learning_rate": 1.731359649122807e-05,
+      "loss": 1.8586,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1913468837738037,
+      "learning_rate": 1.7312682748538013e-05,
+      "loss": 1.6763,
+      "step": 2941
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9786107540130615,
+      "learning_rate": 1.7311769005847954e-05,
+      "loss": 1.628,
+      "step": 2942
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.062408924102783,
+      "learning_rate": 1.7310855263157894e-05,
+      "loss": 1.5986,
+      "step": 2943
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.813668727874756,
+      "learning_rate": 1.7309941520467838e-05,
+      "loss": 2.2766,
+      "step": 2944
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4176695346832275,
+      "learning_rate": 1.730902777777778e-05,
+      "loss": 1.9921,
+      "step": 2945
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.867833137512207,
+      "learning_rate": 1.730811403508772e-05,
+      "loss": 2.0243,
+      "step": 2946
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2775073051452637,
+      "learning_rate": 1.730720029239766e-05,
+      "loss": 1.7362,
+      "step": 2947
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0966804027557373,
+      "learning_rate": 1.7306286549707604e-05,
+      "loss": 1.7459,
+      "step": 2948
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7592406272888184,
+      "learning_rate": 1.7305372807017544e-05,
+      "loss": 1.7341,
+      "step": 2949
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.861111879348755,
+      "learning_rate": 1.7304459064327488e-05,
+      "loss": 1.5795,
+      "step": 2950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.951542854309082,
+      "learning_rate": 1.730354532163743e-05,
+      "loss": 1.6623,
+      "step": 2951
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2604775428771973,
+      "learning_rate": 1.730263157894737e-05,
+      "loss": 1.6944,
+      "step": 2952
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5927605628967285,
+      "learning_rate": 1.7301717836257313e-05,
+      "loss": 1.7564,
+      "step": 2953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.65981125831604,
+      "learning_rate": 1.7300804093567254e-05,
+      "loss": 1.3315,
+      "step": 2954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.065046787261963,
+      "learning_rate": 1.7299890350877194e-05,
+      "loss": 1.5505,
+      "step": 2955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.831800937652588,
+      "learning_rate": 1.7298976608187138e-05,
+      "loss": 2.1079,
+      "step": 2956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3002054691314697,
+      "learning_rate": 1.729806286549708e-05,
+      "loss": 1.7069,
+      "step": 2957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9924685955047607,
+      "learning_rate": 1.729714912280702e-05,
+      "loss": 1.5726,
+      "step": 2958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.71529483795166,
+      "learning_rate": 1.729623538011696e-05,
+      "loss": 1.9081,
+      "step": 2959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4383704662323,
+      "learning_rate": 1.7295321637426904e-05,
+      "loss": 1.9162,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.45229434967041,
+      "learning_rate": 1.7294407894736844e-05,
+      "loss": 1.5392,
+      "step": 2961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.098747730255127,
+      "learning_rate": 1.7293494152046785e-05,
+      "loss": 1.7804,
+      "step": 2962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2166731357574463,
+      "learning_rate": 1.729258040935673e-05,
+      "loss": 1.9528,
+      "step": 2963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.689908981323242,
+      "learning_rate": 1.729166666666667e-05,
+      "loss": 1.5128,
+      "step": 2964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.987074613571167,
+      "learning_rate": 1.729075292397661e-05,
+      "loss": 1.4543,
+      "step": 2965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5326688289642334,
+      "learning_rate": 1.728983918128655e-05,
+      "loss": 1.5155,
+      "step": 2966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4401345252990723,
+      "learning_rate": 1.7288925438596494e-05,
+      "loss": 1.9803,
+      "step": 2967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8100712299346924,
+      "learning_rate": 1.7288011695906435e-05,
+      "loss": 1.848,
+      "step": 2968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.888179302215576,
+      "learning_rate": 1.7287097953216375e-05,
+      "loss": 1.5358,
+      "step": 2969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7369508743286133,
+      "learning_rate": 1.728618421052632e-05,
+      "loss": 1.4375,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8645193576812744,
+      "learning_rate": 1.728527046783626e-05,
+      "loss": 1.5372,
+      "step": 2971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.871143102645874,
+      "learning_rate": 1.72843567251462e-05,
+      "loss": 1.5963,
+      "step": 2972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2448947429656982,
+      "learning_rate": 1.728344298245614e-05,
+      "loss": 1.6632,
+      "step": 2973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.922898292541504,
+      "learning_rate": 1.7282529239766085e-05,
+      "loss": 2.6456,
+      "step": 2974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.605226993560791,
+      "learning_rate": 1.7281615497076025e-05,
+      "loss": 1.3884,
+      "step": 2975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6723434925079346,
+      "learning_rate": 1.7280701754385966e-05,
+      "loss": 1.9702,
+      "step": 2976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.997697353363037,
+      "learning_rate": 1.727978801169591e-05,
+      "loss": 1.5881,
+      "step": 2977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0216195583343506,
+      "learning_rate": 1.727887426900585e-05,
+      "loss": 1.6563,
+      "step": 2978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4902303218841553,
+      "learning_rate": 1.727796052631579e-05,
+      "loss": 1.4466,
+      "step": 2979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.05428409576416,
+      "learning_rate": 1.727704678362573e-05,
+      "loss": 1.3425,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9157841205596924,
+      "learning_rate": 1.7276133040935675e-05,
+      "loss": 1.6368,
+      "step": 2981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9910025596618652,
+      "learning_rate": 1.7275219298245616e-05,
+      "loss": 1.5584,
+      "step": 2982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.605787992477417,
+      "learning_rate": 1.7274305555555556e-05,
+      "loss": 2.1488,
+      "step": 2983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0771830081939697,
+      "learning_rate": 1.72733918128655e-05,
+      "loss": 1.5314,
+      "step": 2984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.800715446472168,
+      "learning_rate": 1.727247807017544e-05,
+      "loss": 1.5924,
+      "step": 2985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1634697914123535,
+      "learning_rate": 1.727156432748538e-05,
+      "loss": 1.5991,
+      "step": 2986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.19392728805542,
+      "learning_rate": 1.727065058479532e-05,
+      "loss": 1.8118,
+      "step": 2987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9422972202301025,
+      "learning_rate": 1.7269736842105265e-05,
+      "loss": 1.912,
+      "step": 2988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5553619861602783,
+      "learning_rate": 1.7268823099415206e-05,
+      "loss": 1.5742,
+      "step": 2989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.79508376121521,
+      "learning_rate": 1.7267909356725147e-05,
+      "loss": 1.6423,
+      "step": 2990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4650774002075195,
+      "learning_rate": 1.726699561403509e-05,
+      "loss": 1.7822,
+      "step": 2991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0107486248016357,
+      "learning_rate": 1.726608187134503e-05,
+      "loss": 1.5967,
+      "step": 2992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.580292224884033,
+      "learning_rate": 1.726516812865497e-05,
+      "loss": 2.079,
+      "step": 2993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 29.01944923400879,
+      "learning_rate": 1.7264254385964912e-05,
+      "loss": 1.4916,
+      "step": 2994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.070211172103882,
+      "learning_rate": 1.7263340643274856e-05,
+      "loss": 1.706,
+      "step": 2995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.875509023666382,
+      "learning_rate": 1.7262426900584796e-05,
+      "loss": 1.5983,
+      "step": 2996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0384087562561035,
+      "learning_rate": 1.7261513157894737e-05,
+      "loss": 2.1234,
+      "step": 2997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.091494083404541,
+      "learning_rate": 1.726059941520468e-05,
+      "loss": 1.6455,
+      "step": 2998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.437821626663208,
+      "learning_rate": 1.725968567251462e-05,
+      "loss": 2.009,
+      "step": 2999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.323944568634033,
+      "learning_rate": 1.7258771929824562e-05,
+      "loss": 1.8353,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0925636291503906,
+      "learning_rate": 1.7257858187134502e-05,
+      "loss": 1.6552,
+      "step": 3001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.054040431976318,
+      "learning_rate": 1.7256944444444446e-05,
+      "loss": 2.001,
+      "step": 3002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.295699119567871,
+      "learning_rate": 1.7256030701754387e-05,
+      "loss": 1.6751,
+      "step": 3003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6656689643859863,
+      "learning_rate": 1.7255116959064327e-05,
+      "loss": 1.2666,
+      "step": 3004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.073344707489014,
+      "learning_rate": 1.725420321637427e-05,
+      "loss": 2.0475,
+      "step": 3005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2931430339813232,
+      "learning_rate": 1.7253289473684212e-05,
+      "loss": 1.8639,
+      "step": 3006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.381152629852295,
+      "learning_rate": 1.7252375730994152e-05,
+      "loss": 1.3896,
+      "step": 3007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9905331134796143,
+      "learning_rate": 1.7251461988304093e-05,
+      "loss": 2.1331,
+      "step": 3008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5089924335479736,
+      "learning_rate": 1.7250548245614037e-05,
+      "loss": 1.6766,
+      "step": 3009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5644569396972656,
+      "learning_rate": 1.7249634502923977e-05,
+      "loss": 1.9093,
+      "step": 3010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.093775987625122,
+      "learning_rate": 1.7248720760233918e-05,
+      "loss": 1.6454,
+      "step": 3011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4026453495025635,
+      "learning_rate": 1.7247807017543862e-05,
+      "loss": 1.8956,
+      "step": 3012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.120699882507324,
+      "learning_rate": 1.7246893274853802e-05,
+      "loss": 1.9183,
+      "step": 3013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.708296537399292,
+      "learning_rate": 1.7245979532163743e-05,
+      "loss": 1.9745,
+      "step": 3014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.391289234161377,
+      "learning_rate": 1.7245065789473683e-05,
+      "loss": 1.6778,
+      "step": 3015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3008859157562256,
+      "learning_rate": 1.7244152046783627e-05,
+      "loss": 2.0719,
+      "step": 3016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5077061653137207,
+      "learning_rate": 1.7243238304093568e-05,
+      "loss": 1.8456,
+      "step": 3017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1940500736236572,
+      "learning_rate": 1.724232456140351e-05,
+      "loss": 1.6435,
+      "step": 3018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.062624931335449,
+      "learning_rate": 1.7241410818713452e-05,
+      "loss": 1.7842,
+      "step": 3019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.782367467880249,
+      "learning_rate": 1.7240497076023393e-05,
+      "loss": 1.4717,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1251518726348877,
+      "learning_rate": 1.7239583333333333e-05,
+      "loss": 1.7227,
+      "step": 3021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.104140520095825,
+      "learning_rate": 1.7238669590643274e-05,
+      "loss": 1.6627,
+      "step": 3022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.966486930847168,
+      "learning_rate": 1.7237755847953218e-05,
+      "loss": 1.4725,
+      "step": 3023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3157105445861816,
+      "learning_rate": 1.723684210526316e-05,
+      "loss": 1.6921,
+      "step": 3024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7875006198883057,
+      "learning_rate": 1.72359283625731e-05,
+      "loss": 1.6044,
+      "step": 3025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.794218063354492,
+      "learning_rate": 1.7235014619883043e-05,
+      "loss": 2.1793,
+      "step": 3026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.275984764099121,
+      "learning_rate": 1.7234100877192983e-05,
+      "loss": 1.2235,
+      "step": 3027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.575380325317383,
+      "learning_rate": 1.7233187134502924e-05,
+      "loss": 1.4864,
+      "step": 3028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.433724880218506,
+      "learning_rate": 1.7232273391812864e-05,
+      "loss": 1.7036,
+      "step": 3029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.221189498901367,
+      "learning_rate": 1.7231359649122808e-05,
+      "loss": 1.7187,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.057345390319824,
+      "learning_rate": 1.723044590643275e-05,
+      "loss": 1.7894,
+      "step": 3031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8345727920532227,
+      "learning_rate": 1.722953216374269e-05,
+      "loss": 1.53,
+      "step": 3032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.628539800643921,
+      "learning_rate": 1.7228618421052633e-05,
+      "loss": 2.0689,
+      "step": 3033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6318976879119873,
+      "learning_rate": 1.7227704678362574e-05,
+      "loss": 1.847,
+      "step": 3034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7870285511016846,
+      "learning_rate": 1.7226790935672514e-05,
+      "loss": 1.7946,
+      "step": 3035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5716323852539062,
+      "learning_rate": 1.7225877192982458e-05,
+      "loss": 1.4861,
+      "step": 3036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.363692283630371,
+      "learning_rate": 1.72249634502924e-05,
+      "loss": 2.0316,
+      "step": 3037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1008334159851074,
+      "learning_rate": 1.7224049707602343e-05,
+      "loss": 1.4358,
+      "step": 3038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5626158714294434,
+      "learning_rate": 1.7223135964912283e-05,
+      "loss": 1.7954,
+      "step": 3039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.932529926300049,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 1.696,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9051427841186523,
+      "learning_rate": 1.7221308479532168e-05,
+      "loss": 1.5839,
+      "step": 3041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.401752471923828,
+      "learning_rate": 1.7220394736842108e-05,
+      "loss": 1.5545,
+      "step": 3042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2030818462371826,
+      "learning_rate": 1.721948099415205e-05,
+      "loss": 1.7761,
+      "step": 3043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4806735515594482,
+      "learning_rate": 1.721856725146199e-05,
+      "loss": 1.3633,
+      "step": 3044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1992413997650146,
+      "learning_rate": 1.7217653508771933e-05,
+      "loss": 1.7826,
+      "step": 3045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.825577974319458,
+      "learning_rate": 1.7216739766081874e-05,
+      "loss": 1.7281,
+      "step": 3046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.734152317047119,
+      "learning_rate": 1.7215826023391814e-05,
+      "loss": 2.0704,
+      "step": 3047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8996634483337402,
+      "learning_rate": 1.7214912280701758e-05,
+      "loss": 1.6174,
+      "step": 3048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0671839714050293,
+      "learning_rate": 1.72139985380117e-05,
+      "loss": 1.8399,
+      "step": 3049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4097323417663574,
+      "learning_rate": 1.721308479532164e-05,
+      "loss": 1.688,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.927175283432007,
+      "learning_rate": 1.721217105263158e-05,
+      "loss": 1.7704,
+      "step": 3051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.837606906890869,
+      "learning_rate": 1.7211257309941524e-05,
+      "loss": 1.8248,
+      "step": 3052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6537256240844727,
+      "learning_rate": 1.7210343567251464e-05,
+      "loss": 2.0853,
+      "step": 3053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3592708110809326,
+      "learning_rate": 1.7209429824561405e-05,
+      "loss": 1.5173,
+      "step": 3054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9006526470184326,
+      "learning_rate": 1.720851608187135e-05,
+      "loss": 1.451,
+      "step": 3055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5423431396484375,
+      "learning_rate": 1.720760233918129e-05,
+      "loss": 2.2228,
+      "step": 3056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.606659412384033,
+      "learning_rate": 1.720668859649123e-05,
+      "loss": 2.117,
+      "step": 3057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.019798755645752,
+      "learning_rate": 1.720577485380117e-05,
+      "loss": 2.0709,
+      "step": 3058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5352935791015625,
+      "learning_rate": 1.7204861111111114e-05,
+      "loss": 1.6468,
+      "step": 3059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.749398946762085,
+      "learning_rate": 1.7203947368421055e-05,
+      "loss": 1.6358,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.709990501403809,
+      "learning_rate": 1.7203033625730995e-05,
+      "loss": 2.5792,
+      "step": 3061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.163615703582764,
+      "learning_rate": 1.720211988304094e-05,
+      "loss": 1.8892,
+      "step": 3062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.691413640975952,
+      "learning_rate": 1.720120614035088e-05,
+      "loss": 1.7706,
+      "step": 3063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.992666721343994,
+      "learning_rate": 1.720029239766082e-05,
+      "loss": 1.6704,
+      "step": 3064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.935342311859131,
+      "learning_rate": 1.719937865497076e-05,
+      "loss": 1.4401,
+      "step": 3065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6707348823547363,
+      "learning_rate": 1.7198464912280705e-05,
+      "loss": 2.143,
+      "step": 3066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9915313720703125,
+      "learning_rate": 1.7197551169590645e-05,
+      "loss": 1.9349,
+      "step": 3067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9275002479553223,
+      "learning_rate": 1.7196637426900586e-05,
+      "loss": 1.7014,
+      "step": 3068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0498101711273193,
+      "learning_rate": 1.719572368421053e-05,
+      "loss": 1.8226,
+      "step": 3069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.757810354232788,
+      "learning_rate": 1.719480994152047e-05,
+      "loss": 1.8372,
+      "step": 3070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.088850498199463,
+      "learning_rate": 1.719389619883041e-05,
+      "loss": 1.8141,
+      "step": 3071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7204439640045166,
+      "learning_rate": 1.719298245614035e-05,
+      "loss": 1.4877,
+      "step": 3072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7721750736236572,
+      "learning_rate": 1.7192068713450295e-05,
+      "loss": 1.4507,
+      "step": 3073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.276620626449585,
+      "learning_rate": 1.7191154970760236e-05,
+      "loss": 2.1214,
+      "step": 3074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7006750106811523,
+      "learning_rate": 1.7190241228070176e-05,
+      "loss": 1.7713,
+      "step": 3075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7481777667999268,
+      "learning_rate": 1.718932748538012e-05,
+      "loss": 1.7922,
+      "step": 3076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.958709478378296,
+      "learning_rate": 1.718841374269006e-05,
+      "loss": 1.7427,
+      "step": 3077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.042184829711914,
+      "learning_rate": 1.71875e-05,
+      "loss": 1.4872,
+      "step": 3078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.320561408996582,
+      "learning_rate": 1.718658625730994e-05,
+      "loss": 1.7922,
+      "step": 3079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1299097537994385,
+      "learning_rate": 1.7185672514619885e-05,
+      "loss": 1.8425,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.155426263809204,
+      "learning_rate": 1.7184758771929826e-05,
+      "loss": 1.8426,
+      "step": 3081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.550656795501709,
+      "learning_rate": 1.7183845029239767e-05,
+      "loss": 1.7153,
+      "step": 3082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.331472158432007,
+      "learning_rate": 1.718293128654971e-05,
+      "loss": 1.6718,
+      "step": 3083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.027604341506958,
+      "learning_rate": 1.718201754385965e-05,
+      "loss": 1.5747,
+      "step": 3084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.500678062438965,
+      "learning_rate": 1.718110380116959e-05,
+      "loss": 1.5507,
+      "step": 3085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8760857582092285,
+      "learning_rate": 1.7180190058479532e-05,
+      "loss": 1.8152,
+      "step": 3086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8533802032470703,
+      "learning_rate": 1.7179276315789476e-05,
+      "loss": 1.7006,
+      "step": 3087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9274139404296875,
+      "learning_rate": 1.7178362573099416e-05,
+      "loss": 1.6181,
+      "step": 3088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.085536479949951,
+      "learning_rate": 1.7177448830409357e-05,
+      "loss": 1.671,
+      "step": 3089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.536608934402466,
+      "learning_rate": 1.71765350877193e-05,
+      "loss": 1.9834,
+      "step": 3090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4977715015411377,
+      "learning_rate": 1.717562134502924e-05,
+      "loss": 2.109,
+      "step": 3091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8454713821411133,
+      "learning_rate": 1.7174707602339182e-05,
+      "loss": 1.8243,
+      "step": 3092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6513500213623047,
+      "learning_rate": 1.7173793859649122e-05,
+      "loss": 2.0889,
+      "step": 3093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.967486619949341,
+      "learning_rate": 1.7172880116959066e-05,
+      "loss": 1.3281,
+      "step": 3094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4039390087127686,
+      "learning_rate": 1.7171966374269007e-05,
+      "loss": 1.7869,
+      "step": 3095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1641387939453125,
+      "learning_rate": 1.7171052631578947e-05,
+      "loss": 1.3458,
+      "step": 3096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1309309005737305,
+      "learning_rate": 1.717013888888889e-05,
+      "loss": 1.6824,
+      "step": 3097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.96863055229187,
+      "learning_rate": 1.7169225146198832e-05,
+      "loss": 2.1044,
+      "step": 3098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.707432508468628,
+      "learning_rate": 1.7168311403508772e-05,
+      "loss": 1.9577,
+      "step": 3099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2345364093780518,
+      "learning_rate": 1.7167397660818713e-05,
+      "loss": 1.8725,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.201688528060913,
+      "learning_rate": 1.7166483918128657e-05,
+      "loss": 1.5108,
+      "step": 3101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6457419395446777,
+      "learning_rate": 1.7165570175438597e-05,
+      "loss": 1.6834,
+      "step": 3102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6936256885528564,
+      "learning_rate": 1.7164656432748538e-05,
+      "loss": 1.8916,
+      "step": 3103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9993717670440674,
+      "learning_rate": 1.7163742690058482e-05,
+      "loss": 1.8701,
+      "step": 3104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8587656021118164,
+      "learning_rate": 1.7162828947368422e-05,
+      "loss": 2.1631,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.308540105819702,
+      "learning_rate": 1.7161915204678363e-05,
+      "loss": 2.3346,
+      "step": 3106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4609744548797607,
+      "learning_rate": 1.7161001461988303e-05,
+      "loss": 1.78,
+      "step": 3107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.615201711654663,
+      "learning_rate": 1.7160087719298247e-05,
+      "loss": 1.6595,
+      "step": 3108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.393958806991577,
+      "learning_rate": 1.7159173976608188e-05,
+      "loss": 1.6182,
+      "step": 3109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.538522720336914,
+      "learning_rate": 1.715826023391813e-05,
+      "loss": 1.466,
+      "step": 3110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.951009750366211,
+      "learning_rate": 1.7157346491228072e-05,
+      "loss": 1.4224,
+      "step": 3111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.001002073287964,
+      "learning_rate": 1.7156432748538013e-05,
+      "loss": 1.9994,
+      "step": 3112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9809303283691406,
+      "learning_rate": 1.7155519005847953e-05,
+      "loss": 1.7508,
+      "step": 3113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.906219244003296,
+      "learning_rate": 1.7154605263157894e-05,
+      "loss": 1.766,
+      "step": 3114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.491847276687622,
+      "learning_rate": 1.7153691520467838e-05,
+      "loss": 2.1459,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.257493019104004,
+      "learning_rate": 1.715277777777778e-05,
+      "loss": 1.8268,
+      "step": 3116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7124674320220947,
+      "learning_rate": 1.715186403508772e-05,
+      "loss": 1.7332,
+      "step": 3117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.104041337966919,
+      "learning_rate": 1.7150950292397663e-05,
+      "loss": 1.8026,
+      "step": 3118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8085594177246094,
+      "learning_rate": 1.7150036549707603e-05,
+      "loss": 1.5752,
+      "step": 3119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4059882164001465,
+      "learning_rate": 1.7149122807017544e-05,
+      "loss": 1.7864,
+      "step": 3120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9923553466796875,
+      "learning_rate": 1.7148209064327488e-05,
+      "loss": 1.8418,
+      "step": 3121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.855747699737549,
+      "learning_rate": 1.7147295321637428e-05,
+      "loss": 1.4824,
+      "step": 3122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.247276544570923,
+      "learning_rate": 1.7146381578947372e-05,
+      "loss": 1.909,
+      "step": 3123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.329528570175171,
+      "learning_rate": 1.7145467836257313e-05,
+      "loss": 1.7206,
+      "step": 3124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3481743335723877,
+      "learning_rate": 1.7144554093567253e-05,
+      "loss": 1.9348,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8519582748413086,
+      "learning_rate": 1.7143640350877194e-05,
+      "loss": 2.2266,
+      "step": 3126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.835561752319336,
+      "learning_rate": 1.7142726608187138e-05,
+      "loss": 1.8379,
+      "step": 3127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1654186248779297,
+      "learning_rate": 1.7141812865497078e-05,
+      "loss": 1.9587,
+      "step": 3128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7079594135284424,
+      "learning_rate": 1.714089912280702e-05,
+      "loss": 1.2664,
+      "step": 3129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2197723388671875,
+      "learning_rate": 1.7139985380116963e-05,
+      "loss": 1.7241,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.105574607849121,
+      "learning_rate": 1.7139071637426903e-05,
+      "loss": 1.7154,
+      "step": 3131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0934157371521,
+      "learning_rate": 1.7138157894736844e-05,
+      "loss": 1.7502,
+      "step": 3132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.324979782104492,
+      "learning_rate": 1.7137244152046784e-05,
+      "loss": 2.0107,
+      "step": 3133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3372855186462402,
+      "learning_rate": 1.7136330409356728e-05,
+      "loss": 1.8557,
+      "step": 3134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.602694034576416,
+      "learning_rate": 1.713541666666667e-05,
+      "loss": 1.1475,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.259317398071289,
+      "learning_rate": 1.713450292397661e-05,
+      "loss": 1.3743,
+      "step": 3136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9142892360687256,
+      "learning_rate": 1.7133589181286553e-05,
+      "loss": 1.7982,
+      "step": 3137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9714927673339844,
+      "learning_rate": 1.7132675438596494e-05,
+      "loss": 1.8141,
+      "step": 3138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8921704292297363,
+      "learning_rate": 1.7131761695906434e-05,
+      "loss": 1.7242,
+      "step": 3139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7172563076019287,
+      "learning_rate": 1.7130847953216375e-05,
+      "loss": 2.0875,
+      "step": 3140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7829060554504395,
+      "learning_rate": 1.712993421052632e-05,
+      "loss": 1.5671,
+      "step": 3141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1941959857940674,
+      "learning_rate": 1.712902046783626e-05,
+      "loss": 2.1909,
+      "step": 3142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.056957244873047,
+      "learning_rate": 1.71281067251462e-05,
+      "loss": 1.6444,
+      "step": 3143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6125845909118652,
+      "learning_rate": 1.7127192982456144e-05,
+      "loss": 1.5177,
+      "step": 3144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.00150990486145,
+      "learning_rate": 1.7126279239766084e-05,
+      "loss": 1.6835,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.868450880050659,
+      "learning_rate": 1.7125365497076025e-05,
+      "loss": 1.6593,
+      "step": 3146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9612419605255127,
+      "learning_rate": 1.7124451754385965e-05,
+      "loss": 1.9306,
+      "step": 3147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1900501251220703,
+      "learning_rate": 1.712353801169591e-05,
+      "loss": 1.4467,
+      "step": 3148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.086914300918579,
+      "learning_rate": 1.712262426900585e-05,
+      "loss": 1.892,
+      "step": 3149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.808814287185669,
+      "learning_rate": 1.712171052631579e-05,
+      "loss": 2.378,
+      "step": 3150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0465481281280518,
+      "learning_rate": 1.7120796783625734e-05,
+      "loss": 1.6848,
+      "step": 3151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.061974048614502,
+      "learning_rate": 1.7119883040935675e-05,
+      "loss": 1.8909,
+      "step": 3152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2804574966430664,
+      "learning_rate": 1.7118969298245615e-05,
+      "loss": 2.0201,
+      "step": 3153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.965174674987793,
+      "learning_rate": 1.7118055555555556e-05,
+      "loss": 1.4757,
+      "step": 3154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.325791835784912,
+      "learning_rate": 1.71171418128655e-05,
+      "loss": 1.5208,
+      "step": 3155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2411623001098633,
+      "learning_rate": 1.711622807017544e-05,
+      "loss": 1.9271,
+      "step": 3156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.536417007446289,
+      "learning_rate": 1.711531432748538e-05,
+      "loss": 1.6871,
+      "step": 3157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2244985103607178,
+      "learning_rate": 1.7114400584795325e-05,
+      "loss": 1.7626,
+      "step": 3158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.213442325592041,
+      "learning_rate": 1.7113486842105265e-05,
+      "loss": 2.1598,
+      "step": 3159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.713350772857666,
+      "learning_rate": 1.7112573099415206e-05,
+      "loss": 1.605,
+      "step": 3160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6849887371063232,
+      "learning_rate": 1.7111659356725146e-05,
+      "loss": 1.5017,
+      "step": 3161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.298104763031006,
+      "learning_rate": 1.711074561403509e-05,
+      "loss": 1.7716,
+      "step": 3162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7762415409088135,
+      "learning_rate": 1.710983187134503e-05,
+      "loss": 1.8525,
+      "step": 3163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9675004482269287,
+      "learning_rate": 1.710891812865497e-05,
+      "loss": 1.2287,
+      "step": 3164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6124069690704346,
+      "learning_rate": 1.7108004385964915e-05,
+      "loss": 1.776,
+      "step": 3165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9033114910125732,
+      "learning_rate": 1.7107090643274856e-05,
+      "loss": 1.6648,
+      "step": 3166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.537710428237915,
+      "learning_rate": 1.7106176900584796e-05,
+      "loss": 2.0614,
+      "step": 3167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7489013671875,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 1.4031,
+      "step": 3168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9177730083465576,
+      "learning_rate": 1.710434941520468e-05,
+      "loss": 1.4205,
+      "step": 3169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.605905294418335,
+      "learning_rate": 1.710343567251462e-05,
+      "loss": 1.4301,
+      "step": 3170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.467700958251953,
+      "learning_rate": 1.710252192982456e-05,
+      "loss": 1.7226,
+      "step": 3171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3252370357513428,
+      "learning_rate": 1.7101608187134505e-05,
+      "loss": 2.0456,
+      "step": 3172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.169764280319214,
+      "learning_rate": 1.7100694444444446e-05,
+      "loss": 1.7297,
+      "step": 3173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.361903190612793,
+      "learning_rate": 1.7099780701754387e-05,
+      "loss": 1.9777,
+      "step": 3174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2199137210845947,
+      "learning_rate": 1.7098866959064327e-05,
+      "loss": 1.5135,
+      "step": 3175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.229008436203003,
+      "learning_rate": 1.709795321637427e-05,
+      "loss": 1.6175,
+      "step": 3176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3625378608703613,
+      "learning_rate": 1.709703947368421e-05,
+      "loss": 1.7083,
+      "step": 3177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.566396474838257,
+      "learning_rate": 1.7096125730994152e-05,
+      "loss": 2.3237,
+      "step": 3178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.696556568145752,
+      "learning_rate": 1.7095211988304096e-05,
+      "loss": 1.6218,
+      "step": 3179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1035521030426025,
+      "learning_rate": 1.7094298245614036e-05,
+      "loss": 1.8491,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8806302547454834,
+      "learning_rate": 1.7093384502923977e-05,
+      "loss": 1.4017,
+      "step": 3181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1449708938598633,
+      "learning_rate": 1.7092470760233918e-05,
+      "loss": 1.7232,
+      "step": 3182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3070261478424072,
+      "learning_rate": 1.709155701754386e-05,
+      "loss": 1.8629,
+      "step": 3183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5867044925689697,
+      "learning_rate": 1.7090643274853802e-05,
+      "loss": 1.8777,
+      "step": 3184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.055776357650757,
+      "learning_rate": 1.7089729532163742e-05,
+      "loss": 1.8076,
+      "step": 3185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.646949052810669,
+      "learning_rate": 1.7088815789473686e-05,
+      "loss": 2.3714,
+      "step": 3186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.116548538208008,
+      "learning_rate": 1.7087902046783627e-05,
+      "loss": 1.8054,
+      "step": 3187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1387360095977783,
+      "learning_rate": 1.7086988304093567e-05,
+      "loss": 1.5988,
+      "step": 3188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.378444194793701,
+      "learning_rate": 1.7086074561403508e-05,
+      "loss": 1.7923,
+      "step": 3189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8743083477020264,
+      "learning_rate": 1.7085160818713452e-05,
+      "loss": 1.5373,
+      "step": 3190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.452683925628662,
+      "learning_rate": 1.7084247076023392e-05,
+      "loss": 2.1292,
+      "step": 3191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2155654430389404,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 1.6817,
+      "step": 3192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.4092607498168945,
+      "learning_rate": 1.7082419590643277e-05,
+      "loss": 1.7228,
+      "step": 3193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.028909206390381,
+      "learning_rate": 1.7081505847953217e-05,
+      "loss": 1.5796,
+      "step": 3194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9131038188934326,
+      "learning_rate": 1.7080592105263158e-05,
+      "loss": 1.6744,
+      "step": 3195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2258996963500977,
+      "learning_rate": 1.70796783625731e-05,
+      "loss": 1.6538,
+      "step": 3196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1507182121276855,
+      "learning_rate": 1.7078764619883042e-05,
+      "loss": 2.0713,
+      "step": 3197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4665491580963135,
+      "learning_rate": 1.7077850877192983e-05,
+      "loss": 1.3282,
+      "step": 3198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3636317253112793,
+      "learning_rate": 1.7076937134502923e-05,
+      "loss": 1.8376,
+      "step": 3199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3552396297454834,
+      "learning_rate": 1.7076023391812867e-05,
+      "loss": 1.7844,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.144268274307251,
+      "learning_rate": 1.7075109649122808e-05,
+      "loss": 1.3481,
+      "step": 3201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2497618198394775,
+      "learning_rate": 1.707419590643275e-05,
+      "loss": 1.9405,
+      "step": 3202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.66597056388855,
+      "learning_rate": 1.707328216374269e-05,
+      "loss": 1.3588,
+      "step": 3203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.384192943572998,
+      "learning_rate": 1.7072368421052633e-05,
+      "loss": 2.1776,
+      "step": 3204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.827254295349121,
+      "learning_rate": 1.7071454678362573e-05,
+      "loss": 1.9162,
+      "step": 3205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7533769607543945,
+      "learning_rate": 1.7070540935672514e-05,
+      "loss": 2.1562,
+      "step": 3206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1076347827911377,
+      "learning_rate": 1.7069627192982458e-05,
+      "loss": 1.5637,
+      "step": 3207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.835230588912964,
+      "learning_rate": 1.7068713450292402e-05,
+      "loss": 1.506,
+      "step": 3208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.789031982421875,
+      "learning_rate": 1.7067799707602342e-05,
+      "loss": 1.3827,
+      "step": 3209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8542702198028564,
+      "learning_rate": 1.7066885964912283e-05,
+      "loss": 1.6922,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.454732656478882,
+      "learning_rate": 1.7065972222222223e-05,
+      "loss": 2.0416,
+      "step": 3211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5391128063201904,
+      "learning_rate": 1.7065058479532167e-05,
+      "loss": 1.7857,
+      "step": 3212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2393574714660645,
+      "learning_rate": 1.7064144736842108e-05,
+      "loss": 1.5252,
+      "step": 3213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.116631507873535,
+      "learning_rate": 1.7063230994152048e-05,
+      "loss": 1.7328,
+      "step": 3214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.234821081161499,
+      "learning_rate": 1.7062317251461992e-05,
+      "loss": 1.5962,
+      "step": 3215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8940110206604004,
+      "learning_rate": 1.7061403508771933e-05,
+      "loss": 1.723,
+      "step": 3216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.171295166015625,
+      "learning_rate": 1.7060489766081873e-05,
+      "loss": 1.5669,
+      "step": 3217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3955202102661133,
+      "learning_rate": 1.7059576023391814e-05,
+      "loss": 1.7085,
+      "step": 3218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7373650074005127,
+      "learning_rate": 1.7058662280701758e-05,
+      "loss": 1.4328,
+      "step": 3219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3877971172332764,
+      "learning_rate": 1.7057748538011698e-05,
+      "loss": 1.6107,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.034006357192993,
+      "learning_rate": 1.705683479532164e-05,
+      "loss": 1.7427,
+      "step": 3221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.221525192260742,
+      "learning_rate": 1.7055921052631583e-05,
+      "loss": 1.5045,
+      "step": 3222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1821482181549072,
+      "learning_rate": 1.7055007309941523e-05,
+      "loss": 1.8063,
+      "step": 3223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8929169178009033,
+      "learning_rate": 1.7054093567251464e-05,
+      "loss": 1.3864,
+      "step": 3224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2155351638793945,
+      "learning_rate": 1.7053179824561404e-05,
+      "loss": 1.7769,
+      "step": 3225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.151146411895752,
+      "learning_rate": 1.7052266081871348e-05,
+      "loss": 1.1953,
+      "step": 3226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.373142719268799,
+      "learning_rate": 1.705135233918129e-05,
+      "loss": 1.8065,
+      "step": 3227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4312918186187744,
+      "learning_rate": 1.705043859649123e-05,
+      "loss": 1.9948,
+      "step": 3228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.736542224884033,
+      "learning_rate": 1.7049524853801173e-05,
+      "loss": 1.4881,
+      "step": 3229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1634862422943115,
+      "learning_rate": 1.7048611111111114e-05,
+      "loss": 1.8471,
+      "step": 3230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.337000846862793,
+      "learning_rate": 1.7047697368421054e-05,
+      "loss": 1.8536,
+      "step": 3231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.437666416168213,
+      "learning_rate": 1.7046783625730995e-05,
+      "loss": 1.925,
+      "step": 3232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2982964515686035,
+      "learning_rate": 1.704586988304094e-05,
+      "loss": 1.7104,
+      "step": 3233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.954113006591797,
+      "learning_rate": 1.704495614035088e-05,
+      "loss": 1.67,
+      "step": 3234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5670080184936523,
+      "learning_rate": 1.704404239766082e-05,
+      "loss": 1.9424,
+      "step": 3235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7582993507385254,
+      "learning_rate": 1.7043128654970764e-05,
+      "loss": 1.3606,
+      "step": 3236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0295658111572266,
+      "learning_rate": 1.7042214912280704e-05,
+      "loss": 1.6655,
+      "step": 3237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.126549243927002,
+      "learning_rate": 1.7041301169590645e-05,
+      "loss": 1.6916,
+      "step": 3238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.815370798110962,
+      "learning_rate": 1.7040387426900585e-05,
+      "loss": 2.1454,
+      "step": 3239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.803979873657227,
+      "learning_rate": 1.703947368421053e-05,
+      "loss": 2.1942,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9335110187530518,
+      "learning_rate": 1.703855994152047e-05,
+      "loss": 1.638,
+      "step": 3241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0497050285339355,
+      "learning_rate": 1.703764619883041e-05,
+      "loss": 1.7565,
+      "step": 3242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.413461685180664,
+      "learning_rate": 1.7036732456140354e-05,
+      "loss": 1.9192,
+      "step": 3243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.223074436187744,
+      "learning_rate": 1.7035818713450295e-05,
+      "loss": 1.9954,
+      "step": 3244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4069254398345947,
+      "learning_rate": 1.7034904970760235e-05,
+      "loss": 2.3027,
+      "step": 3245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.464061737060547,
+      "learning_rate": 1.7033991228070176e-05,
+      "loss": 2.0384,
+      "step": 3246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5111398696899414,
+      "learning_rate": 1.703307748538012e-05,
+      "loss": 1.879,
+      "step": 3247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6917459964752197,
+      "learning_rate": 1.703216374269006e-05,
+      "loss": 1.7405,
+      "step": 3248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1340651512145996,
+      "learning_rate": 1.703125e-05,
+      "loss": 1.6993,
+      "step": 3249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6568126678466797,
+      "learning_rate": 1.7030336257309945e-05,
+      "loss": 1.9249,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.235816478729248,
+      "learning_rate": 1.7029422514619885e-05,
+      "loss": 1.9692,
+      "step": 3251
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.947490930557251,
+      "learning_rate": 1.7028508771929826e-05,
+      "loss": 1.7197,
+      "step": 3252
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.111073017120361,
+      "learning_rate": 1.7027595029239766e-05,
+      "loss": 2.3811,
+      "step": 3253
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.725674867630005,
+      "learning_rate": 1.702668128654971e-05,
+      "loss": 1.7518,
+      "step": 3254
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1800670623779297,
+      "learning_rate": 1.702576754385965e-05,
+      "loss": 1.5048,
+      "step": 3255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.974623918533325,
+      "learning_rate": 1.702485380116959e-05,
+      "loss": 2.1683,
+      "step": 3256
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8739981651306152,
+      "learning_rate": 1.7023940058479535e-05,
+      "loss": 1.5375,
+      "step": 3257
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0636789798736572,
+      "learning_rate": 1.7023026315789476e-05,
+      "loss": 1.7063,
+      "step": 3258
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3138556480407715,
+      "learning_rate": 1.7022112573099416e-05,
+      "loss": 2.0308,
+      "step": 3259
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.383655309677124,
+      "learning_rate": 1.7021198830409357e-05,
+      "loss": 2.0805,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3000025749206543,
+      "learning_rate": 1.70202850877193e-05,
+      "loss": 1.7918,
+      "step": 3261
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.775540590286255,
+      "learning_rate": 1.701937134502924e-05,
+      "loss": 2.1784,
+      "step": 3262
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.801785707473755,
+      "learning_rate": 1.701845760233918e-05,
+      "loss": 1.2944,
+      "step": 3263
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3648030757904053,
+      "learning_rate": 1.7017543859649125e-05,
+      "loss": 1.79,
+      "step": 3264
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7540595531463623,
+      "learning_rate": 1.7016630116959066e-05,
+      "loss": 1.2879,
+      "step": 3265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.250553846359253,
+      "learning_rate": 1.7015716374269007e-05,
+      "loss": 1.709,
+      "step": 3266
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.072406768798828,
+      "learning_rate": 1.7014802631578947e-05,
+      "loss": 1.6487,
+      "step": 3267
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1742770671844482,
+      "learning_rate": 1.701388888888889e-05,
+      "loss": 1.7842,
+      "step": 3268
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6572868824005127,
+      "learning_rate": 1.701297514619883e-05,
+      "loss": 1.4904,
+      "step": 3269
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.658061981201172,
+      "learning_rate": 1.7012061403508772e-05,
+      "loss": 1.5367,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.980959892272949,
+      "learning_rate": 1.7011147660818716e-05,
+      "loss": 1.5952,
+      "step": 3271
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9668381214141846,
+      "learning_rate": 1.7010233918128656e-05,
+      "loss": 1.6142,
+      "step": 3272
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.510850667953491,
+      "learning_rate": 1.7009320175438597e-05,
+      "loss": 1.9413,
+      "step": 3273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.786722421646118,
+      "learning_rate": 1.7008406432748538e-05,
+      "loss": 1.6731,
+      "step": 3274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.0597825050354,
+      "learning_rate": 1.700749269005848e-05,
+      "loss": 2.4628,
+      "step": 3275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1332602500915527,
+      "learning_rate": 1.7006578947368422e-05,
+      "loss": 1.8983,
+      "step": 3276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0083799362182617,
+      "learning_rate": 1.7005665204678362e-05,
+      "loss": 1.8718,
+      "step": 3277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.337623357772827,
+      "learning_rate": 1.7004751461988306e-05,
+      "loss": 1.6981,
+      "step": 3278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4523816108703613,
+      "learning_rate": 1.7003837719298247e-05,
+      "loss": 1.6693,
+      "step": 3279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.167690277099609,
+      "learning_rate": 1.7002923976608187e-05,
+      "loss": 2.1619,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.232893705368042,
+      "learning_rate": 1.7002010233918128e-05,
+      "loss": 1.995,
+      "step": 3281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.505018949508667,
+      "learning_rate": 1.7001096491228072e-05,
+      "loss": 2.0238,
+      "step": 3282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.057861566543579,
+      "learning_rate": 1.7000182748538012e-05,
+      "loss": 1.7539,
+      "step": 3283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.671978235244751,
+      "learning_rate": 1.6999269005847953e-05,
+      "loss": 1.4004,
+      "step": 3284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3055906295776367,
+      "learning_rate": 1.6998355263157897e-05,
+      "loss": 1.9159,
+      "step": 3285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0107333660125732,
+      "learning_rate": 1.6997441520467837e-05,
+      "loss": 1.7472,
+      "step": 3286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.247767925262451,
+      "learning_rate": 1.6996527777777778e-05,
+      "loss": 1.9281,
+      "step": 3287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6599788665771484,
+      "learning_rate": 1.699561403508772e-05,
+      "loss": 1.5558,
+      "step": 3288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.261500835418701,
+      "learning_rate": 1.6994700292397662e-05,
+      "loss": 1.5991,
+      "step": 3289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1953020095825195,
+      "learning_rate": 1.6993786549707603e-05,
+      "loss": 1.975,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.152505874633789,
+      "learning_rate": 1.6992872807017543e-05,
+      "loss": 1.7131,
+      "step": 3291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.802553653717041,
+      "learning_rate": 1.6991959064327487e-05,
+      "loss": 1.6355,
+      "step": 3292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0137853622436523,
+      "learning_rate": 1.6991045321637428e-05,
+      "loss": 1.8452,
+      "step": 3293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.990882158279419,
+      "learning_rate": 1.6990131578947372e-05,
+      "loss": 1.607,
+      "step": 3294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0110437870025635,
+      "learning_rate": 1.6989217836257312e-05,
+      "loss": 1.7162,
+      "step": 3295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.026024103164673,
+      "learning_rate": 1.6988304093567253e-05,
+      "loss": 1.559,
+      "step": 3296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.266671895980835,
+      "learning_rate": 1.6987390350877197e-05,
+      "loss": 1.49,
+      "step": 3297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.020038366317749,
+      "learning_rate": 1.6986476608187137e-05,
+      "loss": 1.846,
+      "step": 3298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.790860891342163,
+      "learning_rate": 1.6985562865497078e-05,
+      "loss": 1.4293,
+      "step": 3299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1167714595794678,
+      "learning_rate": 1.698464912280702e-05,
+      "loss": 1.6307,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3417351245880127,
+      "learning_rate": 1.6983735380116962e-05,
+      "loss": 2.173,
+      "step": 3301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8670692443847656,
+      "learning_rate": 1.6982821637426903e-05,
+      "loss": 1.8627,
+      "step": 3302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1208629608154297,
+      "learning_rate": 1.6981907894736843e-05,
+      "loss": 1.6878,
+      "step": 3303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8776562213897705,
+      "learning_rate": 1.6980994152046787e-05,
+      "loss": 1.7357,
+      "step": 3304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.494249105453491,
+      "learning_rate": 1.6980080409356728e-05,
+      "loss": 1.4718,
+      "step": 3305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1169440746307373,
+      "learning_rate": 1.6979166666666668e-05,
+      "loss": 1.8907,
+      "step": 3306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.122918128967285,
+      "learning_rate": 1.697825292397661e-05,
+      "loss": 1.6593,
+      "step": 3307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.757357120513916,
+      "learning_rate": 1.6977339181286553e-05,
+      "loss": 1.5327,
+      "step": 3308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.069373846054077,
+      "learning_rate": 1.6976425438596493e-05,
+      "loss": 1.4597,
+      "step": 3309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.955864906311035,
+      "learning_rate": 1.6975511695906434e-05,
+      "loss": 1.7331,
+      "step": 3310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.082820415496826,
+      "learning_rate": 1.6974597953216378e-05,
+      "loss": 1.5248,
+      "step": 3311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.765523910522461,
+      "learning_rate": 1.6973684210526318e-05,
+      "loss": 2.1797,
+      "step": 3312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.721017360687256,
+      "learning_rate": 1.697277046783626e-05,
+      "loss": 1.6244,
+      "step": 3313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.620659351348877,
+      "learning_rate": 1.69718567251462e-05,
+      "loss": 1.3457,
+      "step": 3314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6128933429718018,
+      "learning_rate": 1.6970942982456143e-05,
+      "loss": 2.0112,
+      "step": 3315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.011197090148926,
+      "learning_rate": 1.6970029239766084e-05,
+      "loss": 1.5675,
+      "step": 3316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.158548593521118,
+      "learning_rate": 1.6969115497076024e-05,
+      "loss": 1.7179,
+      "step": 3317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.180969476699829,
+      "learning_rate": 1.6968201754385968e-05,
+      "loss": 1.9019,
+      "step": 3318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6877150535583496,
+      "learning_rate": 1.696728801169591e-05,
+      "loss": 1.506,
+      "step": 3319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1160569190979004,
+      "learning_rate": 1.696637426900585e-05,
+      "loss": 1.7792,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.023946762084961,
+      "learning_rate": 1.696546052631579e-05,
+      "loss": 1.7487,
+      "step": 3321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7624340057373047,
+      "learning_rate": 1.6964546783625734e-05,
+      "loss": 1.5235,
+      "step": 3322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8665380477905273,
+      "learning_rate": 1.6963633040935674e-05,
+      "loss": 1.8022,
+      "step": 3323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8895022869110107,
+      "learning_rate": 1.6962719298245615e-05,
+      "loss": 1.7287,
+      "step": 3324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8593900203704834,
+      "learning_rate": 1.696180555555556e-05,
+      "loss": 1.5936,
+      "step": 3325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0804312229156494,
+      "learning_rate": 1.69608918128655e-05,
+      "loss": 1.8504,
+      "step": 3326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2581984996795654,
+      "learning_rate": 1.695997807017544e-05,
+      "loss": 1.7777,
+      "step": 3327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.905151605606079,
+      "learning_rate": 1.695906432748538e-05,
+      "loss": 1.5007,
+      "step": 3328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5869345664978027,
+      "learning_rate": 1.6958150584795324e-05,
+      "loss": 1.9361,
+      "step": 3329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.684255599975586,
+      "learning_rate": 1.6957236842105265e-05,
+      "loss": 1.4607,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0590155124664307,
+      "learning_rate": 1.6956323099415205e-05,
+      "loss": 1.9834,
+      "step": 3331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9123709201812744,
+      "learning_rate": 1.695540935672515e-05,
+      "loss": 1.5515,
+      "step": 3332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.569945812225342,
+      "learning_rate": 1.695449561403509e-05,
+      "loss": 1.7991,
+      "step": 3333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0962326526641846,
+      "learning_rate": 1.695358187134503e-05,
+      "loss": 1.7446,
+      "step": 3334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.158942699432373,
+      "learning_rate": 1.695266812865497e-05,
+      "loss": 1.8438,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1048731803894043,
+      "learning_rate": 1.6951754385964915e-05,
+      "loss": 1.7527,
+      "step": 3336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6700615882873535,
+      "learning_rate": 1.6950840643274855e-05,
+      "loss": 1.9112,
+      "step": 3337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1529839038848877,
+      "learning_rate": 1.6949926900584796e-05,
+      "loss": 1.9969,
+      "step": 3338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1225554943084717,
+      "learning_rate": 1.694901315789474e-05,
+      "loss": 1.7354,
+      "step": 3339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4633984565734863,
+      "learning_rate": 1.694809941520468e-05,
+      "loss": 2.108,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8008642196655273,
+      "learning_rate": 1.694718567251462e-05,
+      "loss": 1.8064,
+      "step": 3341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8734869956970215,
+      "learning_rate": 1.694627192982456e-05,
+      "loss": 1.6795,
+      "step": 3342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5147900581359863,
+      "learning_rate": 1.6945358187134505e-05,
+      "loss": 1.5629,
+      "step": 3343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5762734413146973,
+      "learning_rate": 1.6944444444444446e-05,
+      "loss": 1.5044,
+      "step": 3344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9941513538360596,
+      "learning_rate": 1.6943530701754386e-05,
+      "loss": 1.7848,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0304243564605713,
+      "learning_rate": 1.694261695906433e-05,
+      "loss": 1.4598,
+      "step": 3346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5533130168914795,
+      "learning_rate": 1.694170321637427e-05,
+      "loss": 1.6592,
+      "step": 3347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9867753982543945,
+      "learning_rate": 1.694078947368421e-05,
+      "loss": 1.8534,
+      "step": 3348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6646671295166016,
+      "learning_rate": 1.693987573099415e-05,
+      "loss": 1.4346,
+      "step": 3349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9506137371063232,
+      "learning_rate": 1.6938961988304096e-05,
+      "loss": 1.9886,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0491318702697754,
+      "learning_rate": 1.6938048245614036e-05,
+      "loss": 1.5498,
+      "step": 3351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8656527996063232,
+      "learning_rate": 1.6937134502923977e-05,
+      "loss": 1.6332,
+      "step": 3352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.760280132293701,
+      "learning_rate": 1.693622076023392e-05,
+      "loss": 1.4206,
+      "step": 3353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1863551139831543,
+      "learning_rate": 1.693530701754386e-05,
+      "loss": 1.4506,
+      "step": 3354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1613528728485107,
+      "learning_rate": 1.69343932748538e-05,
+      "loss": 1.6915,
+      "step": 3355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5214085578918457,
+      "learning_rate": 1.6933479532163742e-05,
+      "loss": 1.7959,
+      "step": 3356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.610665798187256,
+      "learning_rate": 1.6932565789473686e-05,
+      "loss": 1.4682,
+      "step": 3357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0218801498413086,
+      "learning_rate": 1.6931652046783627e-05,
+      "loss": 1.5224,
+      "step": 3358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.321129560470581,
+      "learning_rate": 1.6930738304093567e-05,
+      "loss": 1.2635,
+      "step": 3359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.075977087020874,
+      "learning_rate": 1.692982456140351e-05,
+      "loss": 1.8659,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8524298667907715,
+      "learning_rate": 1.692891081871345e-05,
+      "loss": 1.4276,
+      "step": 3361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1310505867004395,
+      "learning_rate": 1.6927997076023392e-05,
+      "loss": 1.8767,
+      "step": 3362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.833348035812378,
+      "learning_rate": 1.6927083333333333e-05,
+      "loss": 1.5996,
+      "step": 3363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.450411558151245,
+      "learning_rate": 1.6926169590643276e-05,
+      "loss": 1.4523,
+      "step": 3364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.422713279724121,
+      "learning_rate": 1.6925255847953217e-05,
+      "loss": 1.584,
+      "step": 3365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.624586343765259,
+      "learning_rate": 1.6924342105263158e-05,
+      "loss": 1.4212,
+      "step": 3366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4649267196655273,
+      "learning_rate": 1.69234283625731e-05,
+      "loss": 1.5469,
+      "step": 3367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4804069995880127,
+      "learning_rate": 1.6922514619883042e-05,
+      "loss": 1.3764,
+      "step": 3368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.737070322036743,
+      "learning_rate": 1.6921600877192982e-05,
+      "loss": 1.6617,
+      "step": 3369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2107720375061035,
+      "learning_rate": 1.6920687134502923e-05,
+      "loss": 2.0155,
+      "step": 3370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.988523483276367,
+      "learning_rate": 1.6919773391812867e-05,
+      "loss": 1.7173,
+      "step": 3371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1351804733276367,
+      "learning_rate": 1.6918859649122807e-05,
+      "loss": 1.7188,
+      "step": 3372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.32625675201416,
+      "learning_rate": 1.6917945906432748e-05,
+      "loss": 1.6916,
+      "step": 3373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.160682439804077,
+      "learning_rate": 1.6917032163742692e-05,
+      "loss": 1.8729,
+      "step": 3374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2138848304748535,
+      "learning_rate": 1.6916118421052632e-05,
+      "loss": 1.9365,
+      "step": 3375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7589361667633057,
+      "learning_rate": 1.6915204678362573e-05,
+      "loss": 1.404,
+      "step": 3376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6698639392852783,
+      "learning_rate": 1.6914290935672513e-05,
+      "loss": 1.4443,
+      "step": 3377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.144542694091797,
+      "learning_rate": 1.6913377192982457e-05,
+      "loss": 1.839,
+      "step": 3378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1060378551483154,
+      "learning_rate": 1.69124634502924e-05,
+      "loss": 1.5912,
+      "step": 3379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.783698320388794,
+      "learning_rate": 1.6911549707602342e-05,
+      "loss": 2.1233,
+      "step": 3380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8811066150665283,
+      "learning_rate": 1.6910635964912282e-05,
+      "loss": 2.0572,
+      "step": 3381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.176147699356079,
+      "learning_rate": 1.6909722222222223e-05,
+      "loss": 1.8019,
+      "step": 3382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3074777126312256,
+      "learning_rate": 1.6908808479532167e-05,
+      "loss": 1.9277,
+      "step": 3383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.306277751922607,
+      "learning_rate": 1.6907894736842107e-05,
+      "loss": 2.3222,
+      "step": 3384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5250515937805176,
+      "learning_rate": 1.6906980994152048e-05,
+      "loss": 1.7142,
+      "step": 3385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.16782283782959,
+      "learning_rate": 1.6906067251461992e-05,
+      "loss": 1.565,
+      "step": 3386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.115044116973877,
+      "learning_rate": 1.6905153508771932e-05,
+      "loss": 1.9055,
+      "step": 3387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1917526721954346,
+      "learning_rate": 1.6904239766081873e-05,
+      "loss": 1.9122,
+      "step": 3388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0721919536590576,
+      "learning_rate": 1.6903326023391813e-05,
+      "loss": 1.7599,
+      "step": 3389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4262986183166504,
+      "learning_rate": 1.6902412280701757e-05,
+      "loss": 1.8368,
+      "step": 3390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.665475606918335,
+      "learning_rate": 1.6901498538011698e-05,
+      "loss": 1.6481,
+      "step": 3391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.339308500289917,
+      "learning_rate": 1.690058479532164e-05,
+      "loss": 1.736,
+      "step": 3392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.571807384490967,
+      "learning_rate": 1.6899671052631582e-05,
+      "loss": 2.0845,
+      "step": 3393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2909696102142334,
+      "learning_rate": 1.6898757309941523e-05,
+      "loss": 2.007,
+      "step": 3394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1836981773376465,
+      "learning_rate": 1.6897843567251463e-05,
+      "loss": 1.6722,
+      "step": 3395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9461302757263184,
+      "learning_rate": 1.6896929824561404e-05,
+      "loss": 1.9758,
+      "step": 3396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5728914737701416,
+      "learning_rate": 1.6896016081871348e-05,
+      "loss": 1.2558,
+      "step": 3397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3128786087036133,
+      "learning_rate": 1.6895102339181288e-05,
+      "loss": 1.854,
+      "step": 3398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9886138439178467,
+      "learning_rate": 1.689418859649123e-05,
+      "loss": 2.1264,
+      "step": 3399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.131190776824951,
+      "learning_rate": 1.6893274853801173e-05,
+      "loss": 1.7579,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.056150197982788,
+      "learning_rate": 1.6892361111111113e-05,
+      "loss": 1.4332,
+      "step": 3401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6969902515411377,
+      "learning_rate": 1.6891447368421054e-05,
+      "loss": 2.2066,
+      "step": 3402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2516374588012695,
+      "learning_rate": 1.6890533625730994e-05,
+      "loss": 1.6169,
+      "step": 3403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.486438035964966,
+      "learning_rate": 1.6889619883040938e-05,
+      "loss": 2.0144,
+      "step": 3404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3712666034698486,
+      "learning_rate": 1.688870614035088e-05,
+      "loss": 1.9954,
+      "step": 3405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.260723114013672,
+      "learning_rate": 1.688779239766082e-05,
+      "loss": 1.7314,
+      "step": 3406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.810244560241699,
+      "learning_rate": 1.6886878654970763e-05,
+      "loss": 1.4815,
+      "step": 3407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5496349334716797,
+      "learning_rate": 1.6885964912280704e-05,
+      "loss": 1.5893,
+      "step": 3408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1207778453826904,
+      "learning_rate": 1.6885051169590644e-05,
+      "loss": 1.5749,
+      "step": 3409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5442323684692383,
+      "learning_rate": 1.6884137426900585e-05,
+      "loss": 1.4041,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0212764739990234,
+      "learning_rate": 1.688322368421053e-05,
+      "loss": 1.7776,
+      "step": 3411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8808412551879883,
+      "learning_rate": 1.688230994152047e-05,
+      "loss": 1.5493,
+      "step": 3412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2121872901916504,
+      "learning_rate": 1.688139619883041e-05,
+      "loss": 1.8051,
+      "step": 3413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.104858160018921,
+      "learning_rate": 1.6880482456140354e-05,
+      "loss": 1.834,
+      "step": 3414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4002599716186523,
+      "learning_rate": 1.6879568713450294e-05,
+      "loss": 1.9841,
+      "step": 3415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.535322427749634,
+      "learning_rate": 1.6878654970760235e-05,
+      "loss": 1.2831,
+      "step": 3416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.645480155944824,
+      "learning_rate": 1.6877741228070175e-05,
+      "loss": 2.2457,
+      "step": 3417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3053877353668213,
+      "learning_rate": 1.687682748538012e-05,
+      "loss": 1.8813,
+      "step": 3418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0844571590423584,
+      "learning_rate": 1.687591374269006e-05,
+      "loss": 1.7215,
+      "step": 3419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0095911026000977,
+      "learning_rate": 1.6875e-05,
+      "loss": 1.6643,
+      "step": 3420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9989373683929443,
+      "learning_rate": 1.6874086257309944e-05,
+      "loss": 1.3799,
+      "step": 3421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0567820072174072,
+      "learning_rate": 1.6873172514619885e-05,
+      "loss": 1.9516,
+      "step": 3422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.808330535888672,
+      "learning_rate": 1.6872258771929825e-05,
+      "loss": 1.7406,
+      "step": 3423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.641040802001953,
+      "learning_rate": 1.6871345029239766e-05,
+      "loss": 1.9625,
+      "step": 3424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.032424211502075,
+      "learning_rate": 1.687043128654971e-05,
+      "loss": 1.5559,
+      "step": 3425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.155975580215454,
+      "learning_rate": 1.686951754385965e-05,
+      "loss": 1.7919,
+      "step": 3426
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8463497161865234,
+      "learning_rate": 1.686860380116959e-05,
+      "loss": 1.7194,
+      "step": 3427
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7136013507843018,
+      "learning_rate": 1.6867690058479535e-05,
+      "loss": 2.1176,
+      "step": 3428
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2222349643707275,
+      "learning_rate": 1.6866776315789475e-05,
+      "loss": 1.8936,
+      "step": 3429
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1770851612091064,
+      "learning_rate": 1.6865862573099416e-05,
+      "loss": 1.8024,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8849847316741943,
+      "learning_rate": 1.6864948830409356e-05,
+      "loss": 1.7663,
+      "step": 3431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.875443458557129,
+      "learning_rate": 1.68640350877193e-05,
+      "loss": 1.4465,
+      "step": 3432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6030426025390625,
+      "learning_rate": 1.686312134502924e-05,
+      "loss": 1.6754,
+      "step": 3433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.675312042236328,
+      "learning_rate": 1.686220760233918e-05,
+      "loss": 1.4348,
+      "step": 3434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1227033138275146,
+      "learning_rate": 1.6861293859649125e-05,
+      "loss": 1.792,
+      "step": 3435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.407623291015625,
+      "learning_rate": 1.6860380116959066e-05,
+      "loss": 1.8651,
+      "step": 3436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5350170135498047,
+      "learning_rate": 1.6859466374269006e-05,
+      "loss": 1.8784,
+      "step": 3437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8886566162109375,
+      "learning_rate": 1.6858552631578947e-05,
+      "loss": 1.7611,
+      "step": 3438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4276273250579834,
+      "learning_rate": 1.685763888888889e-05,
+      "loss": 1.7833,
+      "step": 3439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.513723850250244,
+      "learning_rate": 1.685672514619883e-05,
+      "loss": 1.9088,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3772175312042236,
+      "learning_rate": 1.685581140350877e-05,
+      "loss": 1.3451,
+      "step": 3441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6084914207458496,
+      "learning_rate": 1.6854897660818716e-05,
+      "loss": 2.4273,
+      "step": 3442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.865950584411621,
+      "learning_rate": 1.6853983918128656e-05,
+      "loss": 1.5304,
+      "step": 3443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8008108139038086,
+      "learning_rate": 1.6853070175438597e-05,
+      "loss": 1.3911,
+      "step": 3444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.413135051727295,
+      "learning_rate": 1.6852156432748537e-05,
+      "loss": 1.7868,
+      "step": 3445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.603036642074585,
+      "learning_rate": 1.685124269005848e-05,
+      "loss": 2.2119,
+      "step": 3446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.092048168182373,
+      "learning_rate": 1.685032894736842e-05,
+      "loss": 1.885,
+      "step": 3447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4915566444396973,
+      "learning_rate": 1.6849415204678362e-05,
+      "loss": 1.7753,
+      "step": 3448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.715038776397705,
+      "learning_rate": 1.6848501461988306e-05,
+      "loss": 1.4014,
+      "step": 3449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.018993854522705,
+      "learning_rate": 1.6847587719298247e-05,
+      "loss": 1.664,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.159846305847168,
+      "learning_rate": 1.6846673976608187e-05,
+      "loss": 2.0625,
+      "step": 3451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2315502166748047,
+      "learning_rate": 1.6845760233918128e-05,
+      "loss": 1.7171,
+      "step": 3452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8860890865325928,
+      "learning_rate": 1.684484649122807e-05,
+      "loss": 1.6051,
+      "step": 3453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9292173385620117,
+      "learning_rate": 1.6843932748538012e-05,
+      "loss": 1.6559,
+      "step": 3454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1722586154937744,
+      "learning_rate": 1.6843019005847953e-05,
+      "loss": 1.8303,
+      "step": 3455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9568192958831787,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 1.5961,
+      "step": 3456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.906214475631714,
+      "learning_rate": 1.6841191520467837e-05,
+      "loss": 1.5304,
+      "step": 3457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1942715644836426,
+      "learning_rate": 1.6840277777777778e-05,
+      "loss": 1.6871,
+      "step": 3458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5303635597229004,
+      "learning_rate": 1.6839364035087718e-05,
+      "loss": 2.0012,
+      "step": 3459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8510279655456543,
+      "learning_rate": 1.6838450292397662e-05,
+      "loss": 1.659,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.41807222366333,
+      "learning_rate": 1.6837536549707602e-05,
+      "loss": 1.9285,
+      "step": 3461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0581488609313965,
+      "learning_rate": 1.6836622807017543e-05,
+      "loss": 1.7358,
+      "step": 3462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.867706775665283,
+      "learning_rate": 1.6835709064327487e-05,
+      "loss": 1.3803,
+      "step": 3463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.745589017868042,
+      "learning_rate": 1.683479532163743e-05,
+      "loss": 1.6316,
+      "step": 3464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7968475818634033,
+      "learning_rate": 1.683388157894737e-05,
+      "loss": 1.6111,
+      "step": 3465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.277395009994507,
+      "learning_rate": 1.6832967836257312e-05,
+      "loss": 1.695,
+      "step": 3466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.337796926498413,
+      "learning_rate": 1.6832054093567252e-05,
+      "loss": 1.8633,
+      "step": 3467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5555145740509033,
+      "learning_rate": 1.6831140350877196e-05,
+      "loss": 1.2986,
+      "step": 3468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.710231304168701,
+      "learning_rate": 1.6830226608187137e-05,
+      "loss": 1.5761,
+      "step": 3469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6368753910064697,
+      "learning_rate": 1.6829312865497077e-05,
+      "loss": 2.4142,
+      "step": 3470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5828871726989746,
+      "learning_rate": 1.682839912280702e-05,
+      "loss": 2.0124,
+      "step": 3471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.000549077987671,
+      "learning_rate": 1.6827485380116962e-05,
+      "loss": 1.6915,
+      "step": 3472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7520227432250977,
+      "learning_rate": 1.6826571637426902e-05,
+      "loss": 1.8806,
+      "step": 3473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.137794256210327,
+      "learning_rate": 1.6825657894736843e-05,
+      "loss": 1.7638,
+      "step": 3474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8851430416107178,
+      "learning_rate": 1.6824744152046787e-05,
+      "loss": 1.5192,
+      "step": 3475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0133447647094727,
+      "learning_rate": 1.6823830409356727e-05,
+      "loss": 1.5995,
+      "step": 3476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9777369499206543,
+      "learning_rate": 1.6822916666666668e-05,
+      "loss": 1.9585,
+      "step": 3477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3811521530151367,
+      "learning_rate": 1.6822002923976612e-05,
+      "loss": 1.5222,
+      "step": 3478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.186584711074829,
+      "learning_rate": 1.6821089181286552e-05,
+      "loss": 1.8035,
+      "step": 3479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.570305347442627,
+      "learning_rate": 1.6820175438596493e-05,
+      "loss": 1.5306,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.299650192260742,
+      "learning_rate": 1.6819261695906433e-05,
+      "loss": 1.5977,
+      "step": 3481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.890377998352051,
+      "learning_rate": 1.6818347953216377e-05,
+      "loss": 1.5248,
+      "step": 3482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9810402393341064,
+      "learning_rate": 1.6817434210526318e-05,
+      "loss": 1.7661,
+      "step": 3483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.660581111907959,
+      "learning_rate": 1.681652046783626e-05,
+      "loss": 1.7744,
+      "step": 3484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.313044548034668,
+      "learning_rate": 1.6815606725146202e-05,
+      "loss": 2.039,
+      "step": 3485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1559207439422607,
+      "learning_rate": 1.6814692982456143e-05,
+      "loss": 1.772,
+      "step": 3486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2059671878814697,
+      "learning_rate": 1.6813779239766083e-05,
+      "loss": 1.7856,
+      "step": 3487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.866426944732666,
+      "learning_rate": 1.6812865497076024e-05,
+      "loss": 1.5562,
+      "step": 3488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7014260292053223,
+      "learning_rate": 1.6811951754385968e-05,
+      "loss": 1.5423,
+      "step": 3489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1489899158477783,
+      "learning_rate": 1.6811038011695908e-05,
+      "loss": 2.032,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9446029663085938,
+      "learning_rate": 1.681012426900585e-05,
+      "loss": 1.735,
+      "step": 3491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2946908473968506,
+      "learning_rate": 1.6809210526315793e-05,
+      "loss": 1.9015,
+      "step": 3492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.841278076171875,
+      "learning_rate": 1.6808296783625733e-05,
+      "loss": 1.7897,
+      "step": 3493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.775319814682007,
+      "learning_rate": 1.6807383040935674e-05,
+      "loss": 1.3691,
+      "step": 3494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1901886463165283,
+      "learning_rate": 1.6806469298245614e-05,
+      "loss": 1.6121,
+      "step": 3495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.933124542236328,
+      "learning_rate": 1.6805555555555558e-05,
+      "loss": 1.3414,
+      "step": 3496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.134348392486572,
+      "learning_rate": 1.68046418128655e-05,
+      "loss": 2.2637,
+      "step": 3497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8868329524993896,
+      "learning_rate": 1.680372807017544e-05,
+      "loss": 1.6864,
+      "step": 3498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.055542230606079,
+      "learning_rate": 1.6802814327485383e-05,
+      "loss": 1.6202,
+      "step": 3499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0970993041992188,
+      "learning_rate": 1.6801900584795324e-05,
+      "loss": 1.7938,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.352985382080078,
+      "learning_rate": 1.6800986842105264e-05,
+      "loss": 1.8022,
+      "step": 3501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.970428943634033,
+      "learning_rate": 1.6800073099415205e-05,
+      "loss": 1.8571,
+      "step": 3502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4539575576782227,
+      "learning_rate": 1.679915935672515e-05,
+      "loss": 1.3098,
+      "step": 3503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1274735927581787,
+      "learning_rate": 1.679824561403509e-05,
+      "loss": 1.5272,
+      "step": 3504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1281185150146484,
+      "learning_rate": 1.679733187134503e-05,
+      "loss": 1.6254,
+      "step": 3505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.015827655792236,
+      "learning_rate": 1.6796418128654974e-05,
+      "loss": 1.6775,
+      "step": 3506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1182000637054443,
+      "learning_rate": 1.6795504385964914e-05,
+      "loss": 1.5406,
+      "step": 3507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7614731788635254,
+      "learning_rate": 1.6794590643274855e-05,
+      "loss": 1.5877,
+      "step": 3508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0884499549865723,
+      "learning_rate": 1.6793676900584795e-05,
+      "loss": 1.4877,
+      "step": 3509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9668381214141846,
+      "learning_rate": 1.679276315789474e-05,
+      "loss": 1.6535,
+      "step": 3510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4076855182647705,
+      "learning_rate": 1.679184941520468e-05,
+      "loss": 1.9137,
+      "step": 3511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6413357257843018,
+      "learning_rate": 1.679093567251462e-05,
+      "loss": 1.5523,
+      "step": 3512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1190733909606934,
+      "learning_rate": 1.6790021929824564e-05,
+      "loss": 1.5552,
+      "step": 3513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0670738220214844,
+      "learning_rate": 1.6789108187134505e-05,
+      "loss": 1.7785,
+      "step": 3514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.033735990524292,
+      "learning_rate": 1.6788194444444445e-05,
+      "loss": 1.7883,
+      "step": 3515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7517409324645996,
+      "learning_rate": 1.6787280701754386e-05,
+      "loss": 1.9242,
+      "step": 3516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.27648663520813,
+      "learning_rate": 1.678636695906433e-05,
+      "loss": 1.6585,
+      "step": 3517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.038425922393799,
+      "learning_rate": 1.678545321637427e-05,
+      "loss": 1.712,
+      "step": 3518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.485164165496826,
+      "learning_rate": 1.678453947368421e-05,
+      "loss": 1.8756,
+      "step": 3519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.772684097290039,
+      "learning_rate": 1.6783625730994155e-05,
+      "loss": 1.6586,
+      "step": 3520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8853530883789062,
+      "learning_rate": 1.6782711988304095e-05,
+      "loss": 1.507,
+      "step": 3521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7893245220184326,
+      "learning_rate": 1.6781798245614036e-05,
+      "loss": 1.5102,
+      "step": 3522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.661376953125,
+      "learning_rate": 1.6780884502923976e-05,
+      "loss": 1.6335,
+      "step": 3523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.55686354637146,
+      "learning_rate": 1.677997076023392e-05,
+      "loss": 2.0871,
+      "step": 3524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9409778118133545,
+      "learning_rate": 1.677905701754386e-05,
+      "loss": 1.6089,
+      "step": 3525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4099349975585938,
+      "learning_rate": 1.67781432748538e-05,
+      "loss": 1.3954,
+      "step": 3526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.327326536178589,
+      "learning_rate": 1.6777229532163745e-05,
+      "loss": 1.769,
+      "step": 3527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.951307535171509,
+      "learning_rate": 1.6776315789473686e-05,
+      "loss": 1.732,
+      "step": 3528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1456167697906494,
+      "learning_rate": 1.6775402046783626e-05,
+      "loss": 1.9034,
+      "step": 3529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2361531257629395,
+      "learning_rate": 1.6774488304093567e-05,
+      "loss": 1.9112,
+      "step": 3530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.968881607055664,
+      "learning_rate": 1.677357456140351e-05,
+      "loss": 1.3299,
+      "step": 3531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.956052303314209,
+      "learning_rate": 1.677266081871345e-05,
+      "loss": 1.6499,
+      "step": 3532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8419814109802246,
+      "learning_rate": 1.677174707602339e-05,
+      "loss": 1.8532,
+      "step": 3533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9090566635131836,
+      "learning_rate": 1.6770833333333336e-05,
+      "loss": 1.8054,
+      "step": 3534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.953658103942871,
+      "learning_rate": 1.6769919590643276e-05,
+      "loss": 1.4533,
+      "step": 3535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.025373697280884,
+      "learning_rate": 1.6769005847953217e-05,
+      "loss": 1.5564,
+      "step": 3536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.026625871658325,
+      "learning_rate": 1.6768092105263157e-05,
+      "loss": 1.5184,
+      "step": 3537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0231986045837402,
+      "learning_rate": 1.67671783625731e-05,
+      "loss": 1.5977,
+      "step": 3538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8376336097717285,
+      "learning_rate": 1.676626461988304e-05,
+      "loss": 1.2673,
+      "step": 3539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6299381256103516,
+      "learning_rate": 1.6765350877192982e-05,
+      "loss": 2.2099,
+      "step": 3540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.662499189376831,
+      "learning_rate": 1.6764437134502926e-05,
+      "loss": 1.9856,
+      "step": 3541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2039310932159424,
+      "learning_rate": 1.6763523391812867e-05,
+      "loss": 1.5214,
+      "step": 3542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.307389736175537,
+      "learning_rate": 1.6762609649122807e-05,
+      "loss": 1.7488,
+      "step": 3543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.863039255142212,
+      "learning_rate": 1.6761695906432748e-05,
+      "loss": 2.1896,
+      "step": 3544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.820838212966919,
+      "learning_rate": 1.676078216374269e-05,
+      "loss": 1.3871,
+      "step": 3545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0906591415405273,
+      "learning_rate": 1.6759868421052632e-05,
+      "loss": 1.8753,
+      "step": 3546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.481018543243408,
+      "learning_rate": 1.6758954678362573e-05,
+      "loss": 1.8102,
+      "step": 3547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0223522186279297,
+      "learning_rate": 1.6758040935672516e-05,
+      "loss": 1.8396,
+      "step": 3548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1893599033355713,
+      "learning_rate": 1.6757127192982457e-05,
+      "loss": 1.7241,
+      "step": 3549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0147862434387207,
+      "learning_rate": 1.67562134502924e-05,
+      "loss": 1.4059,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.256582736968994,
+      "learning_rate": 1.675529970760234e-05,
+      "loss": 1.7521,
+      "step": 3551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.794424295425415,
+      "learning_rate": 1.6754385964912282e-05,
+      "loss": 1.6215,
+      "step": 3552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2364981174468994,
+      "learning_rate": 1.6753472222222226e-05,
+      "loss": 1.7159,
+      "step": 3553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2043893337249756,
+      "learning_rate": 1.6752558479532166e-05,
+      "loss": 1.4428,
+      "step": 3554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.966707468032837,
+      "learning_rate": 1.6751644736842107e-05,
+      "loss": 2.0393,
+      "step": 3555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9420664310455322,
+      "learning_rate": 1.6750730994152047e-05,
+      "loss": 1.5748,
+      "step": 3556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7639379501342773,
+      "learning_rate": 1.674981725146199e-05,
+      "loss": 1.6748,
+      "step": 3557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2313334941864014,
+      "learning_rate": 1.6748903508771932e-05,
+      "loss": 1.8605,
+      "step": 3558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5536108016967773,
+      "learning_rate": 1.6747989766081872e-05,
+      "loss": 1.564,
+      "step": 3559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.203167676925659,
+      "learning_rate": 1.6747076023391816e-05,
+      "loss": 1.7462,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1157500743865967,
+      "learning_rate": 1.6746162280701757e-05,
+      "loss": 1.8157,
+      "step": 3561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3158950805664062,
+      "learning_rate": 1.6745248538011697e-05,
+      "loss": 2.042,
+      "step": 3562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8319404125213623,
+      "learning_rate": 1.6744334795321638e-05,
+      "loss": 1.6001,
+      "step": 3563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.537825107574463,
+      "learning_rate": 1.6743421052631582e-05,
+      "loss": 1.6851,
+      "step": 3564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6805245876312256,
+      "learning_rate": 1.6742507309941522e-05,
+      "loss": 1.6131,
+      "step": 3565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.781766653060913,
+      "learning_rate": 1.6741593567251463e-05,
+      "loss": 1.9019,
+      "step": 3566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3558175563812256,
+      "learning_rate": 1.6740679824561407e-05,
+      "loss": 2.3061,
+      "step": 3567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.727294445037842,
+      "learning_rate": 1.6739766081871347e-05,
+      "loss": 1.712,
+      "step": 3568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6517322063446045,
+      "learning_rate": 1.6738852339181288e-05,
+      "loss": 1.4142,
+      "step": 3569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9359002113342285,
+      "learning_rate": 1.673793859649123e-05,
+      "loss": 1.476,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2699673175811768,
+      "learning_rate": 1.6737024853801172e-05,
+      "loss": 2.1056,
+      "step": 3571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0978713035583496,
+      "learning_rate": 1.6736111111111113e-05,
+      "loss": 1.456,
+      "step": 3572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.386380910873413,
+      "learning_rate": 1.6735197368421053e-05,
+      "loss": 1.9087,
+      "step": 3573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9125723838806152,
+      "learning_rate": 1.6734283625730997e-05,
+      "loss": 1.7428,
+      "step": 3574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9263129234313965,
+      "learning_rate": 1.6733369883040938e-05,
+      "loss": 1.8865,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4470725059509277,
+      "learning_rate": 1.673245614035088e-05,
+      "loss": 2.1951,
+      "step": 3576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7323460578918457,
+      "learning_rate": 1.673154239766082e-05,
+      "loss": 1.5653,
+      "step": 3577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.086825132369995,
+      "learning_rate": 1.6730628654970763e-05,
+      "loss": 1.7765,
+      "step": 3578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9412338733673096,
+      "learning_rate": 1.6729714912280703e-05,
+      "loss": 1.4822,
+      "step": 3579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2925193309783936,
+      "learning_rate": 1.6728801169590644e-05,
+      "loss": 1.8764,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0476927757263184,
+      "learning_rate": 1.6727887426900588e-05,
+      "loss": 1.8032,
+      "step": 3581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.249727964401245,
+      "learning_rate": 1.6726973684210528e-05,
+      "loss": 1.5611,
+      "step": 3582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1861979961395264,
+      "learning_rate": 1.672605994152047e-05,
+      "loss": 1.8913,
+      "step": 3583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9637739658355713,
+      "learning_rate": 1.672514619883041e-05,
+      "loss": 1.7,
+      "step": 3584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8842854499816895,
+      "learning_rate": 1.6724232456140353e-05,
+      "loss": 1.9151,
+      "step": 3585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.62839674949646,
+      "learning_rate": 1.6723318713450294e-05,
+      "loss": 2.3201,
+      "step": 3586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4320850372314453,
+      "learning_rate": 1.6722404970760234e-05,
+      "loss": 1.6383,
+      "step": 3587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6051440238952637,
+      "learning_rate": 1.6721491228070178e-05,
+      "loss": 1.8916,
+      "step": 3588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.162785053253174,
+      "learning_rate": 1.672057748538012e-05,
+      "loss": 2.0472,
+      "step": 3589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.771122932434082,
+      "learning_rate": 1.671966374269006e-05,
+      "loss": 2.1593,
+      "step": 3590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.101457357406616,
+      "learning_rate": 1.671875e-05,
+      "loss": 1.4931,
+      "step": 3591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2335212230682373,
+      "learning_rate": 1.6717836257309944e-05,
+      "loss": 1.6719,
+      "step": 3592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3054707050323486,
+      "learning_rate": 1.6716922514619884e-05,
+      "loss": 2.0379,
+      "step": 3593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.428041696548462,
+      "learning_rate": 1.6716008771929825e-05,
+      "loss": 1.7686,
+      "step": 3594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1046195030212402,
+      "learning_rate": 1.671509502923977e-05,
+      "loss": 1.3074,
+      "step": 3595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2152254581451416,
+      "learning_rate": 1.671418128654971e-05,
+      "loss": 1.7188,
+      "step": 3596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6796042919158936,
+      "learning_rate": 1.671326754385965e-05,
+      "loss": 2.1716,
+      "step": 3597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.104811429977417,
+      "learning_rate": 1.671235380116959e-05,
+      "loss": 1.6236,
+      "step": 3598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1122636795043945,
+      "learning_rate": 1.6711440058479534e-05,
+      "loss": 1.7724,
+      "step": 3599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4896180629730225,
+      "learning_rate": 1.6710526315789475e-05,
+      "loss": 1.8035,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.623396873474121,
+      "learning_rate": 1.6709612573099415e-05,
+      "loss": 1.4883,
+      "step": 3601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7124481201171875,
+      "learning_rate": 1.670869883040936e-05,
+      "loss": 1.778,
+      "step": 3602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.366361141204834,
+      "learning_rate": 1.67077850877193e-05,
+      "loss": 1.6188,
+      "step": 3603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7659332752227783,
+      "learning_rate": 1.670687134502924e-05,
+      "loss": 1.9089,
+      "step": 3604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.667855739593506,
+      "learning_rate": 1.670595760233918e-05,
+      "loss": 2.0818,
+      "step": 3605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.133425235748291,
+      "learning_rate": 1.6705043859649125e-05,
+      "loss": 1.5244,
+      "step": 3606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3069088459014893,
+      "learning_rate": 1.6704130116959065e-05,
+      "loss": 2.2517,
+      "step": 3607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5910820960998535,
+      "learning_rate": 1.6703216374269006e-05,
+      "loss": 1.9297,
+      "step": 3608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3286290168762207,
+      "learning_rate": 1.670230263157895e-05,
+      "loss": 1.6702,
+      "step": 3609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6549205780029297,
+      "learning_rate": 1.670138888888889e-05,
+      "loss": 2.0713,
+      "step": 3610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6801042556762695,
+      "learning_rate": 1.670047514619883e-05,
+      "loss": 1.5932,
+      "step": 3611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5554704666137695,
+      "learning_rate": 1.669956140350877e-05,
+      "loss": 1.6018,
+      "step": 3612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2221155166625977,
+      "learning_rate": 1.6698647660818715e-05,
+      "loss": 2.0457,
+      "step": 3613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9340527057647705,
+      "learning_rate": 1.6697733918128656e-05,
+      "loss": 1.7551,
+      "step": 3614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.285922050476074,
+      "learning_rate": 1.6696820175438596e-05,
+      "loss": 1.87,
+      "step": 3615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2865703105926514,
+      "learning_rate": 1.669590643274854e-05,
+      "loss": 1.9079,
+      "step": 3616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6771724224090576,
+      "learning_rate": 1.669499269005848e-05,
+      "loss": 1.4126,
+      "step": 3617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6494827270507812,
+      "learning_rate": 1.669407894736842e-05,
+      "loss": 1.3341,
+      "step": 3618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5868396759033203,
+      "learning_rate": 1.669316520467836e-05,
+      "loss": 1.6545,
+      "step": 3619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.098003387451172,
+      "learning_rate": 1.6692251461988306e-05,
+      "loss": 1.4614,
+      "step": 3620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.718428134918213,
+      "learning_rate": 1.6691337719298246e-05,
+      "loss": 1.4195,
+      "step": 3621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3056743144989014,
+      "learning_rate": 1.6690423976608187e-05,
+      "loss": 1.9916,
+      "step": 3622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.301638603210449,
+      "learning_rate": 1.668951023391813e-05,
+      "loss": 2.0323,
+      "step": 3623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2196848392486572,
+      "learning_rate": 1.668859649122807e-05,
+      "loss": 2.3539,
+      "step": 3624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9050934314727783,
+      "learning_rate": 1.668768274853801e-05,
+      "loss": 1.7627,
+      "step": 3625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.998896598815918,
+      "learning_rate": 1.6686769005847952e-05,
+      "loss": 1.7136,
+      "step": 3626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9559237957000732,
+      "learning_rate": 1.6685855263157896e-05,
+      "loss": 1.6633,
+      "step": 3627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6589815616607666,
+      "learning_rate": 1.6684941520467837e-05,
+      "loss": 1.9555,
+      "step": 3628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.140634536743164,
+      "learning_rate": 1.6684027777777777e-05,
+      "loss": 1.6623,
+      "step": 3629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.064784526824951,
+      "learning_rate": 1.668311403508772e-05,
+      "loss": 1.7581,
+      "step": 3630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8002498149871826,
+      "learning_rate": 1.668220029239766e-05,
+      "loss": 1.3999,
+      "step": 3631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2208895683288574,
+      "learning_rate": 1.6681286549707602e-05,
+      "loss": 1.386,
+      "step": 3632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0847132205963135,
+      "learning_rate": 1.6680372807017543e-05,
+      "loss": 2.143,
+      "step": 3633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.913900852203369,
+      "learning_rate": 1.6679459064327487e-05,
+      "loss": 1.6649,
+      "step": 3634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6590301990509033,
+      "learning_rate": 1.667854532163743e-05,
+      "loss": 1.5081,
+      "step": 3635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.691110610961914,
+      "learning_rate": 1.667763157894737e-05,
+      "loss": 1.8313,
+      "step": 3636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5459556579589844,
+      "learning_rate": 1.667671783625731e-05,
+      "loss": 1.4632,
+      "step": 3637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.537696599960327,
+      "learning_rate": 1.6675804093567255e-05,
+      "loss": 1.7291,
+      "step": 3638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.190802812576294,
+      "learning_rate": 1.6674890350877196e-05,
+      "loss": 1.9716,
+      "step": 3639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9103877544403076,
+      "learning_rate": 1.6673976608187136e-05,
+      "loss": 1.7874,
+      "step": 3640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6950747966766357,
+      "learning_rate": 1.6673062865497077e-05,
+      "loss": 1.328,
+      "step": 3641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.668689250946045,
+      "learning_rate": 1.667214912280702e-05,
+      "loss": 1.5914,
+      "step": 3642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.842125654220581,
+      "learning_rate": 1.667123538011696e-05,
+      "loss": 1.1858,
+      "step": 3643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6375787258148193,
+      "learning_rate": 1.6670321637426902e-05,
+      "loss": 2.2492,
+      "step": 3644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.173124313354492,
+      "learning_rate": 1.6669407894736846e-05,
+      "loss": 1.9444,
+      "step": 3645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6234230995178223,
+      "learning_rate": 1.6668494152046786e-05,
+      "loss": 1.3808,
+      "step": 3646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3693716526031494,
+      "learning_rate": 1.6667580409356727e-05,
+      "loss": 1.9142,
+      "step": 3647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2048776149749756,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.0565,
+      "step": 3648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.298314094543457,
+      "learning_rate": 1.666575292397661e-05,
+      "loss": 2.1562,
+      "step": 3649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4411838054656982,
+      "learning_rate": 1.6664839181286552e-05,
+      "loss": 1.9668,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.696193218231201,
+      "learning_rate": 1.6663925438596492e-05,
+      "loss": 1.3026,
+      "step": 3651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7116036415100098,
+      "learning_rate": 1.6663011695906436e-05,
+      "loss": 1.7458,
+      "step": 3652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.452604293823242,
+      "learning_rate": 1.6662097953216377e-05,
+      "loss": 1.7476,
+      "step": 3653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3701798915863037,
+      "learning_rate": 1.6661184210526317e-05,
+      "loss": 1.837,
+      "step": 3654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.471466064453125,
+      "learning_rate": 1.6660270467836258e-05,
+      "loss": 1.6185,
+      "step": 3655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.41617488861084,
+      "learning_rate": 1.6659356725146202e-05,
+      "loss": 1.9795,
+      "step": 3656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8110084533691406,
+      "learning_rate": 1.6658442982456142e-05,
+      "loss": 1.5042,
+      "step": 3657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.836521625518799,
+      "learning_rate": 1.6657529239766083e-05,
+      "loss": 1.5854,
+      "step": 3658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7161965370178223,
+      "learning_rate": 1.6656615497076027e-05,
+      "loss": 1.2908,
+      "step": 3659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.046539068222046,
+      "learning_rate": 1.6655701754385967e-05,
+      "loss": 1.7523,
+      "step": 3660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7040903568267822,
+      "learning_rate": 1.6654788011695908e-05,
+      "loss": 1.6596,
+      "step": 3661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.848980188369751,
+      "learning_rate": 1.665387426900585e-05,
+      "loss": 1.687,
+      "step": 3662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.029707670211792,
+      "learning_rate": 1.6652960526315792e-05,
+      "loss": 1.8415,
+      "step": 3663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.932130813598633,
+      "learning_rate": 1.6652046783625733e-05,
+      "loss": 1.7343,
+      "step": 3664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.10343337059021,
+      "learning_rate": 1.6651133040935673e-05,
+      "loss": 1.7368,
+      "step": 3665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9361565113067627,
+      "learning_rate": 1.6650219298245617e-05,
+      "loss": 1.6899,
+      "step": 3666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.987217903137207,
+      "learning_rate": 1.6649305555555558e-05,
+      "loss": 2.0365,
+      "step": 3667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9053337574005127,
+      "learning_rate": 1.66483918128655e-05,
+      "loss": 1.4864,
+      "step": 3668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.402717590332031,
+      "learning_rate": 1.664747807017544e-05,
+      "loss": 1.8835,
+      "step": 3669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.181626558303833,
+      "learning_rate": 1.6646564327485383e-05,
+      "loss": 1.9096,
+      "step": 3670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6551475524902344,
+      "learning_rate": 1.6645650584795323e-05,
+      "loss": 1.5513,
+      "step": 3671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5891058444976807,
+      "learning_rate": 1.6644736842105264e-05,
+      "loss": 1.6163,
+      "step": 3672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.355778217315674,
+      "learning_rate": 1.6643823099415208e-05,
+      "loss": 2.3716,
+      "step": 3673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5685746669769287,
+      "learning_rate": 1.6642909356725148e-05,
+      "loss": 1.3995,
+      "step": 3674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0041635036468506,
+      "learning_rate": 1.664199561403509e-05,
+      "loss": 1.5936,
+      "step": 3675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5138819217681885,
+      "learning_rate": 1.664108187134503e-05,
+      "loss": 2.1655,
+      "step": 3676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.585209608078003,
+      "learning_rate": 1.6640168128654973e-05,
+      "loss": 1.5021,
+      "step": 3677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.694563150405884,
+      "learning_rate": 1.6639254385964914e-05,
+      "loss": 2.1851,
+      "step": 3678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8302409648895264,
+      "learning_rate": 1.6638340643274854e-05,
+      "loss": 1.7675,
+      "step": 3679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.479667901992798,
+      "learning_rate": 1.6637426900584798e-05,
+      "loss": 1.5401,
+      "step": 3680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.813415050506592,
+      "learning_rate": 1.663651315789474e-05,
+      "loss": 2.1378,
+      "step": 3681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.563225746154785,
+      "learning_rate": 1.663559941520468e-05,
+      "loss": 1.7893,
+      "step": 3682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7009735107421875,
+      "learning_rate": 1.663468567251462e-05,
+      "loss": 1.5922,
+      "step": 3683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.184849500656128,
+      "learning_rate": 1.6633771929824564e-05,
+      "loss": 1.6791,
+      "step": 3684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5399179458618164,
+      "learning_rate": 1.6632858187134504e-05,
+      "loss": 1.6611,
+      "step": 3685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.255326271057129,
+      "learning_rate": 1.6631944444444445e-05,
+      "loss": 1.5488,
+      "step": 3686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1848323345184326,
+      "learning_rate": 1.663103070175439e-05,
+      "loss": 1.8497,
+      "step": 3687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.338005304336548,
+      "learning_rate": 1.663011695906433e-05,
+      "loss": 2.055,
+      "step": 3688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6748366355895996,
+      "learning_rate": 1.662920321637427e-05,
+      "loss": 1.6896,
+      "step": 3689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.408811092376709,
+      "learning_rate": 1.662828947368421e-05,
+      "loss": 1.683,
+      "step": 3690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6548573970794678,
+      "learning_rate": 1.6627375730994154e-05,
+      "loss": 1.5294,
+      "step": 3691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.777796745300293,
+      "learning_rate": 1.6626461988304095e-05,
+      "loss": 1.6046,
+      "step": 3692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1584486961364746,
+      "learning_rate": 1.6625548245614035e-05,
+      "loss": 1.9349,
+      "step": 3693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8491322994232178,
+      "learning_rate": 1.662463450292398e-05,
+      "loss": 1.4551,
+      "step": 3694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4866950511932373,
+      "learning_rate": 1.662372076023392e-05,
+      "loss": 1.37,
+      "step": 3695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.191906213760376,
+      "learning_rate": 1.662280701754386e-05,
+      "loss": 1.8989,
+      "step": 3696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4264721870422363,
+      "learning_rate": 1.66218932748538e-05,
+      "loss": 2.1234,
+      "step": 3697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3420679569244385,
+      "learning_rate": 1.6620979532163745e-05,
+      "loss": 1.5151,
+      "step": 3698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.592125177383423,
+      "learning_rate": 1.6620065789473685e-05,
+      "loss": 1.8834,
+      "step": 3699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1742660999298096,
+      "learning_rate": 1.6619152046783626e-05,
+      "loss": 1.8642,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2586286067962646,
+      "learning_rate": 1.661823830409357e-05,
+      "loss": 1.8509,
+      "step": 3701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7622568607330322,
+      "learning_rate": 1.661732456140351e-05,
+      "loss": 1.4878,
+      "step": 3702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.684753656387329,
+      "learning_rate": 1.661641081871345e-05,
+      "loss": 2.0522,
+      "step": 3703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4464194774627686,
+      "learning_rate": 1.661549707602339e-05,
+      "loss": 2.0775,
+      "step": 3704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4121532440185547,
+      "learning_rate": 1.6614583333333335e-05,
+      "loss": 1.5295,
+      "step": 3705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.885061264038086,
+      "learning_rate": 1.6613669590643276e-05,
+      "loss": 1.3702,
+      "step": 3706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8799054622650146,
+      "learning_rate": 1.6612755847953216e-05,
+      "loss": 1.6171,
+      "step": 3707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.936150550842285,
+      "learning_rate": 1.661184210526316e-05,
+      "loss": 1.8819,
+      "step": 3708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.126873731613159,
+      "learning_rate": 1.66109283625731e-05,
+      "loss": 1.887,
+      "step": 3709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.836427688598633,
+      "learning_rate": 1.661001461988304e-05,
+      "loss": 1.6524,
+      "step": 3710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1146798133850098,
+      "learning_rate": 1.660910087719298e-05,
+      "loss": 1.6863,
+      "step": 3711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.623368263244629,
+      "learning_rate": 1.6608187134502926e-05,
+      "loss": 1.4838,
+      "step": 3712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7278025150299072,
+      "learning_rate": 1.6607273391812866e-05,
+      "loss": 1.8317,
+      "step": 3713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.791487455368042,
+      "learning_rate": 1.6606359649122807e-05,
+      "loss": 1.5818,
+      "step": 3714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.292532205581665,
+      "learning_rate": 1.660544590643275e-05,
+      "loss": 1.9743,
+      "step": 3715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9273245334625244,
+      "learning_rate": 1.660453216374269e-05,
+      "loss": 1.496,
+      "step": 3716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.189154863357544,
+      "learning_rate": 1.660361842105263e-05,
+      "loss": 1.7889,
+      "step": 3717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.431931257247925,
+      "learning_rate": 1.6602704678362572e-05,
+      "loss": 1.6046,
+      "step": 3718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.714312791824341,
+      "learning_rate": 1.6601790935672516e-05,
+      "loss": 1.4196,
+      "step": 3719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.219238519668579,
+      "learning_rate": 1.660087719298246e-05,
+      "loss": 1.9654,
+      "step": 3720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2777762413024902,
+      "learning_rate": 1.65999634502924e-05,
+      "loss": 1.897,
+      "step": 3721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.583930015563965,
+      "learning_rate": 1.659904970760234e-05,
+      "loss": 1.4151,
+      "step": 3722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0256094932556152,
+      "learning_rate": 1.659813596491228e-05,
+      "loss": 1.8605,
+      "step": 3723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.948554515838623,
+      "learning_rate": 1.6597222222222225e-05,
+      "loss": 1.9891,
+      "step": 3724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.514719009399414,
+      "learning_rate": 1.6596308479532166e-05,
+      "loss": 2.1567,
+      "step": 3725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8748459815979004,
+      "learning_rate": 1.6595394736842107e-05,
+      "loss": 1.7384,
+      "step": 3726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.571763277053833,
+      "learning_rate": 1.659448099415205e-05,
+      "loss": 2.2464,
+      "step": 3727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.430112600326538,
+      "learning_rate": 1.659356725146199e-05,
+      "loss": 1.8441,
+      "step": 3728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.931952953338623,
+      "learning_rate": 1.659265350877193e-05,
+      "loss": 1.6817,
+      "step": 3729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0873594284057617,
+      "learning_rate": 1.6591739766081872e-05,
+      "loss": 1.8589,
+      "step": 3730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3428432941436768,
+      "learning_rate": 1.6590826023391816e-05,
+      "loss": 1.8534,
+      "step": 3731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.099008083343506,
+      "learning_rate": 1.6589912280701756e-05,
+      "loss": 1.8248,
+      "step": 3732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3732385635375977,
+      "learning_rate": 1.6588998538011697e-05,
+      "loss": 1.3715,
+      "step": 3733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8316569328308105,
+      "learning_rate": 1.658808479532164e-05,
+      "loss": 1.826,
+      "step": 3734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8507556915283203,
+      "learning_rate": 1.658717105263158e-05,
+      "loss": 1.5841,
+      "step": 3735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9207966327667236,
+      "learning_rate": 1.6586257309941522e-05,
+      "loss": 1.5248,
+      "step": 3736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.033088445663452,
+      "learning_rate": 1.6585343567251462e-05,
+      "loss": 1.4003,
+      "step": 3737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.685591220855713,
+      "learning_rate": 1.6584429824561406e-05,
+      "loss": 1.6673,
+      "step": 3738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4380016326904297,
+      "learning_rate": 1.6583516081871347e-05,
+      "loss": 1.6847,
+      "step": 3739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.674650192260742,
+      "learning_rate": 1.6582602339181287e-05,
+      "loss": 1.4574,
+      "step": 3740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.040804147720337,
+      "learning_rate": 1.658168859649123e-05,
+      "loss": 1.8329,
+      "step": 3741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.693788528442383,
+      "learning_rate": 1.6580774853801172e-05,
+      "loss": 1.7575,
+      "step": 3742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3303229808807373,
+      "learning_rate": 1.6579861111111112e-05,
+      "loss": 1.7684,
+      "step": 3743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.725069522857666,
+      "learning_rate": 1.6578947368421053e-05,
+      "loss": 2.1403,
+      "step": 3744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3347456455230713,
+      "learning_rate": 1.6578033625730997e-05,
+      "loss": 1.9946,
+      "step": 3745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0243680477142334,
+      "learning_rate": 1.6577119883040937e-05,
+      "loss": 1.7772,
+      "step": 3746
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.894786834716797,
+      "learning_rate": 1.6576206140350878e-05,
+      "loss": 1.4076,
+      "step": 3747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8719635009765625,
+      "learning_rate": 1.6575292397660822e-05,
+      "loss": 2.0202,
+      "step": 3748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.526304006576538,
+      "learning_rate": 1.6574378654970762e-05,
+      "loss": 1.9837,
+      "step": 3749
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.947010040283203,
+      "learning_rate": 1.6573464912280703e-05,
+      "loss": 1.5775,
+      "step": 3750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.072288990020752,
+      "learning_rate": 1.6572551169590643e-05,
+      "loss": 1.9657,
+      "step": 3751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0300278663635254,
+      "learning_rate": 1.6571637426900587e-05,
+      "loss": 1.6313,
+      "step": 3752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.988417863845825,
+      "learning_rate": 1.6570723684210528e-05,
+      "loss": 1.6325,
+      "step": 3753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4537527561187744,
+      "learning_rate": 1.656980994152047e-05,
+      "loss": 1.8312,
+      "step": 3754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.23321533203125,
+      "learning_rate": 1.6568896198830412e-05,
+      "loss": 1.6427,
+      "step": 3755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.04170823097229,
+      "learning_rate": 1.6567982456140353e-05,
+      "loss": 1.7889,
+      "step": 3756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4302220344543457,
+      "learning_rate": 1.6567068713450293e-05,
+      "loss": 2.0963,
+      "step": 3757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5670900344848633,
+      "learning_rate": 1.6566154970760234e-05,
+      "loss": 2.0333,
+      "step": 3758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.447601079940796,
+      "learning_rate": 1.6565241228070178e-05,
+      "loss": 1.9815,
+      "step": 3759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4244308471679688,
+      "learning_rate": 1.656432748538012e-05,
+      "loss": 1.8958,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2991437911987305,
+      "learning_rate": 1.656341374269006e-05,
+      "loss": 1.8726,
+      "step": 3761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.154339551925659,
+      "learning_rate": 1.6562500000000003e-05,
+      "loss": 1.5112,
+      "step": 3762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5319862365722656,
+      "learning_rate": 1.6561586257309943e-05,
+      "loss": 1.9842,
+      "step": 3763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9831466674804688,
+      "learning_rate": 1.6560672514619884e-05,
+      "loss": 1.8987,
+      "step": 3764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1187572479248047,
+      "learning_rate": 1.6559758771929824e-05,
+      "loss": 1.7293,
+      "step": 3765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0376596450805664,
+      "learning_rate": 1.6558845029239768e-05,
+      "loss": 1.6824,
+      "step": 3766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.231243848800659,
+      "learning_rate": 1.655793128654971e-05,
+      "loss": 1.9192,
+      "step": 3767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2394192218780518,
+      "learning_rate": 1.655701754385965e-05,
+      "loss": 1.6448,
+      "step": 3768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7921855449676514,
+      "learning_rate": 1.6556103801169593e-05,
+      "loss": 1.8304,
+      "step": 3769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.53842830657959,
+      "learning_rate": 1.6555190058479534e-05,
+      "loss": 1.8538,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0580294132232666,
+      "learning_rate": 1.6554276315789474e-05,
+      "loss": 1.572,
+      "step": 3771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0375187397003174,
+      "learning_rate": 1.6553362573099415e-05,
+      "loss": 1.6694,
+      "step": 3772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8647522926330566,
+      "learning_rate": 1.655244883040936e-05,
+      "loss": 1.8914,
+      "step": 3773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5000548362731934,
+      "learning_rate": 1.65515350877193e-05,
+      "loss": 1.2859,
+      "step": 3774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.767843723297119,
+      "learning_rate": 1.655062134502924e-05,
+      "loss": 1.896,
+      "step": 3775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4927279949188232,
+      "learning_rate": 1.6549707602339184e-05,
+      "loss": 1.3435,
+      "step": 3776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.92207407951355,
+      "learning_rate": 1.6548793859649124e-05,
+      "loss": 1.5818,
+      "step": 3777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.742069959640503,
+      "learning_rate": 1.6547880116959065e-05,
+      "loss": 1.6305,
+      "step": 3778
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.27388072013855,
+      "learning_rate": 1.6546966374269005e-05,
+      "loss": 1.9485,
+      "step": 3779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7831928730010986,
+      "learning_rate": 1.654605263157895e-05,
+      "loss": 1.5754,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.443812608718872,
+      "learning_rate": 1.654513888888889e-05,
+      "loss": 1.4054,
+      "step": 3781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7035839557647705,
+      "learning_rate": 1.654422514619883e-05,
+      "loss": 1.696,
+      "step": 3782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.530878782272339,
+      "learning_rate": 1.6543311403508774e-05,
+      "loss": 1.3659,
+      "step": 3783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.206550359725952,
+      "learning_rate": 1.6542397660818715e-05,
+      "loss": 1.6141,
+      "step": 3784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.386547088623047,
+      "learning_rate": 1.6541483918128655e-05,
+      "loss": 1.7424,
+      "step": 3785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.625558614730835,
+      "learning_rate": 1.6540570175438596e-05,
+      "loss": 1.6711,
+      "step": 3786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1213300228118896,
+      "learning_rate": 1.653965643274854e-05,
+      "loss": 1.6699,
+      "step": 3787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2415244579315186,
+      "learning_rate": 1.653874269005848e-05,
+      "loss": 2.2169,
+      "step": 3788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.004490613937378,
+      "learning_rate": 1.653782894736842e-05,
+      "loss": 1.6333,
+      "step": 3789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.212613582611084,
+      "learning_rate": 1.6536915204678365e-05,
+      "loss": 1.8679,
+      "step": 3790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.742924451828003,
+      "learning_rate": 1.6536001461988305e-05,
+      "loss": 1.4301,
+      "step": 3791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3634891510009766,
+      "learning_rate": 1.6535087719298246e-05,
+      "loss": 1.6977,
+      "step": 3792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1874499320983887,
+      "learning_rate": 1.6534173976608186e-05,
+      "loss": 1.7032,
+      "step": 3793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.952381134033203,
+      "learning_rate": 1.653326023391813e-05,
+      "loss": 1.8168,
+      "step": 3794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8657636642456055,
+      "learning_rate": 1.653234649122807e-05,
+      "loss": 1.9881,
+      "step": 3795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.84228515625,
+      "learning_rate": 1.653143274853801e-05,
+      "loss": 1.538,
+      "step": 3796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0692837238311768,
+      "learning_rate": 1.6530519005847955e-05,
+      "loss": 1.3886,
+      "step": 3797
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0694808959960938,
+      "learning_rate": 1.6529605263157896e-05,
+      "loss": 1.7665,
+      "step": 3798
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.971748113632202,
+      "learning_rate": 1.6528691520467836e-05,
+      "loss": 1.3187,
+      "step": 3799
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.284050941467285,
+      "learning_rate": 1.6527777777777777e-05,
+      "loss": 1.7003,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0519192218780518,
+      "learning_rate": 1.652686403508772e-05,
+      "loss": 1.6697,
+      "step": 3801
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.115978956222534,
+      "learning_rate": 1.652595029239766e-05,
+      "loss": 1.7903,
+      "step": 3802
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0388832092285156,
+      "learning_rate": 1.65250365497076e-05,
+      "loss": 1.7746,
+      "step": 3803
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0549447536468506,
+      "learning_rate": 1.6524122807017546e-05,
+      "loss": 1.9657,
+      "step": 3804
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0895678997039795,
+      "learning_rate": 1.652320906432749e-05,
+      "loss": 1.8701,
+      "step": 3805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1406004428863525,
+      "learning_rate": 1.652229532163743e-05,
+      "loss": 1.8656,
+      "step": 3806
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7978477478027344,
+      "learning_rate": 1.652138157894737e-05,
+      "loss": 1.7814,
+      "step": 3807
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3099265098571777,
+      "learning_rate": 1.652046783625731e-05,
+      "loss": 2.0625,
+      "step": 3808
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.905871629714966,
+      "learning_rate": 1.6519554093567255e-05,
+      "loss": 1.6542,
+      "step": 3809
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.924025535583496,
+      "learning_rate": 1.6518640350877196e-05,
+      "loss": 1.7663,
+      "step": 3810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.718032121658325,
+      "learning_rate": 1.6517726608187136e-05,
+      "loss": 1.8735,
+      "step": 3811
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8460655212402344,
+      "learning_rate": 1.651681286549708e-05,
+      "loss": 1.4876,
+      "step": 3812
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.020333766937256,
+      "learning_rate": 1.651589912280702e-05,
+      "loss": 1.661,
+      "step": 3813
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2105624675750732,
+      "learning_rate": 1.651498538011696e-05,
+      "loss": 2.0053,
+      "step": 3814
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0397732257843018,
+      "learning_rate": 1.65140716374269e-05,
+      "loss": 1.9327,
+      "step": 3815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.352976083755493,
+      "learning_rate": 1.6513157894736845e-05,
+      "loss": 1.8292,
+      "step": 3816
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0798728466033936,
+      "learning_rate": 1.6512244152046786e-05,
+      "loss": 1.6922,
+      "step": 3817
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.636745929718018,
+      "learning_rate": 1.6511330409356727e-05,
+      "loss": 1.7111,
+      "step": 3818
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3082549571990967,
+      "learning_rate": 1.651041666666667e-05,
+      "loss": 1.7318,
+      "step": 3819
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.661151647567749,
+      "learning_rate": 1.650950292397661e-05,
+      "loss": 1.5092,
+      "step": 3820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.810335159301758,
+      "learning_rate": 1.650858918128655e-05,
+      "loss": 1.4577,
+      "step": 3821
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9017364978790283,
+      "learning_rate": 1.6507675438596492e-05,
+      "loss": 1.4018,
+      "step": 3822
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.495617389678955,
+      "learning_rate": 1.6506761695906436e-05,
+      "loss": 1.434,
+      "step": 3823
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5260114669799805,
+      "learning_rate": 1.6505847953216376e-05,
+      "loss": 1.9825,
+      "step": 3824
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9209043979644775,
+      "learning_rate": 1.6504934210526317e-05,
+      "loss": 1.6121,
+      "step": 3825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.110452175140381,
+      "learning_rate": 1.650402046783626e-05,
+      "loss": 1.6126,
+      "step": 3826
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3086087703704834,
+      "learning_rate": 1.65031067251462e-05,
+      "loss": 1.783,
+      "step": 3827
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.66284441947937,
+      "learning_rate": 1.6502192982456142e-05,
+      "loss": 1.2762,
+      "step": 3828
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.952014923095703,
+      "learning_rate": 1.6501279239766082e-05,
+      "loss": 1.6664,
+      "step": 3829
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6097772121429443,
+      "learning_rate": 1.6500365497076026e-05,
+      "loss": 1.4867,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5308499336242676,
+      "learning_rate": 1.6499451754385967e-05,
+      "loss": 1.2838,
+      "step": 3831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3709096908569336,
+      "learning_rate": 1.6498538011695907e-05,
+      "loss": 1.3173,
+      "step": 3832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.965442657470703,
+      "learning_rate": 1.6497624269005848e-05,
+      "loss": 1.7492,
+      "step": 3833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5540988445281982,
+      "learning_rate": 1.6496710526315792e-05,
+      "loss": 1.5222,
+      "step": 3834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0107924938201904,
+      "learning_rate": 1.6495796783625732e-05,
+      "loss": 1.6728,
+      "step": 3835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1836869716644287,
+      "learning_rate": 1.6494883040935673e-05,
+      "loss": 1.894,
+      "step": 3836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.756305456161499,
+      "learning_rate": 1.6493969298245617e-05,
+      "loss": 2.0455,
+      "step": 3837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.057806968688965,
+      "learning_rate": 1.6493055555555557e-05,
+      "loss": 1.6368,
+      "step": 3838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0231034755706787,
+      "learning_rate": 1.6492141812865498e-05,
+      "loss": 1.5896,
+      "step": 3839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9261550903320312,
+      "learning_rate": 1.649122807017544e-05,
+      "loss": 1.8568,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.079324960708618,
+      "learning_rate": 1.6490314327485382e-05,
+      "loss": 1.6076,
+      "step": 3841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3915469646453857,
+      "learning_rate": 1.6489400584795323e-05,
+      "loss": 2.059,
+      "step": 3842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.496318817138672,
+      "learning_rate": 1.6488486842105263e-05,
+      "loss": 1.3122,
+      "step": 3843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.195962429046631,
+      "learning_rate": 1.6487573099415207e-05,
+      "loss": 1.8429,
+      "step": 3844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.561332941055298,
+      "learning_rate": 1.6486659356725148e-05,
+      "loss": 1.5896,
+      "step": 3845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8939430713653564,
+      "learning_rate": 1.648574561403509e-05,
+      "loss": 1.5057,
+      "step": 3846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1912758350372314,
+      "learning_rate": 1.648483187134503e-05,
+      "loss": 1.6714,
+      "step": 3847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9207351207733154,
+      "learning_rate": 1.6483918128654973e-05,
+      "loss": 1.8636,
+      "step": 3848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0170698165893555,
+      "learning_rate": 1.6483004385964913e-05,
+      "loss": 1.731,
+      "step": 3849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.229871988296509,
+      "learning_rate": 1.6482090643274854e-05,
+      "loss": 1.7664,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.807795524597168,
+      "learning_rate": 1.6481176900584798e-05,
+      "loss": 1.3393,
+      "step": 3851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0275444984436035,
+      "learning_rate": 1.648026315789474e-05,
+      "loss": 1.6407,
+      "step": 3852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3311569690704346,
+      "learning_rate": 1.647934941520468e-05,
+      "loss": 1.3402,
+      "step": 3853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7524099349975586,
+      "learning_rate": 1.647843567251462e-05,
+      "loss": 1.7044,
+      "step": 3854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.033252239227295,
+      "learning_rate": 1.6477521929824563e-05,
+      "loss": 1.5172,
+      "step": 3855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.1672139167785645,
+      "learning_rate": 1.6476608187134504e-05,
+      "loss": 1.5281,
+      "step": 3856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.139703750610352,
+      "learning_rate": 1.6475694444444444e-05,
+      "loss": 2.2138,
+      "step": 3857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.629728317260742,
+      "learning_rate": 1.6474780701754388e-05,
+      "loss": 2.1667,
+      "step": 3858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.852851390838623,
+      "learning_rate": 1.647386695906433e-05,
+      "loss": 1.5247,
+      "step": 3859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3398146629333496,
+      "learning_rate": 1.647295321637427e-05,
+      "loss": 1.9646,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8911235332489014,
+      "learning_rate": 1.647203947368421e-05,
+      "loss": 1.7682,
+      "step": 3861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9582884311676025,
+      "learning_rate": 1.6471125730994154e-05,
+      "loss": 1.8328,
+      "step": 3862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.974587917327881,
+      "learning_rate": 1.6470211988304094e-05,
+      "loss": 1.8543,
+      "step": 3863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1954965591430664,
+      "learning_rate": 1.6469298245614035e-05,
+      "loss": 1.6249,
+      "step": 3864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.478530168533325,
+      "learning_rate": 1.646838450292398e-05,
+      "loss": 1.6056,
+      "step": 3865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7962582111358643,
+      "learning_rate": 1.646747076023392e-05,
+      "loss": 1.5943,
+      "step": 3866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6408352851867676,
+      "learning_rate": 1.646655701754386e-05,
+      "loss": 1.57,
+      "step": 3867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.40956974029541,
+      "learning_rate": 1.64656432748538e-05,
+      "loss": 1.7882,
+      "step": 3868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.969647169113159,
+      "learning_rate": 1.6464729532163744e-05,
+      "loss": 1.7312,
+      "step": 3869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3948564529418945,
+      "learning_rate": 1.6463815789473685e-05,
+      "loss": 2.0972,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9310998916625977,
+      "learning_rate": 1.6462902046783625e-05,
+      "loss": 1.7416,
+      "step": 3871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0633184909820557,
+      "learning_rate": 1.646198830409357e-05,
+      "loss": 1.7391,
+      "step": 3872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5562901496887207,
+      "learning_rate": 1.646107456140351e-05,
+      "loss": 1.5341,
+      "step": 3873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.071254014968872,
+      "learning_rate": 1.646016081871345e-05,
+      "loss": 1.8377,
+      "step": 3874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.010418176651001,
+      "learning_rate": 1.645924707602339e-05,
+      "loss": 1.5827,
+      "step": 3875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6702494621276855,
+      "learning_rate": 1.6458333333333335e-05,
+      "loss": 2.0062,
+      "step": 3876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6637842655181885,
+      "learning_rate": 1.6457419590643275e-05,
+      "loss": 1.2219,
+      "step": 3877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.658235788345337,
+      "learning_rate": 1.6456505847953216e-05,
+      "loss": 2.266,
+      "step": 3878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.579066038131714,
+      "learning_rate": 1.645559210526316e-05,
+      "loss": 1.4662,
+      "step": 3879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.010033130645752,
+      "learning_rate": 1.64546783625731e-05,
+      "loss": 1.9394,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.026294469833374,
+      "learning_rate": 1.645376461988304e-05,
+      "loss": 1.6202,
+      "step": 3881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.702784299850464,
+      "learning_rate": 1.645285087719298e-05,
+      "loss": 1.9806,
+      "step": 3882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.130213737487793,
+      "learning_rate": 1.6451937134502925e-05,
+      "loss": 1.93,
+      "step": 3883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.197869300842285,
+      "learning_rate": 1.6451023391812866e-05,
+      "loss": 1.7105,
+      "step": 3884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.254840850830078,
+      "learning_rate": 1.6450109649122806e-05,
+      "loss": 1.7659,
+      "step": 3885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.426788568496704,
+      "learning_rate": 1.644919590643275e-05,
+      "loss": 1.5675,
+      "step": 3886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1759684085845947,
+      "learning_rate": 1.644828216374269e-05,
+      "loss": 1.5309,
+      "step": 3887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0291597843170166,
+      "learning_rate": 1.644736842105263e-05,
+      "loss": 1.8803,
+      "step": 3888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.856281280517578,
+      "learning_rate": 1.6446454678362572e-05,
+      "loss": 1.8578,
+      "step": 3889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8259832859039307,
+      "learning_rate": 1.6445540935672516e-05,
+      "loss": 1.4115,
+      "step": 3890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7622053623199463,
+      "learning_rate": 1.644462719298246e-05,
+      "loss": 1.5073,
+      "step": 3891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8830952644348145,
+      "learning_rate": 1.64437134502924e-05,
+      "loss": 1.6588,
+      "step": 3892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8154797554016113,
+      "learning_rate": 1.644279970760234e-05,
+      "loss": 1.6949,
+      "step": 3893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.820927381515503,
+      "learning_rate": 1.6441885964912285e-05,
+      "loss": 1.6694,
+      "step": 3894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5737764835357666,
+      "learning_rate": 1.6440972222222225e-05,
+      "loss": 1.5167,
+      "step": 3895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.781960964202881,
+      "learning_rate": 1.6440058479532166e-05,
+      "loss": 1.6588,
+      "step": 3896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.415581703186035,
+      "learning_rate": 1.6439144736842106e-05,
+      "loss": 2.1242,
+      "step": 3897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1401429176330566,
+      "learning_rate": 1.643823099415205e-05,
+      "loss": 1.6586,
+      "step": 3898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2249112129211426,
+      "learning_rate": 1.643731725146199e-05,
+      "loss": 1.7922,
+      "step": 3899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8172953128814697,
+      "learning_rate": 1.643640350877193e-05,
+      "loss": 1.6993,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0006604194641113,
+      "learning_rate": 1.6435489766081875e-05,
+      "loss": 1.5826,
+      "step": 3901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2451682090759277,
+      "learning_rate": 1.6434576023391816e-05,
+      "loss": 1.8272,
+      "step": 3902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0343122482299805,
+      "learning_rate": 1.6433662280701756e-05,
+      "loss": 1.9377,
+      "step": 3903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.121546506881714,
+      "learning_rate": 1.6432748538011697e-05,
+      "loss": 1.7,
+      "step": 3904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.914275646209717,
+      "learning_rate": 1.643183479532164e-05,
+      "loss": 1.516,
+      "step": 3905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.122333526611328,
+      "learning_rate": 1.643092105263158e-05,
+      "loss": 1.7567,
+      "step": 3906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.510927438735962,
+      "learning_rate": 1.643000730994152e-05,
+      "loss": 1.345,
+      "step": 3907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6246097087860107,
+      "learning_rate": 1.6429093567251465e-05,
+      "loss": 1.1166,
+      "step": 3908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.062124490737915,
+      "learning_rate": 1.6428179824561406e-05,
+      "loss": 2.0478,
+      "step": 3909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9422295093536377,
+      "learning_rate": 1.6427266081871347e-05,
+      "loss": 1.5539,
+      "step": 3910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6746373176574707,
+      "learning_rate": 1.6426352339181287e-05,
+      "loss": 1.7291,
+      "step": 3911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6202552318573,
+      "learning_rate": 1.642543859649123e-05,
+      "loss": 1.6057,
+      "step": 3912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4881837368011475,
+      "learning_rate": 1.642452485380117e-05,
+      "loss": 1.5818,
+      "step": 3913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.534904956817627,
+      "learning_rate": 1.6423611111111112e-05,
+      "loss": 1.5503,
+      "step": 3914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7113165855407715,
+      "learning_rate": 1.6422697368421056e-05,
+      "loss": 1.5275,
+      "step": 3915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7890689373016357,
+      "learning_rate": 1.6421783625730996e-05,
+      "loss": 1.484,
+      "step": 3916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1081788539886475,
+      "learning_rate": 1.6420869883040937e-05,
+      "loss": 1.6565,
+      "step": 3917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1978859901428223,
+      "learning_rate": 1.6419956140350878e-05,
+      "loss": 1.7158,
+      "step": 3918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.951540231704712,
+      "learning_rate": 1.641904239766082e-05,
+      "loss": 1.829,
+      "step": 3919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2218446731567383,
+      "learning_rate": 1.6418128654970762e-05,
+      "loss": 1.6931,
+      "step": 3920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.736133098602295,
+      "learning_rate": 1.6417214912280702e-05,
+      "loss": 1.3857,
+      "step": 3921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.328826427459717,
+      "learning_rate": 1.6416301169590646e-05,
+      "loss": 1.6579,
+      "step": 3922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5103578567504883,
+      "learning_rate": 1.6415387426900587e-05,
+      "loss": 1.5896,
+      "step": 3923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.743089437484741,
+      "learning_rate": 1.6414473684210527e-05,
+      "loss": 1.7234,
+      "step": 3924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8773245811462402,
+      "learning_rate": 1.6413559941520468e-05,
+      "loss": 1.6785,
+      "step": 3925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0223045349121094,
+      "learning_rate": 1.6412646198830412e-05,
+      "loss": 1.823,
+      "step": 3926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.510434865951538,
+      "learning_rate": 1.6411732456140352e-05,
+      "loss": 1.9407,
+      "step": 3927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.952934503555298,
+      "learning_rate": 1.6410818713450293e-05,
+      "loss": 1.5546,
+      "step": 3928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.957951068878174,
+      "learning_rate": 1.6409904970760237e-05,
+      "loss": 1.9967,
+      "step": 3929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.568525791168213,
+      "learning_rate": 1.6408991228070177e-05,
+      "loss": 1.3855,
+      "step": 3930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0203142166137695,
+      "learning_rate": 1.6408077485380118e-05,
+      "loss": 1.3891,
+      "step": 3931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9607369899749756,
+      "learning_rate": 1.640716374269006e-05,
+      "loss": 1.6914,
+      "step": 3932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.716486692428589,
+      "learning_rate": 1.6406250000000002e-05,
+      "loss": 1.47,
+      "step": 3933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.008246898651123,
+      "learning_rate": 1.6405336257309943e-05,
+      "loss": 1.6498,
+      "step": 3934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7143523693084717,
+      "learning_rate": 1.6404422514619883e-05,
+      "loss": 1.8828,
+      "step": 3935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9335997104644775,
+      "learning_rate": 1.6403508771929827e-05,
+      "loss": 1.485,
+      "step": 3936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.974940538406372,
+      "learning_rate": 1.6402595029239768e-05,
+      "loss": 1.8227,
+      "step": 3937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6042585372924805,
+      "learning_rate": 1.640168128654971e-05,
+      "loss": 2.3862,
+      "step": 3938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.031363010406494,
+      "learning_rate": 1.640076754385965e-05,
+      "loss": 1.7647,
+      "step": 3939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.020805358886719,
+      "learning_rate": 1.6399853801169593e-05,
+      "loss": 1.656,
+      "step": 3940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1077985763549805,
+      "learning_rate": 1.6398940058479533e-05,
+      "loss": 1.9798,
+      "step": 3941
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8838119506835938,
+      "learning_rate": 1.6398026315789474e-05,
+      "loss": 1.6269,
+      "step": 3942
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9521737098693848,
+      "learning_rate": 1.6397112573099418e-05,
+      "loss": 1.6845,
+      "step": 3943
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5274293422698975,
+      "learning_rate": 1.639619883040936e-05,
+      "loss": 1.6101,
+      "step": 3944
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6735925674438477,
+      "learning_rate": 1.63952850877193e-05,
+      "loss": 1.5667,
+      "step": 3945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.982628345489502,
+      "learning_rate": 1.639437134502924e-05,
+      "loss": 1.91,
+      "step": 3946
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.099775552749634,
+      "learning_rate": 1.6393457602339183e-05,
+      "loss": 1.9292,
+      "step": 3947
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.205028533935547,
+      "learning_rate": 1.6392543859649124e-05,
+      "loss": 1.6039,
+      "step": 3948
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.006326675415039,
+      "learning_rate": 1.6391630116959064e-05,
+      "loss": 1.46,
+      "step": 3949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9421205520629883,
+      "learning_rate": 1.6390716374269008e-05,
+      "loss": 1.5546,
+      "step": 3950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.05517315864563,
+      "learning_rate": 1.638980263157895e-05,
+      "loss": 1.8812,
+      "step": 3951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4527111053466797,
+      "learning_rate": 1.638888888888889e-05,
+      "loss": 1.9486,
+      "step": 3952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8779995441436768,
+      "learning_rate": 1.638797514619883e-05,
+      "loss": 1.2961,
+      "step": 3953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.375159740447998,
+      "learning_rate": 1.6387061403508774e-05,
+      "loss": 2.0493,
+      "step": 3954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9737260341644287,
+      "learning_rate": 1.6386147660818714e-05,
+      "loss": 1.7819,
+      "step": 3955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3721911907196045,
+      "learning_rate": 1.6385233918128655e-05,
+      "loss": 1.6776,
+      "step": 3956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.071023464202881,
+      "learning_rate": 1.63843201754386e-05,
+      "loss": 1.8271,
+      "step": 3957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9032607078552246,
+      "learning_rate": 1.638340643274854e-05,
+      "loss": 1.2771,
+      "step": 3958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9900426864624023,
+      "learning_rate": 1.638249269005848e-05,
+      "loss": 1.4168,
+      "step": 3959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.598111152648926,
+      "learning_rate": 1.638157894736842e-05,
+      "loss": 1.522,
+      "step": 3960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9270477294921875,
+      "learning_rate": 1.6380665204678364e-05,
+      "loss": 1.8215,
+      "step": 3961
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.087198257446289,
+      "learning_rate": 1.6379751461988305e-05,
+      "loss": 1.9197,
+      "step": 3962
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8734347820281982,
+      "learning_rate": 1.6378837719298245e-05,
+      "loss": 1.837,
+      "step": 3963
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7145590782165527,
+      "learning_rate": 1.637792397660819e-05,
+      "loss": 1.7211,
+      "step": 3964
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.083425760269165,
+      "learning_rate": 1.637701023391813e-05,
+      "loss": 2.0605,
+      "step": 3965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3033525943756104,
+      "learning_rate": 1.637609649122807e-05,
+      "loss": 1.517,
+      "step": 3966
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7799179553985596,
+      "learning_rate": 1.637518274853801e-05,
+      "loss": 1.7405,
+      "step": 3967
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.376291275024414,
+      "learning_rate": 1.6374269005847955e-05,
+      "loss": 1.4936,
+      "step": 3968
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6408803462982178,
+      "learning_rate": 1.6373355263157895e-05,
+      "loss": 1.478,
+      "step": 3969
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.857861042022705,
+      "learning_rate": 1.6372441520467836e-05,
+      "loss": 1.9648,
+      "step": 3970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.606172800064087,
+      "learning_rate": 1.637152777777778e-05,
+      "loss": 2.2512,
+      "step": 3971
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.817021131515503,
+      "learning_rate": 1.637061403508772e-05,
+      "loss": 1.7407,
+      "step": 3972
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8968005180358887,
+      "learning_rate": 1.636970029239766e-05,
+      "loss": 1.5899,
+      "step": 3973
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7264299392700195,
+      "learning_rate": 1.63687865497076e-05,
+      "loss": 1.4179,
+      "step": 3974
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2036805152893066,
+      "learning_rate": 1.6367872807017545e-05,
+      "loss": 1.8851,
+      "step": 3975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.098611831665039,
+      "learning_rate": 1.636695906432749e-05,
+      "loss": 1.7893,
+      "step": 3976
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7283692359924316,
+      "learning_rate": 1.636604532163743e-05,
+      "loss": 1.6999,
+      "step": 3977
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1775636672973633,
+      "learning_rate": 1.636513157894737e-05,
+      "loss": 1.9744,
+      "step": 3978
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5343661308288574,
+      "learning_rate": 1.636421783625731e-05,
+      "loss": 1.9347,
+      "step": 3979
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2587730884552,
+      "learning_rate": 1.6363304093567255e-05,
+      "loss": 2.0646,
+      "step": 3980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3648669719696045,
+      "learning_rate": 1.6362390350877195e-05,
+      "loss": 1.8724,
+      "step": 3981
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.621948719024658,
+      "learning_rate": 1.6361476608187136e-05,
+      "loss": 1.4926,
+      "step": 3982
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3696653842926025,
+      "learning_rate": 1.636056286549708e-05,
+      "loss": 1.3092,
+      "step": 3983
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1545822620391846,
+      "learning_rate": 1.635964912280702e-05,
+      "loss": 1.739,
+      "step": 3984
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0960662364959717,
+      "learning_rate": 1.635873538011696e-05,
+      "loss": 1.1805,
+      "step": 3985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.694786787033081,
+      "learning_rate": 1.63578216374269e-05,
+      "loss": 1.9414,
+      "step": 3986
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0517449378967285,
+      "learning_rate": 1.6356907894736845e-05,
+      "loss": 1.5648,
+      "step": 3987
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5165162086486816,
+      "learning_rate": 1.6355994152046786e-05,
+      "loss": 2.3991,
+      "step": 3988
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1600189208984375,
+      "learning_rate": 1.6355080409356726e-05,
+      "loss": 1.6721,
+      "step": 3989
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1732468605041504,
+      "learning_rate": 1.635416666666667e-05,
+      "loss": 2.0849,
+      "step": 3990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.840690851211548,
+      "learning_rate": 1.635325292397661e-05,
+      "loss": 1.7356,
+      "step": 3991
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8710827827453613,
+      "learning_rate": 1.635233918128655e-05,
+      "loss": 1.8315,
+      "step": 3992
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1360585689544678,
+      "learning_rate": 1.635142543859649e-05,
+      "loss": 1.6475,
+      "step": 3993
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5908827781677246,
+      "learning_rate": 1.6350511695906436e-05,
+      "loss": 1.5038,
+      "step": 3994
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.655179977416992,
+      "learning_rate": 1.6349597953216376e-05,
+      "loss": 2.0771,
+      "step": 3995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.160874843597412,
+      "learning_rate": 1.6348684210526317e-05,
+      "loss": 1.7396,
+      "step": 3996
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.698486089706421,
+      "learning_rate": 1.634777046783626e-05,
+      "loss": 1.3668,
+      "step": 3997
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2504348754882812,
+      "learning_rate": 1.63468567251462e-05,
+      "loss": 1.6213,
+      "step": 3998
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.077946424484253,
+      "learning_rate": 1.634594298245614e-05,
+      "loss": 1.5412,
+      "step": 3999
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.222346544265747,
+      "learning_rate": 1.6345029239766082e-05,
+      "loss": 1.7952,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1605608463287354,
+      "learning_rate": 1.6344115497076026e-05,
+      "loss": 1.9423,
+      "step": 4001
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2045586109161377,
+      "learning_rate": 1.6343201754385967e-05,
+      "loss": 2.0595,
+      "step": 4002
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.746356725692749,
+      "learning_rate": 1.6342288011695907e-05,
+      "loss": 1.5772,
+      "step": 4003
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1444520950317383,
+      "learning_rate": 1.634137426900585e-05,
+      "loss": 1.9445,
+      "step": 4004
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.762505531311035,
+      "learning_rate": 1.634046052631579e-05,
+      "loss": 1.5362,
+      "step": 4005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4220943450927734,
+      "learning_rate": 1.6339546783625732e-05,
+      "loss": 1.5502,
+      "step": 4006
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0801916122436523,
+      "learning_rate": 1.6338633040935673e-05,
+      "loss": 2.0018,
+      "step": 4007
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.222203493118286,
+      "learning_rate": 1.6337719298245616e-05,
+      "loss": 1.9738,
+      "step": 4008
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3896596431732178,
+      "learning_rate": 1.6336805555555557e-05,
+      "loss": 1.7578,
+      "step": 4009
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3203907012939453,
+      "learning_rate": 1.6335891812865498e-05,
+      "loss": 1.9823,
+      "step": 4010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.760730028152466,
+      "learning_rate": 1.633497807017544e-05,
+      "loss": 1.6003,
+      "step": 4011
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.042837619781494,
+      "learning_rate": 1.6334064327485382e-05,
+      "loss": 1.58,
+      "step": 4012
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6634955406188965,
+      "learning_rate": 1.6333150584795322e-05,
+      "loss": 1.594,
+      "step": 4013
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3832290172576904,
+      "learning_rate": 1.6332236842105263e-05,
+      "loss": 1.2887,
+      "step": 4014
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1341488361358643,
+      "learning_rate": 1.6331323099415207e-05,
+      "loss": 1.6132,
+      "step": 4015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.558040142059326,
+      "learning_rate": 1.6330409356725147e-05,
+      "loss": 1.7684,
+      "step": 4016
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3302016258239746,
+      "learning_rate": 1.6329495614035088e-05,
+      "loss": 1.9855,
+      "step": 4017
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.545631170272827,
+      "learning_rate": 1.6328581871345032e-05,
+      "loss": 1.3697,
+      "step": 4018
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6917271614074707,
+      "learning_rate": 1.6327668128654972e-05,
+      "loss": 1.5313,
+      "step": 4019
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6739022731781006,
+      "learning_rate": 1.6326754385964913e-05,
+      "loss": 1.6097,
+      "step": 4020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.977407693862915,
+      "learning_rate": 1.6325840643274853e-05,
+      "loss": 2.3101,
+      "step": 4021
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.548057794570923,
+      "learning_rate": 1.6324926900584797e-05,
+      "loss": 1.358,
+      "step": 4022
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0972771644592285,
+      "learning_rate": 1.6324013157894738e-05,
+      "loss": 1.5973,
+      "step": 4023
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.147779703140259,
+      "learning_rate": 1.632309941520468e-05,
+      "loss": 1.8269,
+      "step": 4024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0165915489196777,
+      "learning_rate": 1.6322185672514622e-05,
+      "loss": 1.7834,
+      "step": 4025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9742841720581055,
+      "learning_rate": 1.6321271929824563e-05,
+      "loss": 1.6799,
+      "step": 4026
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.496840000152588,
+      "learning_rate": 1.6320358187134503e-05,
+      "loss": 1.8141,
+      "step": 4027
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1671783924102783,
+      "learning_rate": 1.6319444444444444e-05,
+      "loss": 1.663,
+      "step": 4028
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1571552753448486,
+      "learning_rate": 1.6318530701754388e-05,
+      "loss": 1.4461,
+      "step": 4029
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0211942195892334,
+      "learning_rate": 1.631761695906433e-05,
+      "loss": 1.3312,
+      "step": 4030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.730778455734253,
+      "learning_rate": 1.631670321637427e-05,
+      "loss": 1.6561,
+      "step": 4031
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.06875467300415,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 2.4779,
+      "step": 4032
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1854660511016846,
+      "learning_rate": 1.6314875730994153e-05,
+      "loss": 1.8646,
+      "step": 4033
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.479851007461548,
+      "learning_rate": 1.6313961988304094e-05,
+      "loss": 1.1314,
+      "step": 4034
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9249868392944336,
+      "learning_rate": 1.6313048245614034e-05,
+      "loss": 1.5823,
+      "step": 4035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8777265548706055,
+      "learning_rate": 1.631213450292398e-05,
+      "loss": 1.6541,
+      "step": 4036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1822283267974854,
+      "learning_rate": 1.631122076023392e-05,
+      "loss": 1.3729,
+      "step": 4037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.819002389907837,
+      "learning_rate": 1.631030701754386e-05,
+      "loss": 1.6976,
+      "step": 4038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9309651851654053,
+      "learning_rate": 1.6309393274853803e-05,
+      "loss": 1.339,
+      "step": 4039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.767958402633667,
+      "learning_rate": 1.6308479532163744e-05,
+      "loss": 1.47,
+      "step": 4040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.22359037399292,
+      "learning_rate": 1.6307565789473684e-05,
+      "loss": 1.6241,
+      "step": 4041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.178849697113037,
+      "learning_rate": 1.6306652046783625e-05,
+      "loss": 2.1001,
+      "step": 4042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8723363876342773,
+      "learning_rate": 1.630573830409357e-05,
+      "loss": 1.6504,
+      "step": 4043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.266660690307617,
+      "learning_rate": 1.630482456140351e-05,
+      "loss": 2.0233,
+      "step": 4044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.156106472015381,
+      "learning_rate": 1.630391081871345e-05,
+      "loss": 1.5961,
+      "step": 4045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8053317070007324,
+      "learning_rate": 1.6302997076023394e-05,
+      "loss": 1.6657,
+      "step": 4046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.873465061187744,
+      "learning_rate": 1.6302083333333334e-05,
+      "loss": 1.7111,
+      "step": 4047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.931736946105957,
+      "learning_rate": 1.6301169590643275e-05,
+      "loss": 1.8436,
+      "step": 4048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.103744029998779,
+      "learning_rate": 1.6300255847953215e-05,
+      "loss": 1.8943,
+      "step": 4049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2806475162506104,
+      "learning_rate": 1.629934210526316e-05,
+      "loss": 1.7761,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7905056476593018,
+      "learning_rate": 1.62984283625731e-05,
+      "loss": 1.4052,
+      "step": 4051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1620333194732666,
+      "learning_rate": 1.629751461988304e-05,
+      "loss": 1.8249,
+      "step": 4052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.321428060531616,
+      "learning_rate": 1.6296600877192984e-05,
+      "loss": 1.3108,
+      "step": 4053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.380398750305176,
+      "learning_rate": 1.6295687134502925e-05,
+      "loss": 1.8635,
+      "step": 4054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0229480266571045,
+      "learning_rate": 1.6294773391812865e-05,
+      "loss": 1.6699,
+      "step": 4055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0112664699554443,
+      "learning_rate": 1.6293859649122806e-05,
+      "loss": 1.9,
+      "step": 4056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.895920753479004,
+      "learning_rate": 1.629294590643275e-05,
+      "loss": 2.0744,
+      "step": 4057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6484806537628174,
+      "learning_rate": 1.629203216374269e-05,
+      "loss": 1.8713,
+      "step": 4058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9053232669830322,
+      "learning_rate": 1.629111842105263e-05,
+      "loss": 1.765,
+      "step": 4059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.434453248977661,
+      "learning_rate": 1.6290204678362575e-05,
+      "loss": 1.721,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2683355808258057,
+      "learning_rate": 1.6289290935672515e-05,
+      "loss": 1.2889,
+      "step": 4061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6507534980773926,
+      "learning_rate": 1.628837719298246e-05,
+      "loss": 1.4884,
+      "step": 4062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2341418266296387,
+      "learning_rate": 1.62874634502924e-05,
+      "loss": 1.9227,
+      "step": 4063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.537410259246826,
+      "learning_rate": 1.628654970760234e-05,
+      "loss": 2.1434,
+      "step": 4064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0456271171569824,
+      "learning_rate": 1.6285635964912284e-05,
+      "loss": 1.6916,
+      "step": 4065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1818583011627197,
+      "learning_rate": 1.6284722222222225e-05,
+      "loss": 1.8892,
+      "step": 4066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5925796031951904,
+      "learning_rate": 1.6283808479532165e-05,
+      "loss": 1.9928,
+      "step": 4067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.146639347076416,
+      "learning_rate": 1.628289473684211e-05,
+      "loss": 1.9897,
+      "step": 4068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.524691581726074,
+      "learning_rate": 1.628198099415205e-05,
+      "loss": 1.2714,
+      "step": 4069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.283677101135254,
+      "learning_rate": 1.628106725146199e-05,
+      "loss": 1.7073,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.698970317840576,
+      "learning_rate": 1.628015350877193e-05,
+      "loss": 2.6376,
+      "step": 4071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.706874132156372,
+      "learning_rate": 1.6279239766081875e-05,
+      "loss": 1.7146,
+      "step": 4072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8703665733337402,
+      "learning_rate": 1.6278326023391815e-05,
+      "loss": 1.9065,
+      "step": 4073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0014257431030273,
+      "learning_rate": 1.6277412280701756e-05,
+      "loss": 1.6273,
+      "step": 4074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.708845376968384,
+      "learning_rate": 1.62764985380117e-05,
+      "loss": 1.607,
+      "step": 4075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2546606063842773,
+      "learning_rate": 1.627558479532164e-05,
+      "loss": 1.9543,
+      "step": 4076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.275528907775879,
+      "learning_rate": 1.627467105263158e-05,
+      "loss": 1.7466,
+      "step": 4077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6564078330993652,
+      "learning_rate": 1.627375730994152e-05,
+      "loss": 1.1969,
+      "step": 4078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.979572057723999,
+      "learning_rate": 1.6272843567251465e-05,
+      "loss": 1.7851,
+      "step": 4079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8845343589782715,
+      "learning_rate": 1.6271929824561406e-05,
+      "loss": 1.7661,
+      "step": 4080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2860710620880127,
+      "learning_rate": 1.6271016081871346e-05,
+      "loss": 1.8832,
+      "step": 4081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1449999809265137,
+      "learning_rate": 1.627010233918129e-05,
+      "loss": 1.8471,
+      "step": 4082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4544053077697754,
+      "learning_rate": 1.626918859649123e-05,
+      "loss": 1.4493,
+      "step": 4083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6503794193267822,
+      "learning_rate": 1.626827485380117e-05,
+      "loss": 1.4342,
+      "step": 4084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.716118574142456,
+      "learning_rate": 1.626736111111111e-05,
+      "loss": 1.6102,
+      "step": 4085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.942298173904419,
+      "learning_rate": 1.6266447368421056e-05,
+      "loss": 1.5324,
+      "step": 4086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.052208185195923,
+      "learning_rate": 1.6265533625730996e-05,
+      "loss": 1.9123,
+      "step": 4087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3108630180358887,
+      "learning_rate": 1.6264619883040937e-05,
+      "loss": 1.7839,
+      "step": 4088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.515392780303955,
+      "learning_rate": 1.626370614035088e-05,
+      "loss": 1.9812,
+      "step": 4089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.035765647888184,
+      "learning_rate": 1.626279239766082e-05,
+      "loss": 1.9826,
+      "step": 4090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8924710750579834,
+      "learning_rate": 1.626187865497076e-05,
+      "loss": 1.8111,
+      "step": 4091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.128300428390503,
+      "learning_rate": 1.6260964912280702e-05,
+      "loss": 1.7926,
+      "step": 4092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.68709659576416,
+      "learning_rate": 1.6260051169590646e-05,
+      "loss": 1.4829,
+      "step": 4093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6697447299957275,
+      "learning_rate": 1.6259137426900587e-05,
+      "loss": 1.6613,
+      "step": 4094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.059648275375366,
+      "learning_rate": 1.6258223684210527e-05,
+      "loss": 1.7973,
+      "step": 4095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.199002981185913,
+      "learning_rate": 1.625730994152047e-05,
+      "loss": 1.5389,
+      "step": 4096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.615694046020508,
+      "learning_rate": 1.625639619883041e-05,
+      "loss": 2.2153,
+      "step": 4097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.622971534729004,
+      "learning_rate": 1.6255482456140352e-05,
+      "loss": 1.2209,
+      "step": 4098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.868561267852783,
+      "learning_rate": 1.6254568713450293e-05,
+      "loss": 1.4051,
+      "step": 4099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3477084636688232,
+      "learning_rate": 1.6253654970760236e-05,
+      "loss": 1.7677,
+      "step": 4100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.111135721206665,
+      "learning_rate": 1.6252741228070177e-05,
+      "loss": 1.6338,
+      "step": 4101
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9024317264556885,
+      "learning_rate": 1.6251827485380118e-05,
+      "loss": 1.6586,
+      "step": 4102
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.922410011291504,
+      "learning_rate": 1.625091374269006e-05,
+      "loss": 2.35,
+      "step": 4103
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3004298210144043,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 1.6289,
+      "step": 4104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3702287673950195,
+      "learning_rate": 1.6249086257309942e-05,
+      "loss": 1.2767,
+      "step": 4105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.050954818725586,
+      "learning_rate": 1.6248172514619883e-05,
+      "loss": 2.0468,
+      "step": 4106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8389739990234375,
+      "learning_rate": 1.6247258771929827e-05,
+      "loss": 1.8193,
+      "step": 4107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9184763431549072,
+      "learning_rate": 1.6246345029239767e-05,
+      "loss": 1.8706,
+      "step": 4108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4684970378875732,
+      "learning_rate": 1.6245431286549708e-05,
+      "loss": 1.8038,
+      "step": 4109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0329389572143555,
+      "learning_rate": 1.6244517543859652e-05,
+      "loss": 2.0944,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7452166080474854,
+      "learning_rate": 1.6243603801169592e-05,
+      "loss": 1.6291,
+      "step": 4111
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3909037113189697,
+      "learning_rate": 1.6242690058479533e-05,
+      "loss": 1.541,
+      "step": 4112
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3872690200805664,
+      "learning_rate": 1.6241776315789473e-05,
+      "loss": 1.9082,
+      "step": 4113
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9649438858032227,
+      "learning_rate": 1.6240862573099417e-05,
+      "loss": 1.6375,
+      "step": 4114
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5551130771636963,
+      "learning_rate": 1.6239948830409358e-05,
+      "loss": 1.7663,
+      "step": 4115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.090576648712158,
+      "learning_rate": 1.62390350877193e-05,
+      "loss": 1.6875,
+      "step": 4116
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.934791088104248,
+      "learning_rate": 1.6238121345029242e-05,
+      "loss": 1.4994,
+      "step": 4117
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.082324743270874,
+      "learning_rate": 1.6237207602339183e-05,
+      "loss": 1.7401,
+      "step": 4118
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9690513610839844,
+      "learning_rate": 1.6236293859649123e-05,
+      "loss": 1.7252,
+      "step": 4119
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.638061046600342,
+      "learning_rate": 1.6235380116959064e-05,
+      "loss": 1.6884,
+      "step": 4120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8816752433776855,
+      "learning_rate": 1.6234466374269008e-05,
+      "loss": 1.7081,
+      "step": 4121
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1945834159851074,
+      "learning_rate": 1.623355263157895e-05,
+      "loss": 1.6599,
+      "step": 4122
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.900477647781372,
+      "learning_rate": 1.623263888888889e-05,
+      "loss": 2.056,
+      "step": 4123
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.873551607131958,
+      "learning_rate": 1.6231725146198833e-05,
+      "loss": 2.0588,
+      "step": 4124
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3714725971221924,
+      "learning_rate": 1.6230811403508773e-05,
+      "loss": 1.4316,
+      "step": 4125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.278337240219116,
+      "learning_rate": 1.6229897660818714e-05,
+      "loss": 1.6282,
+      "step": 4126
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0518672466278076,
+      "learning_rate": 1.6228983918128654e-05,
+      "loss": 1.8265,
+      "step": 4127
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4282164573669434,
+      "learning_rate": 1.62280701754386e-05,
+      "loss": 1.8929,
+      "step": 4128
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.51115345954895,
+      "learning_rate": 1.622715643274854e-05,
+      "loss": 1.9713,
+      "step": 4129
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.169052839279175,
+      "learning_rate": 1.622624269005848e-05,
+      "loss": 1.9438,
+      "step": 4130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1328842639923096,
+      "learning_rate": 1.6225328947368423e-05,
+      "loss": 1.5579,
+      "step": 4131
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0230376720428467,
+      "learning_rate": 1.6224415204678364e-05,
+      "loss": 1.6428,
+      "step": 4132
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1105706691741943,
+      "learning_rate": 1.6223501461988304e-05,
+      "loss": 1.6634,
+      "step": 4133
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7739689350128174,
+      "learning_rate": 1.6222587719298245e-05,
+      "loss": 1.6528,
+      "step": 4134
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4915263652801514,
+      "learning_rate": 1.622167397660819e-05,
+      "loss": 1.9401,
+      "step": 4135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.020364999771118,
+      "learning_rate": 1.622076023391813e-05,
+      "loss": 1.7317,
+      "step": 4136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.63305401802063,
+      "learning_rate": 1.621984649122807e-05,
+      "loss": 1.8107,
+      "step": 4137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9306039810180664,
+      "learning_rate": 1.6218932748538014e-05,
+      "loss": 1.438,
+      "step": 4138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.724813222885132,
+      "learning_rate": 1.6218019005847954e-05,
+      "loss": 1.541,
+      "step": 4139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.524379253387451,
+      "learning_rate": 1.6217105263157895e-05,
+      "loss": 1.7867,
+      "step": 4140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.282925605773926,
+      "learning_rate": 1.6216191520467835e-05,
+      "loss": 1.7106,
+      "step": 4141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.286057949066162,
+      "learning_rate": 1.621527777777778e-05,
+      "loss": 2.6275,
+      "step": 4142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5075581073760986,
+      "learning_rate": 1.621436403508772e-05,
+      "loss": 1.654,
+      "step": 4143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2201504707336426,
+      "learning_rate": 1.621345029239766e-05,
+      "loss": 1.7953,
+      "step": 4144
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9259936809539795,
+      "learning_rate": 1.6212536549707604e-05,
+      "loss": 1.6309,
+      "step": 4145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5475099086761475,
+      "learning_rate": 1.6211622807017545e-05,
+      "loss": 1.43,
+      "step": 4146
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.622957944869995,
+      "learning_rate": 1.621070906432749e-05,
+      "loss": 1.5705,
+      "step": 4147
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3274576663970947,
+      "learning_rate": 1.620979532163743e-05,
+      "loss": 1.7367,
+      "step": 4148
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4186205863952637,
+      "learning_rate": 1.620888157894737e-05,
+      "loss": 1.9535,
+      "step": 4149
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5853683948516846,
+      "learning_rate": 1.6207967836257314e-05,
+      "loss": 1.5402,
+      "step": 4150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.064772129058838,
+      "learning_rate": 1.6207054093567254e-05,
+      "loss": 1.6802,
+      "step": 4151
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.24544358253479,
+      "learning_rate": 1.6206140350877195e-05,
+      "loss": 1.6969,
+      "step": 4152
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5852248668670654,
+      "learning_rate": 1.6205226608187135e-05,
+      "loss": 1.706,
+      "step": 4153
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.851750612258911,
+      "learning_rate": 1.620431286549708e-05,
+      "loss": 1.5052,
+      "step": 4154
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2818968296051025,
+      "learning_rate": 1.620339912280702e-05,
+      "loss": 1.7814,
+      "step": 4155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.274871349334717,
+      "learning_rate": 1.620248538011696e-05,
+      "loss": 2.0172,
+      "step": 4156
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.856168508529663,
+      "learning_rate": 1.6201571637426904e-05,
+      "loss": 1.6733,
+      "step": 4157
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.748255491256714,
+      "learning_rate": 1.6200657894736845e-05,
+      "loss": 2.258,
+      "step": 4158
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.77559232711792,
+      "learning_rate": 1.6199744152046785e-05,
+      "loss": 1.6933,
+      "step": 4159
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.152892827987671,
+      "learning_rate": 1.6198830409356726e-05,
+      "loss": 1.8448,
+      "step": 4160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8680098056793213,
+      "learning_rate": 1.619791666666667e-05,
+      "loss": 1.7748,
+      "step": 4161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.046353816986084,
+      "learning_rate": 1.619700292397661e-05,
+      "loss": 1.8124,
+      "step": 4162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7294061183929443,
+      "learning_rate": 1.619608918128655e-05,
+      "loss": 1.8106,
+      "step": 4163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.113415241241455,
+      "learning_rate": 1.6195175438596495e-05,
+      "loss": 1.5375,
+      "step": 4164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.822507381439209,
+      "learning_rate": 1.6194261695906435e-05,
+      "loss": 1.7973,
+      "step": 4165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.006650924682617,
+      "learning_rate": 1.6193347953216376e-05,
+      "loss": 1.4757,
+      "step": 4166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4247324466705322,
+      "learning_rate": 1.6192434210526316e-05,
+      "loss": 1.3089,
+      "step": 4167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.634148120880127,
+      "learning_rate": 1.619152046783626e-05,
+      "loss": 1.4318,
+      "step": 4168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.82003116607666,
+      "learning_rate": 1.61906067251462e-05,
+      "loss": 1.6333,
+      "step": 4169
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.15803599357605,
+      "learning_rate": 1.618969298245614e-05,
+      "loss": 1.8058,
+      "step": 4170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.978445291519165,
+      "learning_rate": 1.6188779239766085e-05,
+      "loss": 1.6664,
+      "step": 4171
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4575085639953613,
+      "learning_rate": 1.6187865497076026e-05,
+      "loss": 1.5988,
+      "step": 4172
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2619621753692627,
+      "learning_rate": 1.6186951754385966e-05,
+      "loss": 2.0169,
+      "step": 4173
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4957151412963867,
+      "learning_rate": 1.6186038011695907e-05,
+      "loss": 1.2466,
+      "step": 4174
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.927896499633789,
+      "learning_rate": 1.618512426900585e-05,
+      "loss": 1.7941,
+      "step": 4175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.055778980255127,
+      "learning_rate": 1.618421052631579e-05,
+      "loss": 1.4726,
+      "step": 4176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1355936527252197,
+      "learning_rate": 1.618329678362573e-05,
+      "loss": 1.9492,
+      "step": 4177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8316822052001953,
+      "learning_rate": 1.6182383040935676e-05,
+      "loss": 1.5492,
+      "step": 4178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0471513271331787,
+      "learning_rate": 1.6181469298245616e-05,
+      "loss": 1.8247,
+      "step": 4179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.440769910812378,
+      "learning_rate": 1.6180555555555557e-05,
+      "loss": 2.0359,
+      "step": 4180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.181163787841797,
+      "learning_rate": 1.6179641812865497e-05,
+      "loss": 1.6972,
+      "step": 4181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6368114948272705,
+      "learning_rate": 1.617872807017544e-05,
+      "loss": 1.9172,
+      "step": 4182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0048394203186035,
+      "learning_rate": 1.617781432748538e-05,
+      "loss": 1.5305,
+      "step": 4183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.927168130874634,
+      "learning_rate": 1.6176900584795322e-05,
+      "loss": 1.6754,
+      "step": 4184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1351726055145264,
+      "learning_rate": 1.6175986842105266e-05,
+      "loss": 2.1194,
+      "step": 4185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.600863218307495,
+      "learning_rate": 1.6175073099415207e-05,
+      "loss": 1.6512,
+      "step": 4186
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6677753925323486,
+      "learning_rate": 1.6174159356725147e-05,
+      "loss": 1.8032,
+      "step": 4187
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1505208015441895,
+      "learning_rate": 1.6173245614035088e-05,
+      "loss": 1.757,
+      "step": 4188
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9816243648529053,
+      "learning_rate": 1.617233187134503e-05,
+      "loss": 1.5888,
+      "step": 4189
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2045047283172607,
+      "learning_rate": 1.6171418128654972e-05,
+      "loss": 2.331,
+      "step": 4190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9488728046417236,
+      "learning_rate": 1.6170504385964913e-05,
+      "loss": 1.5722,
+      "step": 4191
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.791131019592285,
+      "learning_rate": 1.6169590643274856e-05,
+      "loss": 1.4729,
+      "step": 4192
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9773716926574707,
+      "learning_rate": 1.6168676900584797e-05,
+      "loss": 1.4387,
+      "step": 4193
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8914949893951416,
+      "learning_rate": 1.6167763157894738e-05,
+      "loss": 1.5129,
+      "step": 4194
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2444651126861572,
+      "learning_rate": 1.6166849415204678e-05,
+      "loss": 1.6572,
+      "step": 4195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2660889625549316,
+      "learning_rate": 1.6165935672514622e-05,
+      "loss": 1.7359,
+      "step": 4196
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1084187030792236,
+      "learning_rate": 1.6165021929824562e-05,
+      "loss": 1.8248,
+      "step": 4197
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9082701206207275,
+      "learning_rate": 1.6164108187134503e-05,
+      "loss": 1.6615,
+      "step": 4198
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6165623664855957,
+      "learning_rate": 1.6163194444444447e-05,
+      "loss": 1.564,
+      "step": 4199
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6787757873535156,
+      "learning_rate": 1.6162280701754387e-05,
+      "loss": 1.4785,
+      "step": 4200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.256049394607544,
+      "learning_rate": 1.6161366959064328e-05,
+      "loss": 1.6903,
+      "step": 4201
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5897092819213867,
+      "learning_rate": 1.616045321637427e-05,
+      "loss": 1.2969,
+      "step": 4202
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1067707538604736,
+      "learning_rate": 1.6159539473684212e-05,
+      "loss": 2.0182,
+      "step": 4203
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2196848392486572,
+      "learning_rate": 1.6158625730994153e-05,
+      "loss": 1.8061,
+      "step": 4204
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2544946670532227,
+      "learning_rate": 1.6157711988304093e-05,
+      "loss": 1.688,
+      "step": 4205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.058533191680908,
+      "learning_rate": 1.6156798245614037e-05,
+      "loss": 1.7693,
+      "step": 4206
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.036405086517334,
+      "learning_rate": 1.6155884502923978e-05,
+      "loss": 1.7508,
+      "step": 4207
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.617772340774536,
+      "learning_rate": 1.615497076023392e-05,
+      "loss": 1.5243,
+      "step": 4208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.991262197494507,
+      "learning_rate": 1.615405701754386e-05,
+      "loss": 1.547,
+      "step": 4209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.271881341934204,
+      "learning_rate": 1.6153143274853803e-05,
+      "loss": 1.6981,
+      "step": 4210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.738328695297241,
+      "learning_rate": 1.6152229532163743e-05,
+      "loss": 1.3172,
+      "step": 4211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.162337303161621,
+      "learning_rate": 1.6151315789473684e-05,
+      "loss": 1.607,
+      "step": 4212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.826538562774658,
+      "learning_rate": 1.6150402046783628e-05,
+      "loss": 1.6125,
+      "step": 4213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3605523109436035,
+      "learning_rate": 1.614948830409357e-05,
+      "loss": 2.2168,
+      "step": 4214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.592932939529419,
+      "learning_rate": 1.614857456140351e-05,
+      "loss": 2.3879,
+      "step": 4215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0075724124908447,
+      "learning_rate": 1.614766081871345e-05,
+      "loss": 1.6408,
+      "step": 4216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.774578094482422,
+      "learning_rate": 1.6146747076023393e-05,
+      "loss": 2.3143,
+      "step": 4217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8361146450042725,
+      "learning_rate": 1.6145833333333334e-05,
+      "loss": 1.8842,
+      "step": 4218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6724600791931152,
+      "learning_rate": 1.6144919590643274e-05,
+      "loss": 1.1826,
+      "step": 4219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.447619676589966,
+      "learning_rate": 1.614400584795322e-05,
+      "loss": 1.4343,
+      "step": 4220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.852787733078003,
+      "learning_rate": 1.614309210526316e-05,
+      "loss": 1.613,
+      "step": 4221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.385531187057495,
+      "learning_rate": 1.61421783625731e-05,
+      "loss": 1.7269,
+      "step": 4222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4216256141662598,
+      "learning_rate": 1.614126461988304e-05,
+      "loss": 1.3403,
+      "step": 4223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9417502880096436,
+      "learning_rate": 1.6140350877192984e-05,
+      "loss": 1.5171,
+      "step": 4224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4929256439208984,
+      "learning_rate": 1.6139437134502924e-05,
+      "loss": 1.611,
+      "step": 4225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.700739622116089,
+      "learning_rate": 1.6138523391812865e-05,
+      "loss": 2.0209,
+      "step": 4226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.127969264984131,
+      "learning_rate": 1.613760964912281e-05,
+      "loss": 1.7936,
+      "step": 4227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.813840627670288,
+      "learning_rate": 1.613669590643275e-05,
+      "loss": 1.5049,
+      "step": 4228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8096561431884766,
+      "learning_rate": 1.613578216374269e-05,
+      "loss": 1.6675,
+      "step": 4229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.498161554336548,
+      "learning_rate": 1.613486842105263e-05,
+      "loss": 2.1156,
+      "step": 4230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2985315322875977,
+      "learning_rate": 1.6133954678362574e-05,
+      "loss": 1.502,
+      "step": 4231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8615615367889404,
+      "learning_rate": 1.6133040935672515e-05,
+      "loss": 1.5362,
+      "step": 4232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9905152320861816,
+      "learning_rate": 1.613212719298246e-05,
+      "loss": 1.656,
+      "step": 4233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1153712272644043,
+      "learning_rate": 1.61312134502924e-05,
+      "loss": 1.812,
+      "step": 4234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.565375328063965,
+      "learning_rate": 1.6130299707602343e-05,
+      "loss": 1.5122,
+      "step": 4235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3222451210021973,
+      "learning_rate": 1.6129385964912284e-05,
+      "loss": 1.4359,
+      "step": 4236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9147701263427734,
+      "learning_rate": 1.6128472222222224e-05,
+      "loss": 1.5024,
+      "step": 4237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9337985515594482,
+      "learning_rate": 1.6127558479532165e-05,
+      "loss": 1.984,
+      "step": 4238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8145647048950195,
+      "learning_rate": 1.612664473684211e-05,
+      "loss": 1.6628,
+      "step": 4239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.309110403060913,
+      "learning_rate": 1.612573099415205e-05,
+      "loss": 1.9389,
+      "step": 4240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2979843616485596,
+      "learning_rate": 1.612481725146199e-05,
+      "loss": 1.963,
+      "step": 4241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4911162853240967,
+      "learning_rate": 1.6123903508771934e-05,
+      "loss": 1.3651,
+      "step": 4242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9451048374176025,
+      "learning_rate": 1.6122989766081874e-05,
+      "loss": 1.8198,
+      "step": 4243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.695621967315674,
+      "learning_rate": 1.6122076023391815e-05,
+      "loss": 1.5618,
+      "step": 4244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0198631286621094,
+      "learning_rate": 1.6121162280701755e-05,
+      "loss": 1.8654,
+      "step": 4245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3127598762512207,
+      "learning_rate": 1.61202485380117e-05,
+      "loss": 1.7818,
+      "step": 4246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9173636436462402,
+      "learning_rate": 1.611933479532164e-05,
+      "loss": 1.56,
+      "step": 4247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.960876703262329,
+      "learning_rate": 1.611842105263158e-05,
+      "loss": 1.5333,
+      "step": 4248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0622453689575195,
+      "learning_rate": 1.6117507309941524e-05,
+      "loss": 1.6841,
+      "step": 4249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.034623384475708,
+      "learning_rate": 1.6116593567251465e-05,
+      "loss": 1.7136,
+      "step": 4250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7920048236846924,
+      "learning_rate": 1.6115679824561405e-05,
+      "loss": 1.468,
+      "step": 4251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6015408039093018,
+      "learning_rate": 1.6114766081871346e-05,
+      "loss": 1.1926,
+      "step": 4252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0478851795196533,
+      "learning_rate": 1.611385233918129e-05,
+      "loss": 1.7226,
+      "step": 4253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.880676746368408,
+      "learning_rate": 1.611293859649123e-05,
+      "loss": 2.0006,
+      "step": 4254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0180587768554688,
+      "learning_rate": 1.611202485380117e-05,
+      "loss": 1.8394,
+      "step": 4255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1941704750061035,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 1.3846,
+      "step": 4256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1921675205230713,
+      "learning_rate": 1.6110197368421055e-05,
+      "loss": 1.7995,
+      "step": 4257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6443965435028076,
+      "learning_rate": 1.6109283625730996e-05,
+      "loss": 1.6084,
+      "step": 4258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.22910213470459,
+      "learning_rate": 1.6108369883040936e-05,
+      "loss": 1.5303,
+      "step": 4259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.179961681365967,
+      "learning_rate": 1.610745614035088e-05,
+      "loss": 1.9723,
+      "step": 4260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8173775672912598,
+      "learning_rate": 1.610654239766082e-05,
+      "loss": 1.7616,
+      "step": 4261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8815767765045166,
+      "learning_rate": 1.610562865497076e-05,
+      "loss": 1.9309,
+      "step": 4262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3773624897003174,
+      "learning_rate": 1.6104714912280705e-05,
+      "loss": 1.7689,
+      "step": 4263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0695395469665527,
+      "learning_rate": 1.6103801169590646e-05,
+      "loss": 1.7839,
+      "step": 4264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3439722061157227,
+      "learning_rate": 1.6102887426900586e-05,
+      "loss": 1.2557,
+      "step": 4265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0969738960266113,
+      "learning_rate": 1.6101973684210527e-05,
+      "loss": 1.6724,
+      "step": 4266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.002995252609253,
+      "learning_rate": 1.610105994152047e-05,
+      "loss": 1.8548,
+      "step": 4267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.863779306411743,
+      "learning_rate": 1.610014619883041e-05,
+      "loss": 1.7359,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.599498987197876,
+      "learning_rate": 1.609923245614035e-05,
+      "loss": 2.188,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6153314113616943,
+      "learning_rate": 1.6098318713450296e-05,
+      "loss": 1.733,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7221519947052,
+      "learning_rate": 1.6097404970760236e-05,
+      "loss": 1.7981,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0489962100982666,
+      "learning_rate": 1.6096491228070177e-05,
+      "loss": 1.7487,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7013070583343506,
+      "learning_rate": 1.6095577485380117e-05,
+      "loss": 1.3841,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.514292001724243,
+      "learning_rate": 1.609466374269006e-05,
+      "loss": 1.9366,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.104212760925293,
+      "learning_rate": 1.609375e-05,
+      "loss": 1.6827,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.708313226699829,
+      "learning_rate": 1.6092836257309942e-05,
+      "loss": 1.9286,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.909184694290161,
+      "learning_rate": 1.6091922514619886e-05,
+      "loss": 1.6125,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8057727813720703,
+      "learning_rate": 1.6091008771929827e-05,
+      "loss": 1.8162,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.068103790283203,
+      "learning_rate": 1.6090095029239767e-05,
+      "loss": 1.955,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8452770709991455,
+      "learning_rate": 1.6089181286549708e-05,
+      "loss": 1.5142,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4960310459136963,
+      "learning_rate": 1.608826754385965e-05,
+      "loss": 1.2728,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7935588359832764,
+      "learning_rate": 1.6087353801169592e-05,
+      "loss": 1.5451,
+      "step": 4282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9748549461364746,
+      "learning_rate": 1.6086440058479533e-05,
+      "loss": 1.6673,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.062300682067871,
+      "learning_rate": 1.6085526315789473e-05,
+      "loss": 1.4251,
+      "step": 4284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8663535118103027,
+      "learning_rate": 1.6084612573099417e-05,
+      "loss": 1.7056,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9530038833618164,
+      "learning_rate": 1.6083698830409358e-05,
+      "loss": 1.5732,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.282352924346924,
+      "learning_rate": 1.6082785087719298e-05,
+      "loss": 1.599,
+      "step": 4287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7590768337249756,
+      "learning_rate": 1.6081871345029242e-05,
+      "loss": 1.4263,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0225493907928467,
+      "learning_rate": 1.6080957602339182e-05,
+      "loss": 1.5464,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9105629920959473,
+      "learning_rate": 1.6080043859649123e-05,
+      "loss": 1.71,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.323789596557617,
+      "learning_rate": 1.6079130116959064e-05,
+      "loss": 1.8804,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.028101682662964,
+      "learning_rate": 1.6078216374269007e-05,
+      "loss": 1.68,
+      "step": 4292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.482696294784546,
+      "learning_rate": 1.6077302631578948e-05,
+      "loss": 1.4784,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7463760375976562,
+      "learning_rate": 1.607638888888889e-05,
+      "loss": 1.6125,
+      "step": 4294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.919189214706421,
+      "learning_rate": 1.6075475146198832e-05,
+      "loss": 1.6092,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.561166524887085,
+      "learning_rate": 1.6074561403508773e-05,
+      "loss": 2.1646,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.539572238922119,
+      "learning_rate": 1.6073647660818713e-05,
+      "loss": 2.1968,
+      "step": 4297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5926315784454346,
+      "learning_rate": 1.6072733918128654e-05,
+      "loss": 2.0379,
+      "step": 4298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.467419385910034,
+      "learning_rate": 1.6071820175438598e-05,
+      "loss": 2.0628,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.332350730895996,
+      "learning_rate": 1.607090643274854e-05,
+      "loss": 1.9613,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.983732223510742,
+      "learning_rate": 1.606999269005848e-05,
+      "loss": 1.698,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9902737140655518,
+      "learning_rate": 1.6069078947368423e-05,
+      "loss": 1.8936,
+      "step": 4302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1474053859710693,
+      "learning_rate": 1.6068165204678363e-05,
+      "loss": 1.789,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6655466556549072,
+      "learning_rate": 1.6067251461988304e-05,
+      "loss": 1.6445,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2977724075317383,
+      "learning_rate": 1.6066337719298244e-05,
+      "loss": 1.6458,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0953590869903564,
+      "learning_rate": 1.606542397660819e-05,
+      "loss": 1.878,
+      "step": 4306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.004068613052368,
+      "learning_rate": 1.606451023391813e-05,
+      "loss": 1.5063,
+      "step": 4307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9667088985443115,
+      "learning_rate": 1.606359649122807e-05,
+      "loss": 1.7346,
+      "step": 4308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2628281116485596,
+      "learning_rate": 1.6062682748538013e-05,
+      "loss": 1.3253,
+      "step": 4309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.880250930786133,
+      "learning_rate": 1.6061769005847954e-05,
+      "loss": 1.7294,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.036142587661743,
+      "learning_rate": 1.6060855263157894e-05,
+      "loss": 1.4033,
+      "step": 4311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.343825340270996,
+      "learning_rate": 1.6059941520467835e-05,
+      "loss": 1.8602,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5490899085998535,
+      "learning_rate": 1.605902777777778e-05,
+      "loss": 1.5162,
+      "step": 4313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5641016960144043,
+      "learning_rate": 1.605811403508772e-05,
+      "loss": 2.1225,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.704962730407715,
+      "learning_rate": 1.605720029239766e-05,
+      "loss": 1.4644,
+      "step": 4315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3145430088043213,
+      "learning_rate": 1.6056286549707604e-05,
+      "loss": 1.8111,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.121286630630493,
+      "learning_rate": 1.6055372807017544e-05,
+      "loss": 1.5639,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8787307739257812,
+      "learning_rate": 1.6054459064327488e-05,
+      "loss": 1.7517,
+      "step": 4318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7553181648254395,
+      "learning_rate": 1.605354532163743e-05,
+      "loss": 1.6274,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.603983163833618,
+      "learning_rate": 1.605263157894737e-05,
+      "loss": 2.0341,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.566904306411743,
+      "learning_rate": 1.6051717836257313e-05,
+      "loss": 1.4221,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.635730504989624,
+      "learning_rate": 1.6050804093567254e-05,
+      "loss": 1.8952,
+      "step": 4322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.51788592338562,
+      "learning_rate": 1.6049890350877194e-05,
+      "loss": 1.3605,
+      "step": 4323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.630483865737915,
+      "learning_rate": 1.6048976608187138e-05,
+      "loss": 1.7699,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6701486110687256,
+      "learning_rate": 1.604806286549708e-05,
+      "loss": 1.661,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7446084022521973,
+      "learning_rate": 1.604714912280702e-05,
+      "loss": 1.7159,
+      "step": 4326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3897862434387207,
+      "learning_rate": 1.604623538011696e-05,
+      "loss": 1.4162,
+      "step": 4327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9896838665008545,
+      "learning_rate": 1.6045321637426904e-05,
+      "loss": 1.9876,
+      "step": 4328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.793782949447632,
+      "learning_rate": 1.6044407894736844e-05,
+      "loss": 1.608,
+      "step": 4329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5763227939605713,
+      "learning_rate": 1.6043494152046785e-05,
+      "loss": 1.438,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2081220149993896,
+      "learning_rate": 1.604258040935673e-05,
+      "loss": 1.6463,
+      "step": 4331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7949929237365723,
+      "learning_rate": 1.604166666666667e-05,
+      "loss": 1.3194,
+      "step": 4332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.904641628265381,
+      "learning_rate": 1.604075292397661e-05,
+      "loss": 1.8621,
+      "step": 4333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4553158283233643,
+      "learning_rate": 1.603983918128655e-05,
+      "loss": 1.5753,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2794389724731445,
+      "learning_rate": 1.6038925438596494e-05,
+      "loss": 1.9466,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.491302728652954,
+      "learning_rate": 1.6038011695906435e-05,
+      "loss": 1.5527,
+      "step": 4336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5697734355926514,
+      "learning_rate": 1.6037097953216375e-05,
+      "loss": 1.3927,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6610376834869385,
+      "learning_rate": 1.603618421052632e-05,
+      "loss": 1.765,
+      "step": 4338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9023797512054443,
+      "learning_rate": 1.603527046783626e-05,
+      "loss": 1.6706,
+      "step": 4339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.135497570037842,
+      "learning_rate": 1.60343567251462e-05,
+      "loss": 1.8646,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.284219741821289,
+      "learning_rate": 1.603344298245614e-05,
+      "loss": 1.3546,
+      "step": 4341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3224780559539795,
+      "learning_rate": 1.6032529239766085e-05,
+      "loss": 1.625,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0757856369018555,
+      "learning_rate": 1.6031615497076025e-05,
+      "loss": 1.9993,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3526620864868164,
+      "learning_rate": 1.6030701754385966e-05,
+      "loss": 2.2132,
+      "step": 4344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7538087368011475,
+      "learning_rate": 1.602978801169591e-05,
+      "loss": 1.9207,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1460540294647217,
+      "learning_rate": 1.602887426900585e-05,
+      "loss": 1.9089,
+      "step": 4346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3215973377227783,
+      "learning_rate": 1.602796052631579e-05,
+      "loss": 1.5205,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0402019023895264,
+      "learning_rate": 1.602704678362573e-05,
+      "loss": 1.6222,
+      "step": 4348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0353262424468994,
+      "learning_rate": 1.6026133040935675e-05,
+      "loss": 1.7426,
+      "step": 4349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.716582775115967,
+      "learning_rate": 1.6025219298245616e-05,
+      "loss": 1.5569,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7187724113464355,
+      "learning_rate": 1.6024305555555556e-05,
+      "loss": 1.673,
+      "step": 4351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0095269680023193,
+      "learning_rate": 1.60233918128655e-05,
+      "loss": 2.0499,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.403569459915161,
+      "learning_rate": 1.602247807017544e-05,
+      "loss": 1.7629,
+      "step": 4353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.388188600540161,
+      "learning_rate": 1.602156432748538e-05,
+      "loss": 1.8044,
+      "step": 4354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3740217685699463,
+      "learning_rate": 1.602065058479532e-05,
+      "loss": 1.9628,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7702810764312744,
+      "learning_rate": 1.6019736842105266e-05,
+      "loss": 1.6753,
+      "step": 4356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.932826042175293,
+      "learning_rate": 1.6018823099415206e-05,
+      "loss": 1.7323,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.542206764221191,
+      "learning_rate": 1.6017909356725147e-05,
+      "loss": 1.4093,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2108683586120605,
+      "learning_rate": 1.601699561403509e-05,
+      "loss": 1.6457,
+      "step": 4359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5035653114318848,
+      "learning_rate": 1.601608187134503e-05,
+      "loss": 1.7523,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4108095169067383,
+      "learning_rate": 1.601516812865497e-05,
+      "loss": 1.3631,
+      "step": 4361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.622563600540161,
+      "learning_rate": 1.6014254385964912e-05,
+      "loss": 1.7586,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0612738132476807,
+      "learning_rate": 1.6013340643274856e-05,
+      "loss": 1.956,
+      "step": 4363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.205875873565674,
+      "learning_rate": 1.6012426900584797e-05,
+      "loss": 1.6052,
+      "step": 4364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.531952381134033,
+      "learning_rate": 1.6011513157894737e-05,
+      "loss": 2.1759,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.809760808944702,
+      "learning_rate": 1.601059941520468e-05,
+      "loss": 1.3984,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7147772312164307,
+      "learning_rate": 1.600968567251462e-05,
+      "loss": 1.3984,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1666393280029297,
+      "learning_rate": 1.6008771929824562e-05,
+      "loss": 1.5959,
+      "step": 4368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.640089273452759,
+      "learning_rate": 1.6007858187134503e-05,
+      "loss": 1.6469,
+      "step": 4369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.286285161972046,
+      "learning_rate": 1.6006944444444447e-05,
+      "loss": 1.5356,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7872250080108643,
+      "learning_rate": 1.6006030701754387e-05,
+      "loss": 1.6938,
+      "step": 4371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.121894121170044,
+      "learning_rate": 1.6005116959064328e-05,
+      "loss": 1.8556,
+      "step": 4372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5340847969055176,
+      "learning_rate": 1.600420321637427e-05,
+      "loss": 1.669,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2957799434661865,
+      "learning_rate": 1.6003289473684212e-05,
+      "loss": 1.5804,
+      "step": 4374
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0516107082366943,
+      "learning_rate": 1.6002375730994153e-05,
+      "loss": 1.6143,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.205249786376953,
+      "learning_rate": 1.6001461988304093e-05,
+      "loss": 1.8249,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5966241359710693,
+      "learning_rate": 1.6000548245614037e-05,
+      "loss": 1.4362,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.85953950881958,
+      "learning_rate": 1.5999634502923978e-05,
+      "loss": 1.6023,
+      "step": 4378
+    },
+    {
+      "epoch": 0.2,
+      "eval_train_bleu": 0.22515474268017505,
+      "eval_train_loss": 2.1745991706848145,
+      "eval_train_runtime": 23.4069,
+      "eval_train_samples_per_second": 42.765,
+      "eval_train_steps_per_second": 1.367,
+      "step": 4378
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.893463611602783,
+      "learning_rate": 1.5998720760233918e-05,
+      "loss": 1.7377,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6054344177246094,
+      "learning_rate": 1.5997807017543862e-05,
+      "loss": 1.6316,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3410568237304688,
+      "learning_rate": 1.5996893274853802e-05,
+      "loss": 1.7454,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9931013584136963,
+      "learning_rate": 1.5995979532163743e-05,
+      "loss": 1.843,
+      "step": 4382
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2709944248199463,
+      "learning_rate": 1.5995065789473684e-05,
+      "loss": 1.9658,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4128966331481934,
+      "learning_rate": 1.5994152046783627e-05,
+      "loss": 1.9234,
+      "step": 4384
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.441795825958252,
+      "learning_rate": 1.5993238304093568e-05,
+      "loss": 2.3506,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9552536010742188,
+      "learning_rate": 1.599232456140351e-05,
+      "loss": 1.9477,
+      "step": 4386
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.502366065979004,
+      "learning_rate": 1.5991410818713452e-05,
+      "loss": 1.5864,
+      "step": 4387
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.902116060256958,
+      "learning_rate": 1.5990497076023393e-05,
+      "loss": 1.6767,
+      "step": 4388
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.688615322113037,
+      "learning_rate": 1.5989583333333333e-05,
+      "loss": 1.8796,
+      "step": 4389
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.891735553741455,
+      "learning_rate": 1.5988669590643274e-05,
+      "loss": 1.4424,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.799225091934204,
+      "learning_rate": 1.5987755847953218e-05,
+      "loss": 1.4998,
+      "step": 4391
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2195394039154053,
+      "learning_rate": 1.598684210526316e-05,
+      "loss": 1.8988,
+      "step": 4392
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0111911296844482,
+      "learning_rate": 1.59859283625731e-05,
+      "loss": 1.5209,
+      "step": 4393
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6122946739196777,
+      "learning_rate": 1.5985014619883043e-05,
+      "loss": 1.6482,
+      "step": 4394
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.251645803451538,
+      "learning_rate": 1.5984100877192983e-05,
+      "loss": 1.9723,
+      "step": 4395
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.038771867752075,
+      "learning_rate": 1.5983187134502924e-05,
+      "loss": 1.6195,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.693330764770508,
+      "learning_rate": 1.5982273391812864e-05,
+      "loss": 1.567,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4366981983184814,
+      "learning_rate": 1.598135964912281e-05,
+      "loss": 1.6592,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2326505184173584,
+      "learning_rate": 1.598044590643275e-05,
+      "loss": 1.5697,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6505842208862305,
+      "learning_rate": 1.597953216374269e-05,
+      "loss": 1.4361,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5864598751068115,
+      "learning_rate": 1.5978618421052633e-05,
+      "loss": 2.0577,
+      "step": 4401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.662689208984375,
+      "learning_rate": 1.5977704678362574e-05,
+      "loss": 1.5449,
+      "step": 4402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2195346355438232,
+      "learning_rate": 1.5976790935672514e-05,
+      "loss": 1.5614,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0345516204833984,
+      "learning_rate": 1.597587719298246e-05,
+      "loss": 1.4523,
+      "step": 4404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7500696182250977,
+      "learning_rate": 1.59749634502924e-05,
+      "loss": 1.5805,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2926394939422607,
+      "learning_rate": 1.5974049707602343e-05,
+      "loss": 1.8692,
+      "step": 4406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7384908199310303,
+      "learning_rate": 1.5973135964912283e-05,
+      "loss": 1.3613,
+      "step": 4407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6780917644500732,
+      "learning_rate": 1.5972222222222224e-05,
+      "loss": 1.4709,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9847428798675537,
+      "learning_rate": 1.5971308479532164e-05,
+      "loss": 1.7483,
+      "step": 4409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6829614639282227,
+      "learning_rate": 1.5970394736842108e-05,
+      "loss": 1.3778,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4266536235809326,
+      "learning_rate": 1.596948099415205e-05,
+      "loss": 1.8364,
+      "step": 4411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8932251930236816,
+      "learning_rate": 1.596856725146199e-05,
+      "loss": 1.5049,
+      "step": 4412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8646278381347656,
+      "learning_rate": 1.5967653508771933e-05,
+      "loss": 1.6296,
+      "step": 4413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.815739631652832,
+      "learning_rate": 1.5966739766081874e-05,
+      "loss": 1.444,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.944594621658325,
+      "learning_rate": 1.5965826023391814e-05,
+      "loss": 1.6599,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1036319732666016,
+      "learning_rate": 1.5964912280701755e-05,
+      "loss": 2.0204,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.604436159133911,
+      "learning_rate": 1.59639985380117e-05,
+      "loss": 1.6737,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.981194496154785,
+      "learning_rate": 1.596308479532164e-05,
+      "loss": 1.5151,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.879513740539551,
+      "learning_rate": 1.596217105263158e-05,
+      "loss": 2.1019,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8735272884368896,
+      "learning_rate": 1.5961257309941524e-05,
+      "loss": 1.8277,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4077670574188232,
+      "learning_rate": 1.5960343567251464e-05,
+      "loss": 2.0949,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6676523685455322,
+      "learning_rate": 1.5959429824561405e-05,
+      "loss": 1.5507,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.792837381362915,
+      "learning_rate": 1.5958516081871345e-05,
+      "loss": 2.12,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1904189586639404,
+      "learning_rate": 1.595760233918129e-05,
+      "loss": 2.0233,
+      "step": 4424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.172532320022583,
+      "learning_rate": 1.595668859649123e-05,
+      "loss": 1.2514,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.365793466567993,
+      "learning_rate": 1.595577485380117e-05,
+      "loss": 1.9753,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0165340900421143,
+      "learning_rate": 1.5954861111111114e-05,
+      "loss": 1.6842,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.214016914367676,
+      "learning_rate": 1.5953947368421055e-05,
+      "loss": 1.601,
+      "step": 4428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6832504272460938,
+      "learning_rate": 1.5953033625730995e-05,
+      "loss": 1.4638,
+      "step": 4429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.775275707244873,
+      "learning_rate": 1.5952119883040936e-05,
+      "loss": 1.583,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5494308471679688,
+      "learning_rate": 1.595120614035088e-05,
+      "loss": 1.4632,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9192306995391846,
+      "learning_rate": 1.595029239766082e-05,
+      "loss": 1.5249,
+      "step": 4432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.291775703430176,
+      "learning_rate": 1.594937865497076e-05,
+      "loss": 1.3198,
+      "step": 4433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.755812406539917,
+      "learning_rate": 1.5948464912280705e-05,
+      "loss": 1.6426,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.006155967712402,
+      "learning_rate": 1.5947551169590645e-05,
+      "loss": 2.3149,
+      "step": 4435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9363837242126465,
+      "learning_rate": 1.5946637426900586e-05,
+      "loss": 1.7299,
+      "step": 4436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.131256580352783,
+      "learning_rate": 1.5945723684210526e-05,
+      "loss": 1.5693,
+      "step": 4437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7433042526245117,
+      "learning_rate": 1.594480994152047e-05,
+      "loss": 1.9039,
+      "step": 4438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.835063934326172,
+      "learning_rate": 1.594389619883041e-05,
+      "loss": 1.7219,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.54982852935791,
+      "learning_rate": 1.594298245614035e-05,
+      "loss": 1.3213,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9560375213623047,
+      "learning_rate": 1.5942068713450295e-05,
+      "loss": 1.3966,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9048871994018555,
+      "learning_rate": 1.5941154970760236e-05,
+      "loss": 1.8168,
+      "step": 4442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9209115505218506,
+      "learning_rate": 1.5940241228070176e-05,
+      "loss": 1.8966,
+      "step": 4443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0429725646972656,
+      "learning_rate": 1.5939327485380117e-05,
+      "loss": 1.7613,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0732760429382324,
+      "learning_rate": 1.593841374269006e-05,
+      "loss": 1.6085,
+      "step": 4445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.420713424682617,
+      "learning_rate": 1.59375e-05,
+      "loss": 1.6052,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4328815937042236,
+      "learning_rate": 1.593658625730994e-05,
+      "loss": 2.0572,
+      "step": 4447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.769810438156128,
+      "learning_rate": 1.5935672514619886e-05,
+      "loss": 2.3223,
+      "step": 4448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.272365093231201,
+      "learning_rate": 1.5934758771929826e-05,
+      "loss": 1.839,
+      "step": 4449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.162778854370117,
+      "learning_rate": 1.5933845029239767e-05,
+      "loss": 2.2638,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.400428295135498,
+      "learning_rate": 1.5932931286549707e-05,
+      "loss": 1.864,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.373326539993286,
+      "learning_rate": 1.593201754385965e-05,
+      "loss": 1.5403,
+      "step": 4452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.89599347114563,
+      "learning_rate": 1.593110380116959e-05,
+      "loss": 2.0205,
+      "step": 4453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9643802642822266,
+      "learning_rate": 1.5930190058479532e-05,
+      "loss": 1.5883,
+      "step": 4454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.565126657485962,
+      "learning_rate": 1.5929276315789476e-05,
+      "loss": 1.1808,
+      "step": 4455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6637110710144043,
+      "learning_rate": 1.5928362573099417e-05,
+      "loss": 1.7608,
+      "step": 4456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6494288444519043,
+      "learning_rate": 1.5927448830409357e-05,
+      "loss": 1.5573,
+      "step": 4457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.227224111557007,
+      "learning_rate": 1.5926535087719298e-05,
+      "loss": 2.0118,
+      "step": 4458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4243905544281006,
+      "learning_rate": 1.592562134502924e-05,
+      "loss": 1.427,
+      "step": 4459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.848346471786499,
+      "learning_rate": 1.5924707602339182e-05,
+      "loss": 1.6814,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8249809741973877,
+      "learning_rate": 1.5923793859649123e-05,
+      "loss": 1.5975,
+      "step": 4461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.744182586669922,
+      "learning_rate": 1.5922880116959067e-05,
+      "loss": 1.4631,
+      "step": 4462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.392242431640625,
+      "learning_rate": 1.5921966374269007e-05,
+      "loss": 1.4668,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.144699811935425,
+      "learning_rate": 1.5921052631578948e-05,
+      "loss": 1.7042,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4652254581451416,
+      "learning_rate": 1.5920138888888888e-05,
+      "loss": 1.5194,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6328649520874023,
+      "learning_rate": 1.5919225146198832e-05,
+      "loss": 1.7537,
+      "step": 4466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.607720136642456,
+      "learning_rate": 1.5918311403508773e-05,
+      "loss": 1.505,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7183620929718018,
+      "learning_rate": 1.5917397660818713e-05,
+      "loss": 1.4988,
+      "step": 4468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.803009271621704,
+      "learning_rate": 1.5916483918128657e-05,
+      "loss": 1.5886,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.956653594970703,
+      "learning_rate": 1.5915570175438598e-05,
+      "loss": 1.7353,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.785128355026245,
+      "learning_rate": 1.5914656432748538e-05,
+      "loss": 1.8176,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.073688268661499,
+      "learning_rate": 1.591374269005848e-05,
+      "loss": 1.5974,
+      "step": 4472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5816309452056885,
+      "learning_rate": 1.5912828947368422e-05,
+      "loss": 1.4443,
+      "step": 4473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8907155990600586,
+      "learning_rate": 1.5911915204678363e-05,
+      "loss": 1.3854,
+      "step": 4474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3694112300872803,
+      "learning_rate": 1.5911001461988304e-05,
+      "loss": 1.4444,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8100569248199463,
+      "learning_rate": 1.5910087719298247e-05,
+      "loss": 1.641,
+      "step": 4476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7863829135894775,
+      "learning_rate": 1.5909173976608188e-05,
+      "loss": 1.7363,
+      "step": 4477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.353630781173706,
+      "learning_rate": 1.590826023391813e-05,
+      "loss": 1.7111,
+      "step": 4478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1183834075927734,
+      "learning_rate": 1.590734649122807e-05,
+      "loss": 2.2147,
+      "step": 4479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8729395866394043,
+      "learning_rate": 1.5906432748538013e-05,
+      "loss": 1.5183,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.057482957839966,
+      "learning_rate": 1.5905519005847953e-05,
+      "loss": 1.2675,
+      "step": 4481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8498897552490234,
+      "learning_rate": 1.5904605263157894e-05,
+      "loss": 1.5366,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.133025646209717,
+      "learning_rate": 1.5903691520467838e-05,
+      "loss": 2.2142,
+      "step": 4483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.564995050430298,
+      "learning_rate": 1.590277777777778e-05,
+      "loss": 1.3679,
+      "step": 4484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.598114490509033,
+      "learning_rate": 1.590186403508772e-05,
+      "loss": 1.9086,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.16255259513855,
+      "learning_rate": 1.590095029239766e-05,
+      "loss": 1.5646,
+      "step": 4486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.028782844543457,
+      "learning_rate": 1.5900036549707603e-05,
+      "loss": 1.6302,
+      "step": 4487
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.382321834564209,
+      "learning_rate": 1.5899122807017544e-05,
+      "loss": 2.2265,
+      "step": 4488
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3282337188720703,
+      "learning_rate": 1.5898209064327488e-05,
+      "loss": 1.616,
+      "step": 4489
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7781009674072266,
+      "learning_rate": 1.589729532163743e-05,
+      "loss": 1.6365,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6002438068389893,
+      "learning_rate": 1.5896381578947372e-05,
+      "loss": 1.4168,
+      "step": 4491
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.108018159866333,
+      "learning_rate": 1.5895467836257313e-05,
+      "loss": 2.1247,
+      "step": 4492
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.651303291320801,
+      "learning_rate": 1.5894554093567253e-05,
+      "loss": 1.611,
+      "step": 4493
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.468636989593506,
+      "learning_rate": 1.5893640350877194e-05,
+      "loss": 1.3351,
+      "step": 4494
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3522517681121826,
+      "learning_rate": 1.5892726608187138e-05,
+      "loss": 1.8911,
+      "step": 4495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9735450744628906,
+      "learning_rate": 1.589181286549708e-05,
+      "loss": 1.7075,
+      "step": 4496
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.034592390060425,
+      "learning_rate": 1.589089912280702e-05,
+      "loss": 1.3983,
+      "step": 4497
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.725198984146118,
+      "learning_rate": 1.5889985380116963e-05,
+      "loss": 1.3093,
+      "step": 4498
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.681213855743408,
+      "learning_rate": 1.5889071637426903e-05,
+      "loss": 1.5112,
+      "step": 4499
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.786651134490967,
+      "learning_rate": 1.5888157894736844e-05,
+      "loss": 1.7935,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6632022857666016,
+      "learning_rate": 1.5887244152046784e-05,
+      "loss": 1.5627,
+      "step": 4501
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9170806407928467,
+      "learning_rate": 1.5886330409356728e-05,
+      "loss": 1.9292,
+      "step": 4502
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0872726440429688,
+      "learning_rate": 1.588541666666667e-05,
+      "loss": 1.6312,
+      "step": 4503
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1020944118499756,
+      "learning_rate": 1.588450292397661e-05,
+      "loss": 1.9149,
+      "step": 4504
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8527064323425293,
+      "learning_rate": 1.5883589181286553e-05,
+      "loss": 1.642,
+      "step": 4505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.251718521118164,
+      "learning_rate": 1.5882675438596494e-05,
+      "loss": 1.9913,
+      "step": 4506
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6563220024108887,
+      "learning_rate": 1.5881761695906434e-05,
+      "loss": 1.5456,
+      "step": 4507
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.000225067138672,
+      "learning_rate": 1.5880847953216375e-05,
+      "loss": 1.9156,
+      "step": 4508
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7090108394622803,
+      "learning_rate": 1.587993421052632e-05,
+      "loss": 2.2844,
+      "step": 4509
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8133416175842285,
+      "learning_rate": 1.587902046783626e-05,
+      "loss": 1.3976,
+      "step": 4510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1781959533691406,
+      "learning_rate": 1.58781067251462e-05,
+      "loss": 1.4229,
+      "step": 4511
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6540539264678955,
+      "learning_rate": 1.5877192982456144e-05,
+      "loss": 1.5455,
+      "step": 4512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.82838773727417,
+      "learning_rate": 1.5876279239766084e-05,
+      "loss": 1.6397,
+      "step": 4513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.421006917953491,
+      "learning_rate": 1.5875365497076025e-05,
+      "loss": 2.0525,
+      "step": 4514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8512940406799316,
+      "learning_rate": 1.5874451754385965e-05,
+      "loss": 1.4832,
+      "step": 4515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.506040096282959,
+      "learning_rate": 1.587353801169591e-05,
+      "loss": 1.8152,
+      "step": 4516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6314854621887207,
+      "learning_rate": 1.587262426900585e-05,
+      "loss": 1.4677,
+      "step": 4517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.199639320373535,
+      "learning_rate": 1.587171052631579e-05,
+      "loss": 1.9671,
+      "step": 4518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4105515480041504,
+      "learning_rate": 1.5870796783625734e-05,
+      "loss": 2.1741,
+      "step": 4519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6965417861938477,
+      "learning_rate": 1.5869883040935675e-05,
+      "loss": 1.7743,
+      "step": 4520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4409799575805664,
+      "learning_rate": 1.5868969298245615e-05,
+      "loss": 1.8921,
+      "step": 4521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.050553321838379,
+      "learning_rate": 1.5868055555555556e-05,
+      "loss": 1.5594,
+      "step": 4522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2766268253326416,
+      "learning_rate": 1.58671418128655e-05,
+      "loss": 1.6994,
+      "step": 4523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.053864002227783,
+      "learning_rate": 1.586622807017544e-05,
+      "loss": 1.6058,
+      "step": 4524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2637381553649902,
+      "learning_rate": 1.586531432748538e-05,
+      "loss": 1.8624,
+      "step": 4525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3041749000549316,
+      "learning_rate": 1.5864400584795325e-05,
+      "loss": 1.8547,
+      "step": 4526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.189941883087158,
+      "learning_rate": 1.5863486842105265e-05,
+      "loss": 1.8404,
+      "step": 4527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1806108951568604,
+      "learning_rate": 1.5862573099415206e-05,
+      "loss": 1.683,
+      "step": 4528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5878942012786865,
+      "learning_rate": 1.5861659356725146e-05,
+      "loss": 1.3616,
+      "step": 4529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.871833086013794,
+      "learning_rate": 1.586074561403509e-05,
+      "loss": 1.6724,
+      "step": 4530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5401482582092285,
+      "learning_rate": 1.585983187134503e-05,
+      "loss": 1.6857,
+      "step": 4531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7208516597747803,
+      "learning_rate": 1.585891812865497e-05,
+      "loss": 1.7806,
+      "step": 4532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.908097267150879,
+      "learning_rate": 1.5858004385964915e-05,
+      "loss": 1.7383,
+      "step": 4533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.162386894226074,
+      "learning_rate": 1.5857090643274856e-05,
+      "loss": 1.383,
+      "step": 4534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7544596195220947,
+      "learning_rate": 1.5856176900584796e-05,
+      "loss": 1.9258,
+      "step": 4535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.882290840148926,
+      "learning_rate": 1.5855263157894737e-05,
+      "loss": 2.2533,
+      "step": 4536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9036996364593506,
+      "learning_rate": 1.585434941520468e-05,
+      "loss": 2.0401,
+      "step": 4537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.178602695465088,
+      "learning_rate": 1.585343567251462e-05,
+      "loss": 2.2532,
+      "step": 4538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8373992443084717,
+      "learning_rate": 1.585252192982456e-05,
+      "loss": 1.6236,
+      "step": 4539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1733858585357666,
+      "learning_rate": 1.5851608187134506e-05,
+      "loss": 1.3313,
+      "step": 4540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.02180814743042,
+      "learning_rate": 1.5850694444444446e-05,
+      "loss": 1.8915,
+      "step": 4541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2518653869628906,
+      "learning_rate": 1.5849780701754387e-05,
+      "loss": 1.7178,
+      "step": 4542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6301889419555664,
+      "learning_rate": 1.5848866959064327e-05,
+      "loss": 1.6151,
+      "step": 4543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5253546237945557,
+      "learning_rate": 1.584795321637427e-05,
+      "loss": 2.2198,
+      "step": 4544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8626537322998047,
+      "learning_rate": 1.584703947368421e-05,
+      "loss": 1.6627,
+      "step": 4545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.975304126739502,
+      "learning_rate": 1.5846125730994152e-05,
+      "loss": 1.7236,
+      "step": 4546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3112998008728027,
+      "learning_rate": 1.5845211988304096e-05,
+      "loss": 1.795,
+      "step": 4547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.540381669998169,
+      "learning_rate": 1.5844298245614037e-05,
+      "loss": 2.0025,
+      "step": 4548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.388821840286255,
+      "learning_rate": 1.5843384502923977e-05,
+      "loss": 1.6254,
+      "step": 4549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.800487756729126,
+      "learning_rate": 1.5842470760233918e-05,
+      "loss": 1.6568,
+      "step": 4550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7779695987701416,
+      "learning_rate": 1.584155701754386e-05,
+      "loss": 1.6977,
+      "step": 4551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.981557846069336,
+      "learning_rate": 1.5840643274853802e-05,
+      "loss": 1.4264,
+      "step": 4552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8880956172943115,
+      "learning_rate": 1.5839729532163743e-05,
+      "loss": 1.7685,
+      "step": 4553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9829602241516113,
+      "learning_rate": 1.5838815789473687e-05,
+      "loss": 1.8123,
+      "step": 4554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.0969390869140625,
+      "learning_rate": 1.5837902046783627e-05,
+      "loss": 2.1444,
+      "step": 4555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.29550838470459,
+      "learning_rate": 1.5836988304093568e-05,
+      "loss": 1.7538,
+      "step": 4556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8277831077575684,
+      "learning_rate": 1.5836074561403508e-05,
+      "loss": 1.5164,
+      "step": 4557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.870809316635132,
+      "learning_rate": 1.5835160818713452e-05,
+      "loss": 1.5128,
+      "step": 4558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1251556873321533,
+      "learning_rate": 1.5834247076023393e-05,
+      "loss": 1.5853,
+      "step": 4559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.242136240005493,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 1.9998,
+      "step": 4560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6978235244750977,
+      "learning_rate": 1.5832419590643277e-05,
+      "loss": 1.5507,
+      "step": 4561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0244104862213135,
+      "learning_rate": 1.5831505847953218e-05,
+      "loss": 1.9742,
+      "step": 4562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9782040119171143,
+      "learning_rate": 1.5830592105263158e-05,
+      "loss": 1.5366,
+      "step": 4563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.254754066467285,
+      "learning_rate": 1.58296783625731e-05,
+      "loss": 1.763,
+      "step": 4564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2389516830444336,
+      "learning_rate": 1.5828764619883042e-05,
+      "loss": 1.1935,
+      "step": 4565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0764546394348145,
+      "learning_rate": 1.5827850877192983e-05,
+      "loss": 1.9053,
+      "step": 4566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9991536140441895,
+      "learning_rate": 1.5826937134502924e-05,
+      "loss": 1.5871,
+      "step": 4567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6269187927246094,
+      "learning_rate": 1.5826023391812867e-05,
+      "loss": 1.3289,
+      "step": 4568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7619824409484863,
+      "learning_rate": 1.5825109649122808e-05,
+      "loss": 1.752,
+      "step": 4569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9921352863311768,
+      "learning_rate": 1.582419590643275e-05,
+      "loss": 1.5195,
+      "step": 4570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1972506046295166,
+      "learning_rate": 1.582328216374269e-05,
+      "loss": 2.0626,
+      "step": 4571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.275169610977173,
+      "learning_rate": 1.5822368421052633e-05,
+      "loss": 1.9786,
+      "step": 4572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.965033531188965,
+      "learning_rate": 1.5821454678362573e-05,
+      "loss": 1.5805,
+      "step": 4573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.516829013824463,
+      "learning_rate": 1.5820540935672514e-05,
+      "loss": 1.6088,
+      "step": 4574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.849350929260254,
+      "learning_rate": 1.5819627192982458e-05,
+      "loss": 1.7764,
+      "step": 4575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8151986598968506,
+      "learning_rate": 1.58187134502924e-05,
+      "loss": 1.8294,
+      "step": 4576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8602163791656494,
+      "learning_rate": 1.5817799707602342e-05,
+      "loss": 1.4695,
+      "step": 4577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.75510835647583,
+      "learning_rate": 1.5816885964912283e-05,
+      "loss": 1.3179,
+      "step": 4578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.755689859390259,
+      "learning_rate": 1.5815972222222223e-05,
+      "loss": 1.7452,
+      "step": 4579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.227029800415039,
+      "learning_rate": 1.5815058479532167e-05,
+      "loss": 1.6239,
+      "step": 4580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.223414182662964,
+      "learning_rate": 1.5814144736842108e-05,
+      "loss": 1.8436,
+      "step": 4581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.850429058074951,
+      "learning_rate": 1.581323099415205e-05,
+      "loss": 1.5945,
+      "step": 4582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4166829586029053,
+      "learning_rate": 1.581231725146199e-05,
+      "loss": 1.5971,
+      "step": 4583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2758076190948486,
+      "learning_rate": 1.5811403508771933e-05,
+      "loss": 1.3456,
+      "step": 4584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7951860427856445,
+      "learning_rate": 1.5810489766081873e-05,
+      "loss": 1.3509,
+      "step": 4585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.680433750152588,
+      "learning_rate": 1.5809576023391814e-05,
+      "loss": 1.5052,
+      "step": 4586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0980429649353027,
+      "learning_rate": 1.5808662280701758e-05,
+      "loss": 1.433,
+      "step": 4587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3066155910491943,
+      "learning_rate": 1.58077485380117e-05,
+      "loss": 2.0155,
+      "step": 4588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1934690475463867,
+      "learning_rate": 1.580683479532164e-05,
+      "loss": 1.9747,
+      "step": 4589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6036386489868164,
+      "learning_rate": 1.580592105263158e-05,
+      "loss": 1.5003,
+      "step": 4590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.21891713142395,
+      "learning_rate": 1.5805007309941523e-05,
+      "loss": 1.88,
+      "step": 4591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0244650840759277,
+      "learning_rate": 1.5804093567251464e-05,
+      "loss": 1.6819,
+      "step": 4592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5206916332244873,
+      "learning_rate": 1.5803179824561404e-05,
+      "loss": 1.2616,
+      "step": 4593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2634828090667725,
+      "learning_rate": 1.5802266081871348e-05,
+      "loss": 1.3289,
+      "step": 4594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9730770587921143,
+      "learning_rate": 1.580135233918129e-05,
+      "loss": 1.7732,
+      "step": 4595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4546310901641846,
+      "learning_rate": 1.580043859649123e-05,
+      "loss": 1.9108,
+      "step": 4596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5974719524383545,
+      "learning_rate": 1.579952485380117e-05,
+      "loss": 1.9884,
+      "step": 4597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.966522455215454,
+      "learning_rate": 1.5798611111111114e-05,
+      "loss": 1.4365,
+      "step": 4598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9179298877716064,
+      "learning_rate": 1.5797697368421054e-05,
+      "loss": 1.8881,
+      "step": 4599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.198608875274658,
+      "learning_rate": 1.5796783625730995e-05,
+      "loss": 1.6949,
+      "step": 4600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2873167991638184,
+      "learning_rate": 1.579586988304094e-05,
+      "loss": 1.7852,
+      "step": 4601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.123399019241333,
+      "learning_rate": 1.579495614035088e-05,
+      "loss": 1.9739,
+      "step": 4602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7739298343658447,
+      "learning_rate": 1.579404239766082e-05,
+      "loss": 1.6507,
+      "step": 4603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8916754722595215,
+      "learning_rate": 1.579312865497076e-05,
+      "loss": 1.8092,
+      "step": 4604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7865777015686035,
+      "learning_rate": 1.5792214912280704e-05,
+      "loss": 1.5704,
+      "step": 4605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3007054328918457,
+      "learning_rate": 1.5791301169590645e-05,
+      "loss": 2.3128,
+      "step": 4606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.360140085220337,
+      "learning_rate": 1.5790387426900585e-05,
+      "loss": 1.7773,
+      "step": 4607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.847383499145508,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 1.7354,
+      "step": 4608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6271069049835205,
+      "learning_rate": 1.578855994152047e-05,
+      "loss": 1.349,
+      "step": 4609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7210423946380615,
+      "learning_rate": 1.578764619883041e-05,
+      "loss": 1.5024,
+      "step": 4610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2849490642547607,
+      "learning_rate": 1.578673245614035e-05,
+      "loss": 2.0945,
+      "step": 4611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.976679563522339,
+      "learning_rate": 1.5785818713450295e-05,
+      "loss": 2.2966,
+      "step": 4612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4945805072784424,
+      "learning_rate": 1.5784904970760235e-05,
+      "loss": 1.438,
+      "step": 4613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.880603551864624,
+      "learning_rate": 1.5783991228070176e-05,
+      "loss": 1.4889,
+      "step": 4614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7893478870391846,
+      "learning_rate": 1.578307748538012e-05,
+      "loss": 1.653,
+      "step": 4615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1246237754821777,
+      "learning_rate": 1.578216374269006e-05,
+      "loss": 1.8163,
+      "step": 4616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5801029205322266,
+      "learning_rate": 1.578125e-05,
+      "loss": 1.6381,
+      "step": 4617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7117443084716797,
+      "learning_rate": 1.578033625730994e-05,
+      "loss": 1.399,
+      "step": 4618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5574069023132324,
+      "learning_rate": 1.5779422514619885e-05,
+      "loss": 1.3938,
+      "step": 4619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.107362747192383,
+      "learning_rate": 1.5778508771929826e-05,
+      "loss": 1.7055,
+      "step": 4620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.884260654449463,
+      "learning_rate": 1.5777595029239766e-05,
+      "loss": 1.6343,
+      "step": 4621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.332123279571533,
+      "learning_rate": 1.577668128654971e-05,
+      "loss": 2.0345,
+      "step": 4622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.454075336456299,
+      "learning_rate": 1.577576754385965e-05,
+      "loss": 1.6047,
+      "step": 4623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0050148963928223,
+      "learning_rate": 1.577485380116959e-05,
+      "loss": 1.4681,
+      "step": 4624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9376325607299805,
+      "learning_rate": 1.5773940058479532e-05,
+      "loss": 1.9339,
+      "step": 4625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.302737236022949,
+      "learning_rate": 1.5773026315789476e-05,
+      "loss": 1.7486,
+      "step": 4626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2733242511749268,
+      "learning_rate": 1.5772112573099416e-05,
+      "loss": 1.873,
+      "step": 4627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.092524766921997,
+      "learning_rate": 1.5771198830409357e-05,
+      "loss": 1.4024,
+      "step": 4628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8402960300445557,
+      "learning_rate": 1.57702850877193e-05,
+      "loss": 1.4736,
+      "step": 4629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.480546474456787,
+      "learning_rate": 1.576937134502924e-05,
+      "loss": 1.374,
+      "step": 4630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3656039237976074,
+      "learning_rate": 1.576845760233918e-05,
+      "loss": 1.9093,
+      "step": 4631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0903360843658447,
+      "learning_rate": 1.5767543859649122e-05,
+      "loss": 1.6286,
+      "step": 4632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1825952529907227,
+      "learning_rate": 1.5766630116959066e-05,
+      "loss": 2.1152,
+      "step": 4633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.741260051727295,
+      "learning_rate": 1.5765716374269007e-05,
+      "loss": 1.6938,
+      "step": 4634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7461631298065186,
+      "learning_rate": 1.5764802631578947e-05,
+      "loss": 1.6685,
+      "step": 4635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1095352172851562,
+      "learning_rate": 1.576388888888889e-05,
+      "loss": 1.7912,
+      "step": 4636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.413888931274414,
+      "learning_rate": 1.576297514619883e-05,
+      "loss": 1.6958,
+      "step": 4637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.562039375305176,
+      "learning_rate": 1.5762061403508772e-05,
+      "loss": 1.4649,
+      "step": 4638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8887503147125244,
+      "learning_rate": 1.5761147660818713e-05,
+      "loss": 1.7176,
+      "step": 4639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.896366596221924,
+      "learning_rate": 1.5760233918128657e-05,
+      "loss": 1.6787,
+      "step": 4640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.032025098800659,
+      "learning_rate": 1.5759320175438597e-05,
+      "loss": 1.3959,
+      "step": 4641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2364370822906494,
+      "learning_rate": 1.5758406432748538e-05,
+      "loss": 1.712,
+      "step": 4642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.815758228302002,
+      "learning_rate": 1.575749269005848e-05,
+      "loss": 1.72,
+      "step": 4643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1245288848876953,
+      "learning_rate": 1.5756578947368422e-05,
+      "loss": 1.8095,
+      "step": 4644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.079111099243164,
+      "learning_rate": 1.5755665204678363e-05,
+      "loss": 1.8791,
+      "step": 4645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0052103996276855,
+      "learning_rate": 1.5754751461988303e-05,
+      "loss": 1.4567,
+      "step": 4646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.807241916656494,
+      "learning_rate": 1.5753837719298247e-05,
+      "loss": 1.6406,
+      "step": 4647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.77191162109375,
+      "learning_rate": 1.5752923976608188e-05,
+      "loss": 1.8681,
+      "step": 4648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.410060167312622,
+      "learning_rate": 1.5752010233918128e-05,
+      "loss": 1.6766,
+      "step": 4649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1398227214813232,
+      "learning_rate": 1.5751096491228072e-05,
+      "loss": 1.6847,
+      "step": 4650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.309185743331909,
+      "learning_rate": 1.5750182748538013e-05,
+      "loss": 1.2907,
+      "step": 4651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0625622272491455,
+      "learning_rate": 1.5749269005847953e-05,
+      "loss": 1.834,
+      "step": 4652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0788981914520264,
+      "learning_rate": 1.5748355263157894e-05,
+      "loss": 1.8689,
+      "step": 4653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.15027117729187,
+      "learning_rate": 1.5747441520467838e-05,
+      "loss": 1.8534,
+      "step": 4654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5741934776306152,
+      "learning_rate": 1.5746527777777778e-05,
+      "loss": 1.6653,
+      "step": 4655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.511821746826172,
+      "learning_rate": 1.574561403508772e-05,
+      "loss": 1.3334,
+      "step": 4656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4119760990142822,
+      "learning_rate": 1.5744700292397662e-05,
+      "loss": 1.9228,
+      "step": 4657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.972257375717163,
+      "learning_rate": 1.5743786549707603e-05,
+      "loss": 1.6534,
+      "step": 4658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0145914554595947,
+      "learning_rate": 1.5742872807017544e-05,
+      "loss": 1.8274,
+      "step": 4659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5690014362335205,
+      "learning_rate": 1.5741959064327487e-05,
+      "loss": 1.5514,
+      "step": 4660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8841793537139893,
+      "learning_rate": 1.5741045321637428e-05,
+      "loss": 1.8161,
+      "step": 4661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.137716054916382,
+      "learning_rate": 1.5740131578947372e-05,
+      "loss": 1.8573,
+      "step": 4662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.876774549484253,
+      "learning_rate": 1.5739217836257312e-05,
+      "loss": 1.547,
+      "step": 4663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.896462917327881,
+      "learning_rate": 1.5738304093567253e-05,
+      "loss": 1.315,
+      "step": 4664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.642777681350708,
+      "learning_rate": 1.5737390350877197e-05,
+      "loss": 2.5345,
+      "step": 4665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6083152294158936,
+      "learning_rate": 1.5736476608187137e-05,
+      "loss": 1.5277,
+      "step": 4666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.861556053161621,
+      "learning_rate": 1.5735562865497078e-05,
+      "loss": 1.7735,
+      "step": 4667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0706593990325928,
+      "learning_rate": 1.573464912280702e-05,
+      "loss": 1.6355,
+      "step": 4668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.792858839035034,
+      "learning_rate": 1.5733735380116962e-05,
+      "loss": 2.0316,
+      "step": 4669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.707589864730835,
+      "learning_rate": 1.5732821637426903e-05,
+      "loss": 1.5871,
+      "step": 4670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7160427570343018,
+      "learning_rate": 1.5731907894736843e-05,
+      "loss": 1.6775,
+      "step": 4671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.69624662399292,
+      "learning_rate": 1.5730994152046787e-05,
+      "loss": 1.4321,
+      "step": 4672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.949951648712158,
+      "learning_rate": 1.5730080409356728e-05,
+      "loss": 1.6099,
+      "step": 4673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.194641590118408,
+      "learning_rate": 1.572916666666667e-05,
+      "loss": 1.8915,
+      "step": 4674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9103095531463623,
+      "learning_rate": 1.572825292397661e-05,
+      "loss": 1.3673,
+      "step": 4675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7013256549835205,
+      "learning_rate": 1.5727339181286553e-05,
+      "loss": 1.7879,
+      "step": 4676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.647383213043213,
+      "learning_rate": 1.5726425438596493e-05,
+      "loss": 1.5888,
+      "step": 4677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.54247784614563,
+      "learning_rate": 1.5725511695906434e-05,
+      "loss": 1.3951,
+      "step": 4678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.436572313308716,
+      "learning_rate": 1.5724597953216378e-05,
+      "loss": 1.4112,
+      "step": 4679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.469869375228882,
+      "learning_rate": 1.572368421052632e-05,
+      "loss": 1.4765,
+      "step": 4680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6935536861419678,
+      "learning_rate": 1.572277046783626e-05,
+      "loss": 1.5975,
+      "step": 4681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6863350868225098,
+      "learning_rate": 1.57218567251462e-05,
+      "loss": 1.8406,
+      "step": 4682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.261455535888672,
+      "learning_rate": 1.5720942982456143e-05,
+      "loss": 1.113,
+      "step": 4683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0957229137420654,
+      "learning_rate": 1.5720029239766084e-05,
+      "loss": 1.4568,
+      "step": 4684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.568903684616089,
+      "learning_rate": 1.5719115497076024e-05,
+      "loss": 1.3158,
+      "step": 4685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.096583366394043,
+      "learning_rate": 1.5718201754385968e-05,
+      "loss": 1.633,
+      "step": 4686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4512484073638916,
+      "learning_rate": 1.571728801169591e-05,
+      "loss": 1.3737,
+      "step": 4687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.276453733444214,
+      "learning_rate": 1.571637426900585e-05,
+      "loss": 1.8138,
+      "step": 4688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9602675437927246,
+      "learning_rate": 1.571546052631579e-05,
+      "loss": 1.7975,
+      "step": 4689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3386576175689697,
+      "learning_rate": 1.5714546783625734e-05,
+      "loss": 2.0247,
+      "step": 4690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1745293140411377,
+      "learning_rate": 1.5713633040935674e-05,
+      "loss": 1.9176,
+      "step": 4691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.067033290863037,
+      "learning_rate": 1.5712719298245615e-05,
+      "loss": 1.5691,
+      "step": 4692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0986201763153076,
+      "learning_rate": 1.571180555555556e-05,
+      "loss": 1.7484,
+      "step": 4693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5861711502075195,
+      "learning_rate": 1.57108918128655e-05,
+      "loss": 1.5849,
+      "step": 4694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.872352123260498,
+      "learning_rate": 1.570997807017544e-05,
+      "loss": 2.1102,
+      "step": 4695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.124964952468872,
+      "learning_rate": 1.570906432748538e-05,
+      "loss": 1.6177,
+      "step": 4696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0315420627593994,
+      "learning_rate": 1.5708150584795324e-05,
+      "loss": 1.3745,
+      "step": 4697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.124243974685669,
+      "learning_rate": 1.5707236842105265e-05,
+      "loss": 1.6465,
+      "step": 4698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5868794918060303,
+      "learning_rate": 1.5706323099415205e-05,
+      "loss": 1.4009,
+      "step": 4699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4300522804260254,
+      "learning_rate": 1.570540935672515e-05,
+      "loss": 2.0515,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9083077907562256,
+      "learning_rate": 1.570449561403509e-05,
+      "loss": 1.6642,
+      "step": 4701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.193276882171631,
+      "learning_rate": 1.570358187134503e-05,
+      "loss": 2.0589,
+      "step": 4702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5795648097991943,
+      "learning_rate": 1.570266812865497e-05,
+      "loss": 1.4139,
+      "step": 4703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.412381887435913,
+      "learning_rate": 1.5701754385964915e-05,
+      "loss": 1.89,
+      "step": 4704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.691423177719116,
+      "learning_rate": 1.5700840643274855e-05,
+      "loss": 1.5274,
+      "step": 4705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.365316152572632,
+      "learning_rate": 1.5699926900584796e-05,
+      "loss": 1.843,
+      "step": 4706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4562275409698486,
+      "learning_rate": 1.569901315789474e-05,
+      "loss": 1.6102,
+      "step": 4707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.904341697692871,
+      "learning_rate": 1.569809941520468e-05,
+      "loss": 1.7404,
+      "step": 4708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.242997407913208,
+      "learning_rate": 1.569718567251462e-05,
+      "loss": 2.0976,
+      "step": 4709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7067883014678955,
+      "learning_rate": 1.569627192982456e-05,
+      "loss": 1.4952,
+      "step": 4710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.859494924545288,
+      "learning_rate": 1.5695358187134505e-05,
+      "loss": 1.4717,
+      "step": 4711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6182756423950195,
+      "learning_rate": 1.5694444444444446e-05,
+      "loss": 1.7566,
+      "step": 4712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6446359157562256,
+      "learning_rate": 1.5693530701754386e-05,
+      "loss": 1.5018,
+      "step": 4713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0960805416107178,
+      "learning_rate": 1.569261695906433e-05,
+      "loss": 1.7481,
+      "step": 4714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9004015922546387,
+      "learning_rate": 1.569170321637427e-05,
+      "loss": 1.5715,
+      "step": 4715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9998719692230225,
+      "learning_rate": 1.569078947368421e-05,
+      "loss": 1.5639,
+      "step": 4716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3744261264801025,
+      "learning_rate": 1.5689875730994152e-05,
+      "loss": 1.9836,
+      "step": 4717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.706223726272583,
+      "learning_rate": 1.5688961988304096e-05,
+      "loss": 1.3911,
+      "step": 4718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.286738634109497,
+      "learning_rate": 1.5688048245614036e-05,
+      "loss": 1.7106,
+      "step": 4719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.120374917984009,
+      "learning_rate": 1.5687134502923977e-05,
+      "loss": 1.9259,
+      "step": 4720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.298659563064575,
+      "learning_rate": 1.568622076023392e-05,
+      "loss": 1.3145,
+      "step": 4721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1304056644439697,
+      "learning_rate": 1.568530701754386e-05,
+      "loss": 1.8317,
+      "step": 4722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.644009828567505,
+      "learning_rate": 1.56843932748538e-05,
+      "loss": 1.4201,
+      "step": 4723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7063405513763428,
+      "learning_rate": 1.5683479532163742e-05,
+      "loss": 1.5195,
+      "step": 4724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.734212636947632,
+      "learning_rate": 1.5682565789473686e-05,
+      "loss": 1.9147,
+      "step": 4725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0494940280914307,
+      "learning_rate": 1.5681652046783627e-05,
+      "loss": 1.8984,
+      "step": 4726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.500028610229492,
+      "learning_rate": 1.5680738304093567e-05,
+      "loss": 1.5923,
+      "step": 4727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0826518535614014,
+      "learning_rate": 1.567982456140351e-05,
+      "loss": 1.8413,
+      "step": 4728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.277522087097168,
+      "learning_rate": 1.567891081871345e-05,
+      "loss": 1.7361,
+      "step": 4729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8833224773406982,
+      "learning_rate": 1.5677997076023392e-05,
+      "loss": 1.639,
+      "step": 4730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.541712760925293,
+      "learning_rate": 1.5677083333333333e-05,
+      "loss": 2.0952,
+      "step": 4731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9483325481414795,
+      "learning_rate": 1.5676169590643277e-05,
+      "loss": 1.7462,
+      "step": 4732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7089285850524902,
+      "learning_rate": 1.5675255847953217e-05,
+      "loss": 1.3466,
+      "step": 4733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2471137046813965,
+      "learning_rate": 1.5674342105263158e-05,
+      "loss": 1.5349,
+      "step": 4734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8236749172210693,
+      "learning_rate": 1.5673428362573098e-05,
+      "loss": 1.5872,
+      "step": 4735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.092465400695801,
+      "learning_rate": 1.5672514619883042e-05,
+      "loss": 1.6367,
+      "step": 4736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6031644344329834,
+      "learning_rate": 1.5671600877192983e-05,
+      "loss": 1.6169,
+      "step": 4737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8120534420013428,
+      "learning_rate": 1.5670687134502923e-05,
+      "loss": 1.7046,
+      "step": 4738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.975604295730591,
+      "learning_rate": 1.5669773391812867e-05,
+      "loss": 1.6501,
+      "step": 4739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.108663558959961,
+      "learning_rate": 1.5668859649122808e-05,
+      "loss": 1.6855,
+      "step": 4740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0201168060302734,
+      "learning_rate": 1.5667945906432748e-05,
+      "loss": 1.7814,
+      "step": 4741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6907591819763184,
+      "learning_rate": 1.566703216374269e-05,
+      "loss": 1.4366,
+      "step": 4742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.027693748474121,
+      "learning_rate": 1.5666118421052633e-05,
+      "loss": 1.9576,
+      "step": 4743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1577627658843994,
+      "learning_rate": 1.5665204678362573e-05,
+      "loss": 1.6888,
+      "step": 4744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9321448802948,
+      "learning_rate": 1.5664290935672514e-05,
+      "loss": 1.635,
+      "step": 4745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.235653877258301,
+      "learning_rate": 1.5663377192982458e-05,
+      "loss": 1.8575,
+      "step": 4746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4750306606292725,
+      "learning_rate": 1.56624634502924e-05,
+      "loss": 1.3297,
+      "step": 4747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.640348196029663,
+      "learning_rate": 1.5661549707602342e-05,
+      "loss": 1.7417,
+      "step": 4748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3029592037200928,
+      "learning_rate": 1.5660635964912282e-05,
+      "loss": 1.8379,
+      "step": 4749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.006809711456299,
+      "learning_rate": 1.5659722222222223e-05,
+      "loss": 1.7452,
+      "step": 4750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0070769786834717,
+      "learning_rate": 1.5658808479532167e-05,
+      "loss": 1.8468,
+      "step": 4751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7470719814300537,
+      "learning_rate": 1.5657894736842107e-05,
+      "loss": 1.5155,
+      "step": 4752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8003921508789062,
+      "learning_rate": 1.5656980994152048e-05,
+      "loss": 1.3418,
+      "step": 4753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.684035301208496,
+      "learning_rate": 1.5656067251461992e-05,
+      "loss": 1.3618,
+      "step": 4754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6463117599487305,
+      "learning_rate": 1.5655153508771932e-05,
+      "loss": 1.3249,
+      "step": 4755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.765740156173706,
+      "learning_rate": 1.5654239766081873e-05,
+      "loss": 1.8297,
+      "step": 4756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.081480026245117,
+      "learning_rate": 1.5653326023391813e-05,
+      "loss": 2.063,
+      "step": 4757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0070929527282715,
+      "learning_rate": 1.5652412280701757e-05,
+      "loss": 1.5906,
+      "step": 4758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.161302089691162,
+      "learning_rate": 1.5651498538011698e-05,
+      "loss": 2.1838,
+      "step": 4759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4560866355895996,
+      "learning_rate": 1.565058479532164e-05,
+      "loss": 1.5433,
+      "step": 4760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7607882022857666,
+      "learning_rate": 1.5649671052631582e-05,
+      "loss": 1.8435,
+      "step": 4761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9382505416870117,
+      "learning_rate": 1.5648757309941523e-05,
+      "loss": 1.4727,
+      "step": 4762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6912968158721924,
+      "learning_rate": 1.5647843567251463e-05,
+      "loss": 1.6562,
+      "step": 4763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5928382873535156,
+      "learning_rate": 1.5646929824561404e-05,
+      "loss": 1.4105,
+      "step": 4764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3170735836029053,
+      "learning_rate": 1.5646016081871348e-05,
+      "loss": 1.7736,
+      "step": 4765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.91703724861145,
+      "learning_rate": 1.564510233918129e-05,
+      "loss": 1.7347,
+      "step": 4766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1866538524627686,
+      "learning_rate": 1.564418859649123e-05,
+      "loss": 2.0805,
+      "step": 4767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.120358943939209,
+      "learning_rate": 1.5643274853801173e-05,
+      "loss": 1.7441,
+      "step": 4768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6975796222686768,
+      "learning_rate": 1.5642361111111113e-05,
+      "loss": 1.2748,
+      "step": 4769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8925883769989014,
+      "learning_rate": 1.5641447368421054e-05,
+      "loss": 1.8133,
+      "step": 4770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.961229085922241,
+      "learning_rate": 1.5640533625730994e-05,
+      "loss": 1.6639,
+      "step": 4771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0635390281677246,
+      "learning_rate": 1.563961988304094e-05,
+      "loss": 1.6956,
+      "step": 4772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.182980537414551,
+      "learning_rate": 1.563870614035088e-05,
+      "loss": 2.0622,
+      "step": 4773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6808767318725586,
+      "learning_rate": 1.563779239766082e-05,
+      "loss": 1.5536,
+      "step": 4774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7895541191101074,
+      "learning_rate": 1.5636878654970763e-05,
+      "loss": 1.5704,
+      "step": 4775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5948426723480225,
+      "learning_rate": 1.5635964912280704e-05,
+      "loss": 1.9765,
+      "step": 4776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.891986846923828,
+      "learning_rate": 1.5635051169590644e-05,
+      "loss": 1.6055,
+      "step": 4777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9161438941955566,
+      "learning_rate": 1.5634137426900585e-05,
+      "loss": 1.7612,
+      "step": 4778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9897217750549316,
+      "learning_rate": 1.563322368421053e-05,
+      "loss": 1.7825,
+      "step": 4779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.778377056121826,
+      "learning_rate": 1.563230994152047e-05,
+      "loss": 1.6668,
+      "step": 4780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.007204294204712,
+      "learning_rate": 1.563139619883041e-05,
+      "loss": 1.9005,
+      "step": 4781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.950641393661499,
+      "learning_rate": 1.5630482456140354e-05,
+      "loss": 1.7835,
+      "step": 4782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.445819616317749,
+      "learning_rate": 1.5629568713450294e-05,
+      "loss": 1.9566,
+      "step": 4783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.723961114883423,
+      "learning_rate": 1.5628654970760235e-05,
+      "loss": 1.6757,
+      "step": 4784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.738471269607544,
+      "learning_rate": 1.5627741228070175e-05,
+      "loss": 1.4873,
+      "step": 4785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.501117706298828,
+      "learning_rate": 1.562682748538012e-05,
+      "loss": 1.7514,
+      "step": 4786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.302816867828369,
+      "learning_rate": 1.562591374269006e-05,
+      "loss": 1.9942,
+      "step": 4787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.846230983734131,
+      "learning_rate": 1.5625e-05,
+      "loss": 1.7658,
+      "step": 4788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.934565305709839,
+      "learning_rate": 1.5624086257309944e-05,
+      "loss": 1.384,
+      "step": 4789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.35151743888855,
+      "learning_rate": 1.5623172514619885e-05,
+      "loss": 1.6792,
+      "step": 4790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.682515859603882,
+      "learning_rate": 1.5622258771929825e-05,
+      "loss": 1.5061,
+      "step": 4791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3980050086975098,
+      "learning_rate": 1.5621345029239766e-05,
+      "loss": 1.5664,
+      "step": 4792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.773832082748413,
+      "learning_rate": 1.562043128654971e-05,
+      "loss": 1.6759,
+      "step": 4793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.843050241470337,
+      "learning_rate": 1.561951754385965e-05,
+      "loss": 2.3293,
+      "step": 4794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.532238483428955,
+      "learning_rate": 1.561860380116959e-05,
+      "loss": 1.4183,
+      "step": 4795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.865839719772339,
+      "learning_rate": 1.5617690058479535e-05,
+      "loss": 1.6965,
+      "step": 4796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8788552284240723,
+      "learning_rate": 1.5616776315789475e-05,
+      "loss": 1.2733,
+      "step": 4797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5607223510742188,
+      "learning_rate": 1.5615862573099416e-05,
+      "loss": 1.5912,
+      "step": 4798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.77599835395813,
+      "learning_rate": 1.5614948830409356e-05,
+      "loss": 1.8807,
+      "step": 4799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.697467803955078,
+      "learning_rate": 1.56140350877193e-05,
+      "loss": 1.945,
+      "step": 4800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7953455448150635,
+      "learning_rate": 1.561312134502924e-05,
+      "loss": 1.3486,
+      "step": 4801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6689095497131348,
+      "learning_rate": 1.561220760233918e-05,
+      "loss": 1.5857,
+      "step": 4802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.789794921875,
+      "learning_rate": 1.5611293859649125e-05,
+      "loss": 1.8104,
+      "step": 4803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.997546672821045,
+      "learning_rate": 1.5610380116959066e-05,
+      "loss": 1.549,
+      "step": 4804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.539017677307129,
+      "learning_rate": 1.5609466374269006e-05,
+      "loss": 1.921,
+      "step": 4805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7901275157928467,
+      "learning_rate": 1.5608552631578947e-05,
+      "loss": 1.6354,
+      "step": 4806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5366859436035156,
+      "learning_rate": 1.560763888888889e-05,
+      "loss": 1.8759,
+      "step": 4807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.736224889755249,
+      "learning_rate": 1.560672514619883e-05,
+      "loss": 1.6658,
+      "step": 4808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.072404623031616,
+      "learning_rate": 1.5605811403508772e-05,
+      "loss": 1.7072,
+      "step": 4809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1883509159088135,
+      "learning_rate": 1.5604897660818716e-05,
+      "loss": 2.0447,
+      "step": 4810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0029921531677246,
+      "learning_rate": 1.5603983918128656e-05,
+      "loss": 1.6436,
+      "step": 4811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4719021320343018,
+      "learning_rate": 1.5603070175438597e-05,
+      "loss": 1.6575,
+      "step": 4812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.88631272315979,
+      "learning_rate": 1.5602156432748537e-05,
+      "loss": 1.7443,
+      "step": 4813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.739619493484497,
+      "learning_rate": 1.560124269005848e-05,
+      "loss": 1.6354,
+      "step": 4814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.215144634246826,
+      "learning_rate": 1.560032894736842e-05,
+      "loss": 1.7099,
+      "step": 4815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.48648738861084,
+      "learning_rate": 1.5599415204678362e-05,
+      "loss": 1.4902,
+      "step": 4816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.089094638824463,
+      "learning_rate": 1.5598501461988306e-05,
+      "loss": 1.8126,
+      "step": 4817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.929959774017334,
+      "learning_rate": 1.5597587719298247e-05,
+      "loss": 1.8478,
+      "step": 4818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8088228702545166,
+      "learning_rate": 1.5596673976608187e-05,
+      "loss": 1.4922,
+      "step": 4819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8208870887756348,
+      "learning_rate": 1.5595760233918128e-05,
+      "loss": 1.5606,
+      "step": 4820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6114935874938965,
+      "learning_rate": 1.559484649122807e-05,
+      "loss": 1.8229,
+      "step": 4821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1699180603027344,
+      "learning_rate": 1.5593932748538012e-05,
+      "loss": 1.7264,
+      "step": 4822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4615368843078613,
+      "learning_rate": 1.5593019005847953e-05,
+      "loss": 1.1756,
+      "step": 4823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.048421859741211,
+      "learning_rate": 1.5592105263157897e-05,
+      "loss": 1.912,
+      "step": 4824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.93373703956604,
+      "learning_rate": 1.5591191520467837e-05,
+      "loss": 1.4975,
+      "step": 4825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6321022510528564,
+      "learning_rate": 1.5590277777777778e-05,
+      "loss": 1.5373,
+      "step": 4826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8432281017303467,
+      "learning_rate": 1.5589364035087718e-05,
+      "loss": 1.9477,
+      "step": 4827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.531209945678711,
+      "learning_rate": 1.5588450292397662e-05,
+      "loss": 1.4394,
+      "step": 4828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9780428409576416,
+      "learning_rate": 1.5587536549707603e-05,
+      "loss": 1.6335,
+      "step": 4829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.370763063430786,
+      "learning_rate": 1.5586622807017543e-05,
+      "loss": 1.4414,
+      "step": 4830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2067198753356934,
+      "learning_rate": 1.5585709064327487e-05,
+      "loss": 1.6732,
+      "step": 4831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3916947841644287,
+      "learning_rate": 1.5584795321637428e-05,
+      "loss": 1.1768,
+      "step": 4832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.818880081176758,
+      "learning_rate": 1.558388157894737e-05,
+      "loss": 1.4617,
+      "step": 4833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0769920349121094,
+      "learning_rate": 1.5582967836257312e-05,
+      "loss": 1.7715,
+      "step": 4834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.87927508354187,
+      "learning_rate": 1.5582054093567253e-05,
+      "loss": 1.4434,
+      "step": 4835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4853994846343994,
+      "learning_rate": 1.5581140350877196e-05,
+      "loss": 1.7656,
+      "step": 4836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.562906265258789,
+      "learning_rate": 1.5580226608187137e-05,
+      "loss": 1.9012,
+      "step": 4837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9544246196746826,
+      "learning_rate": 1.5579312865497078e-05,
+      "loss": 1.6618,
+      "step": 4838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.744149923324585,
+      "learning_rate": 1.5578399122807018e-05,
+      "loss": 1.6171,
+      "step": 4839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9377667903900146,
+      "learning_rate": 1.5577485380116962e-05,
+      "loss": 1.4222,
+      "step": 4840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9224965572357178,
+      "learning_rate": 1.5576571637426902e-05,
+      "loss": 1.584,
+      "step": 4841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.762390613555908,
+      "learning_rate": 1.5575657894736843e-05,
+      "loss": 1.4402,
+      "step": 4842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8707096576690674,
+      "learning_rate": 1.5574744152046787e-05,
+      "loss": 1.4766,
+      "step": 4843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.912248373031616,
+      "learning_rate": 1.5573830409356727e-05,
+      "loss": 1.66,
+      "step": 4844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9477474689483643,
+      "learning_rate": 1.5572916666666668e-05,
+      "loss": 1.6007,
+      "step": 4845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9745798110961914,
+      "learning_rate": 1.557200292397661e-05,
+      "loss": 1.5659,
+      "step": 4846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2967007160186768,
+      "learning_rate": 1.5571089181286552e-05,
+      "loss": 1.3505,
+      "step": 4847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.034245491027832,
+      "learning_rate": 1.5570175438596493e-05,
+      "loss": 1.4189,
+      "step": 4848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1182682514190674,
+      "learning_rate": 1.5569261695906433e-05,
+      "loss": 1.4479,
+      "step": 4849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.168952465057373,
+      "learning_rate": 1.5568347953216377e-05,
+      "loss": 1.7732,
+      "step": 4850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3150904178619385,
+      "learning_rate": 1.5567434210526318e-05,
+      "loss": 1.5933,
+      "step": 4851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8732409477233887,
+      "learning_rate": 1.556652046783626e-05,
+      "loss": 1.6451,
+      "step": 4852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.652301073074341,
+      "learning_rate": 1.55656067251462e-05,
+      "loss": 1.4655,
+      "step": 4853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.125321388244629,
+      "learning_rate": 1.5564692982456143e-05,
+      "loss": 1.6095,
+      "step": 4854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.925379991531372,
+      "learning_rate": 1.5563779239766083e-05,
+      "loss": 2.2651,
+      "step": 4855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0908334255218506,
+      "learning_rate": 1.5562865497076024e-05,
+      "loss": 1.5617,
+      "step": 4856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9291698932647705,
+      "learning_rate": 1.5561951754385968e-05,
+      "loss": 1.932,
+      "step": 4857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.093365430831909,
+      "learning_rate": 1.556103801169591e-05,
+      "loss": 1.5194,
+      "step": 4858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.093942642211914,
+      "learning_rate": 1.556012426900585e-05,
+      "loss": 1.7031,
+      "step": 4859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2472949028015137,
+      "learning_rate": 1.555921052631579e-05,
+      "loss": 1.326,
+      "step": 4860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1328868865966797,
+      "learning_rate": 1.5558296783625733e-05,
+      "loss": 1.6962,
+      "step": 4861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9816079139709473,
+      "learning_rate": 1.5557383040935674e-05,
+      "loss": 1.7969,
+      "step": 4862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7618167400360107,
+      "learning_rate": 1.5556469298245614e-05,
+      "loss": 1.6883,
+      "step": 4863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5773749351501465,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 2.0099,
+      "step": 4864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.247504711151123,
+      "learning_rate": 1.55546418128655e-05,
+      "loss": 1.8605,
+      "step": 4865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9291188716888428,
+      "learning_rate": 1.555372807017544e-05,
+      "loss": 1.8736,
+      "step": 4866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7380478382110596,
+      "learning_rate": 1.555281432748538e-05,
+      "loss": 1.527,
+      "step": 4867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.970156669616699,
+      "learning_rate": 1.5551900584795324e-05,
+      "loss": 2.0017,
+      "step": 4868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.654865264892578,
+      "learning_rate": 1.5550986842105264e-05,
+      "loss": 2.117,
+      "step": 4869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2305126190185547,
+      "learning_rate": 1.5550073099415205e-05,
+      "loss": 1.4608,
+      "step": 4870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6727981567382812,
+      "learning_rate": 1.554915935672515e-05,
+      "loss": 1.7876,
+      "step": 4871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0256662368774414,
+      "learning_rate": 1.554824561403509e-05,
+      "loss": 1.4969,
+      "step": 4872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5021533966064453,
+      "learning_rate": 1.554733187134503e-05,
+      "loss": 1.3866,
+      "step": 4873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.978728771209717,
+      "learning_rate": 1.554641812865497e-05,
+      "loss": 1.6056,
+      "step": 4874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7683145999908447,
+      "learning_rate": 1.5545504385964914e-05,
+      "loss": 1.5904,
+      "step": 4875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.164928913116455,
+      "learning_rate": 1.5544590643274855e-05,
+      "loss": 1.5188,
+      "step": 4876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9274981021881104,
+      "learning_rate": 1.5543676900584795e-05,
+      "loss": 1.5317,
+      "step": 4877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8331661224365234,
+      "learning_rate": 1.554276315789474e-05,
+      "loss": 1.7003,
+      "step": 4878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.879101514816284,
+      "learning_rate": 1.554184941520468e-05,
+      "loss": 1.6712,
+      "step": 4879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.93538236618042,
+      "learning_rate": 1.554093567251462e-05,
+      "loss": 1.8575,
+      "step": 4880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.238291025161743,
+      "learning_rate": 1.554002192982456e-05,
+      "loss": 1.9826,
+      "step": 4881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.223165988922119,
+      "learning_rate": 1.5539108187134505e-05,
+      "loss": 1.7774,
+      "step": 4882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8770718574523926,
+      "learning_rate": 1.5538194444444445e-05,
+      "loss": 1.9863,
+      "step": 4883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6853177547454834,
+      "learning_rate": 1.5537280701754386e-05,
+      "loss": 1.6281,
+      "step": 4884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3584437370300293,
+      "learning_rate": 1.553636695906433e-05,
+      "loss": 1.4118,
+      "step": 4885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6908373832702637,
+      "learning_rate": 1.553545321637427e-05,
+      "loss": 1.7084,
+      "step": 4886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7132034301757812,
+      "learning_rate": 1.553453947368421e-05,
+      "loss": 1.6267,
+      "step": 4887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2487809658050537,
+      "learning_rate": 1.553362573099415e-05,
+      "loss": 2.0602,
+      "step": 4888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4678564071655273,
+      "learning_rate": 1.5532711988304095e-05,
+      "loss": 1.7053,
+      "step": 4889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9739623069763184,
+      "learning_rate": 1.5531798245614036e-05,
+      "loss": 1.7946,
+      "step": 4890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.870793342590332,
+      "learning_rate": 1.5530884502923976e-05,
+      "loss": 1.8059,
+      "step": 4891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9794723987579346,
+      "learning_rate": 1.552997076023392e-05,
+      "loss": 1.6311,
+      "step": 4892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1480185985565186,
+      "learning_rate": 1.552905701754386e-05,
+      "loss": 1.7627,
+      "step": 4893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.552929401397705,
+      "learning_rate": 1.55281432748538e-05,
+      "loss": 1.369,
+      "step": 4894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3877127170562744,
+      "learning_rate": 1.5527229532163742e-05,
+      "loss": 2.0336,
+      "step": 4895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.982271671295166,
+      "learning_rate": 1.5526315789473686e-05,
+      "loss": 1.6908,
+      "step": 4896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.085123062133789,
+      "learning_rate": 1.5525402046783626e-05,
+      "loss": 1.8569,
+      "step": 4897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.711656332015991,
+      "learning_rate": 1.5524488304093567e-05,
+      "loss": 1.4365,
+      "step": 4898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.73016357421875,
+      "learning_rate": 1.552357456140351e-05,
+      "loss": 2.2054,
+      "step": 4899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.894793748855591,
+      "learning_rate": 1.552266081871345e-05,
+      "loss": 1.6236,
+      "step": 4900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0049948692321777,
+      "learning_rate": 1.5521747076023392e-05,
+      "loss": 1.896,
+      "step": 4901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7240395545959473,
+      "learning_rate": 1.5520833333333332e-05,
+      "loss": 1.7501,
+      "step": 4902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5520923137664795,
+      "learning_rate": 1.5519919590643276e-05,
+      "loss": 2.2909,
+      "step": 4903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5361998081207275,
+      "learning_rate": 1.5519005847953217e-05,
+      "loss": 1.4217,
+      "step": 4904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.089050054550171,
+      "learning_rate": 1.5518092105263157e-05,
+      "loss": 1.6592,
+      "step": 4905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0498039722442627,
+      "learning_rate": 1.55171783625731e-05,
+      "loss": 1.9527,
+      "step": 4906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9979653358459473,
+      "learning_rate": 1.551626461988304e-05,
+      "loss": 1.7749,
+      "step": 4907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8656551837921143,
+      "learning_rate": 1.5515350877192982e-05,
+      "loss": 2.1243,
+      "step": 4908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.867556095123291,
+      "learning_rate": 1.5514437134502923e-05,
+      "loss": 1.6741,
+      "step": 4909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.031636953353882,
+      "learning_rate": 1.5513523391812867e-05,
+      "loss": 1.7069,
+      "step": 4910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4977269172668457,
+      "learning_rate": 1.5512609649122807e-05,
+      "loss": 1.5115,
+      "step": 4911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8096399307250977,
+      "learning_rate": 1.5511695906432748e-05,
+      "loss": 1.3046,
+      "step": 4912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5248782634735107,
+      "learning_rate": 1.551078216374269e-05,
+      "loss": 1.4202,
+      "step": 4913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4441542625427246,
+      "learning_rate": 1.5509868421052632e-05,
+      "loss": 1.9247,
+      "step": 4914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0633246898651123,
+      "learning_rate": 1.5508954678362573e-05,
+      "loss": 1.9917,
+      "step": 4915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.162158489227295,
+      "learning_rate": 1.5508040935672513e-05,
+      "loss": 1.5933,
+      "step": 4916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3218421936035156,
+      "learning_rate": 1.5507127192982457e-05,
+      "loss": 1.5695,
+      "step": 4917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5381505489349365,
+      "learning_rate": 1.55062134502924e-05,
+      "loss": 1.4265,
+      "step": 4918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.198403835296631,
+      "learning_rate": 1.550529970760234e-05,
+      "loss": 2.0644,
+      "step": 4919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.34541654586792,
+      "learning_rate": 1.5504385964912282e-05,
+      "loss": 2.0047,
+      "step": 4920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6786694526672363,
+      "learning_rate": 1.5503472222222226e-05,
+      "loss": 1.5281,
+      "step": 4921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8028197288513184,
+      "learning_rate": 1.5502558479532167e-05,
+      "loss": 1.4449,
+      "step": 4922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9855473041534424,
+      "learning_rate": 1.5501644736842107e-05,
+      "loss": 1.7689,
+      "step": 4923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6753554344177246,
+      "learning_rate": 1.5500730994152048e-05,
+      "loss": 1.6495,
+      "step": 4924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.898066282272339,
+      "learning_rate": 1.549981725146199e-05,
+      "loss": 1.8136,
+      "step": 4925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5423529148101807,
+      "learning_rate": 1.5498903508771932e-05,
+      "loss": 1.4627,
+      "step": 4926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7315866947174072,
+      "learning_rate": 1.5497989766081873e-05,
+      "loss": 1.5796,
+      "step": 4927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.064008951187134,
+      "learning_rate": 1.5497076023391816e-05,
+      "loss": 1.6828,
+      "step": 4928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7146639823913574,
+      "learning_rate": 1.5496162280701757e-05,
+      "loss": 1.3434,
+      "step": 4929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.641686201095581,
+      "learning_rate": 1.5495248538011698e-05,
+      "loss": 1.8381,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.676365375518799,
+      "learning_rate": 1.5494334795321638e-05,
+      "loss": 1.3439,
+      "step": 4931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.727055788040161,
+      "learning_rate": 1.5493421052631582e-05,
+      "loss": 1.4581,
+      "step": 4932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.230264186859131,
+      "learning_rate": 1.5492507309941522e-05,
+      "loss": 1.5483,
+      "step": 4933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8708181381225586,
+      "learning_rate": 1.5491593567251463e-05,
+      "loss": 1.4452,
+      "step": 4934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6363162994384766,
+      "learning_rate": 1.5490679824561407e-05,
+      "loss": 1.5035,
+      "step": 4935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.071805477142334,
+      "learning_rate": 1.5489766081871347e-05,
+      "loss": 1.7226,
+      "step": 4936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.085858106613159,
+      "learning_rate": 1.5488852339181288e-05,
+      "loss": 1.4879,
+      "step": 4937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.563526153564453,
+      "learning_rate": 1.548793859649123e-05,
+      "loss": 1.5714,
+      "step": 4938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7446906566619873,
+      "learning_rate": 1.5487024853801172e-05,
+      "loss": 1.6276,
+      "step": 4939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0959322452545166,
+      "learning_rate": 1.5486111111111113e-05,
+      "loss": 1.8119,
+      "step": 4940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8383209705352783,
+      "learning_rate": 1.5485197368421053e-05,
+      "loss": 1.4812,
+      "step": 4941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7139432430267334,
+      "learning_rate": 1.5484283625730997e-05,
+      "loss": 1.4732,
+      "step": 4942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.665698766708374,
+      "learning_rate": 1.5483369883040938e-05,
+      "loss": 1.8187,
+      "step": 4943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.386328935623169,
+      "learning_rate": 1.548245614035088e-05,
+      "loss": 1.5436,
+      "step": 4944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.747189998626709,
+      "learning_rate": 1.548154239766082e-05,
+      "loss": 1.8679,
+      "step": 4945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2078604698181152,
+      "learning_rate": 1.5480628654970763e-05,
+      "loss": 1.7878,
+      "step": 4946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3366646766662598,
+      "learning_rate": 1.5479714912280703e-05,
+      "loss": 1.3665,
+      "step": 4947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.684182643890381,
+      "learning_rate": 1.5478801169590644e-05,
+      "loss": 1.5371,
+      "step": 4948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.918748378753662,
+      "learning_rate": 1.5477887426900588e-05,
+      "loss": 1.8865,
+      "step": 4949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1220450401306152,
+      "learning_rate": 1.547697368421053e-05,
+      "loss": 2.0244,
+      "step": 4950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.763571262359619,
+      "learning_rate": 1.547605994152047e-05,
+      "loss": 1.4439,
+      "step": 4951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5019736289978027,
+      "learning_rate": 1.547514619883041e-05,
+      "loss": 1.4808,
+      "step": 4952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8403804302215576,
+      "learning_rate": 1.5474232456140353e-05,
+      "loss": 1.6403,
+      "step": 4953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3372397422790527,
+      "learning_rate": 1.5473318713450294e-05,
+      "loss": 1.695,
+      "step": 4954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.103210687637329,
+      "learning_rate": 1.5472404970760234e-05,
+      "loss": 1.915,
+      "step": 4955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.622426748275757,
+      "learning_rate": 1.547149122807018e-05,
+      "loss": 1.4534,
+      "step": 4956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.360441207885742,
+      "learning_rate": 1.547057748538012e-05,
+      "loss": 1.7296,
+      "step": 4957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.388422727584839,
+      "learning_rate": 1.546966374269006e-05,
+      "loss": 1.1996,
+      "step": 4958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7679948806762695,
+      "learning_rate": 1.546875e-05,
+      "loss": 1.6657,
+      "step": 4959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0346426963806152,
+      "learning_rate": 1.5467836257309944e-05,
+      "loss": 1.7626,
+      "step": 4960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3688275814056396,
+      "learning_rate": 1.5466922514619884e-05,
+      "loss": 1.2979,
+      "step": 4961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5283071994781494,
+      "learning_rate": 1.5466008771929825e-05,
+      "loss": 1.4961,
+      "step": 4962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5404458045959473,
+      "learning_rate": 1.546509502923977e-05,
+      "loss": 1.861,
+      "step": 4963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.787226915359497,
+      "learning_rate": 1.546418128654971e-05,
+      "loss": 1.7915,
+      "step": 4964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.652961492538452,
+      "learning_rate": 1.546326754385965e-05,
+      "loss": 1.2263,
+      "step": 4965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.696200370788574,
+      "learning_rate": 1.546235380116959e-05,
+      "loss": 1.4188,
+      "step": 4966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.454693555831909,
+      "learning_rate": 1.5461440058479534e-05,
+      "loss": 1.7269,
+      "step": 4967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.509117364883423,
+      "learning_rate": 1.5460526315789475e-05,
+      "loss": 1.4788,
+      "step": 4968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0143074989318848,
+      "learning_rate": 1.5459612573099415e-05,
+      "loss": 1.8866,
+      "step": 4969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8382182121276855,
+      "learning_rate": 1.545869883040936e-05,
+      "loss": 1.6199,
+      "step": 4970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.349717855453491,
+      "learning_rate": 1.54577850877193e-05,
+      "loss": 1.3291,
+      "step": 4971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6588757038116455,
+      "learning_rate": 1.545687134502924e-05,
+      "loss": 1.6981,
+      "step": 4972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9402918815612793,
+      "learning_rate": 1.545595760233918e-05,
+      "loss": 1.8315,
+      "step": 4973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.518007755279541,
+      "learning_rate": 1.5455043859649125e-05,
+      "loss": 1.467,
+      "step": 4974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.489570140838623,
+      "learning_rate": 1.5454130116959065e-05,
+      "loss": 1.9867,
+      "step": 4975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.033050298690796,
+      "learning_rate": 1.5453216374269006e-05,
+      "loss": 1.7883,
+      "step": 4976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.766024112701416,
+      "learning_rate": 1.545230263157895e-05,
+      "loss": 1.599,
+      "step": 4977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.946197509765625,
+      "learning_rate": 1.545138888888889e-05,
+      "loss": 1.6446,
+      "step": 4978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4144930839538574,
+      "learning_rate": 1.545047514619883e-05,
+      "loss": 1.4564,
+      "step": 4979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.078728437423706,
+      "learning_rate": 1.544956140350877e-05,
+      "loss": 1.9701,
+      "step": 4980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.914295196533203,
+      "learning_rate": 1.5448647660818715e-05,
+      "loss": 1.646,
+      "step": 4981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7332847118377686,
+      "learning_rate": 1.5447733918128656e-05,
+      "loss": 1.6874,
+      "step": 4982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.236567735671997,
+      "learning_rate": 1.5446820175438596e-05,
+      "loss": 2.1183,
+      "step": 4983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.30000376701355,
+      "learning_rate": 1.544590643274854e-05,
+      "loss": 1.722,
+      "step": 4984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8506627082824707,
+      "learning_rate": 1.544499269005848e-05,
+      "loss": 1.7344,
+      "step": 4985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5847043991088867,
+      "learning_rate": 1.544407894736842e-05,
+      "loss": 1.5313,
+      "step": 4986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7519097328186035,
+      "learning_rate": 1.5443165204678362e-05,
+      "loss": 1.5813,
+      "step": 4987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3353939056396484,
+      "learning_rate": 1.5442251461988306e-05,
+      "loss": 1.5608,
+      "step": 4988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.359806776046753,
+      "learning_rate": 1.5441337719298246e-05,
+      "loss": 1.944,
+      "step": 4989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7071783542633057,
+      "learning_rate": 1.5440423976608187e-05,
+      "loss": 1.4055,
+      "step": 4990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.394477605819702,
+      "learning_rate": 1.543951023391813e-05,
+      "loss": 1.7437,
+      "step": 4991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8444817066192627,
+      "learning_rate": 1.543859649122807e-05,
+      "loss": 1.8142,
+      "step": 4992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.020051956176758,
+      "learning_rate": 1.5437682748538012e-05,
+      "loss": 2.2277,
+      "step": 4993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6714212894439697,
+      "learning_rate": 1.5436769005847952e-05,
+      "loss": 1.5957,
+      "step": 4994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7994678020477295,
+      "learning_rate": 1.5435855263157896e-05,
+      "loss": 1.6673,
+      "step": 4995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.804523229598999,
+      "learning_rate": 1.5434941520467837e-05,
+      "loss": 1.4779,
+      "step": 4996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2115604877471924,
+      "learning_rate": 1.5434027777777777e-05,
+      "loss": 1.4789,
+      "step": 4997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.337390184402466,
+      "learning_rate": 1.543311403508772e-05,
+      "loss": 1.8623,
+      "step": 4998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1034843921661377,
+      "learning_rate": 1.543220029239766e-05,
+      "loss": 1.6984,
+      "step": 4999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7036924362182617,
+      "learning_rate": 1.5431286549707602e-05,
+      "loss": 1.5477,
+      "step": 5000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0291831493377686,
+      "learning_rate": 1.5430372807017543e-05,
+      "loss": 1.7679,
+      "step": 5001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5623726844787598,
+      "learning_rate": 1.5429459064327487e-05,
+      "loss": 1.6549,
+      "step": 5002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.206385374069214,
+      "learning_rate": 1.542854532163743e-05,
+      "loss": 1.9045,
+      "step": 5003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.022303581237793,
+      "learning_rate": 1.542763157894737e-05,
+      "loss": 1.9694,
+      "step": 5004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.875985622406006,
+      "learning_rate": 1.542671783625731e-05,
+      "loss": 1.4334,
+      "step": 5005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1314809322357178,
+      "learning_rate": 1.5425804093567252e-05,
+      "loss": 2.0996,
+      "step": 5006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4322447776794434,
+      "learning_rate": 1.5424890350877196e-05,
+      "loss": 1.4291,
+      "step": 5007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.474940061569214,
+      "learning_rate": 1.5423976608187137e-05,
+      "loss": 1.876,
+      "step": 5008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2792811393737793,
+      "learning_rate": 1.5423062865497077e-05,
+      "loss": 1.7269,
+      "step": 5009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1618967056274414,
+      "learning_rate": 1.542214912280702e-05,
+      "loss": 1.7333,
+      "step": 5010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.342834234237671,
+      "learning_rate": 1.542123538011696e-05,
+      "loss": 1.5155,
+      "step": 5011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1381630897521973,
+      "learning_rate": 1.5420321637426902e-05,
+      "loss": 1.7088,
+      "step": 5012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.330033540725708,
+      "learning_rate": 1.5419407894736843e-05,
+      "loss": 1.43,
+      "step": 5013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.444673538208008,
+      "learning_rate": 1.5418494152046787e-05,
+      "loss": 1.3371,
+      "step": 5014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.650242805480957,
+      "learning_rate": 1.5417580409356727e-05,
+      "loss": 1.2456,
+      "step": 5015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5632193088531494,
+      "learning_rate": 1.5416666666666668e-05,
+      "loss": 1.6135,
+      "step": 5016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.51031494140625,
+      "learning_rate": 1.541575292397661e-05,
+      "loss": 1.2899,
+      "step": 5017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.190422773361206,
+      "learning_rate": 1.5414839181286552e-05,
+      "loss": 1.957,
+      "step": 5018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6784563064575195,
+      "learning_rate": 1.5413925438596493e-05,
+      "loss": 1.3002,
+      "step": 5019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.998429536819458,
+      "learning_rate": 1.5413011695906433e-05,
+      "loss": 1.9622,
+      "step": 5020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.469226360321045,
+      "learning_rate": 1.5412097953216377e-05,
+      "loss": 1.3024,
+      "step": 5021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3533685207366943,
+      "learning_rate": 1.5411184210526318e-05,
+      "loss": 1.3864,
+      "step": 5022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.162466526031494,
+      "learning_rate": 1.5410270467836258e-05,
+      "loss": 1.7812,
+      "step": 5023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8254802227020264,
+      "learning_rate": 1.5409356725146202e-05,
+      "loss": 1.4459,
+      "step": 5024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.187955617904663,
+      "learning_rate": 1.5408442982456142e-05,
+      "loss": 1.8613,
+      "step": 5025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9989302158355713,
+      "learning_rate": 1.5407529239766083e-05,
+      "loss": 1.7433,
+      "step": 5026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4944100379943848,
+      "learning_rate": 1.5406615497076024e-05,
+      "loss": 1.5487,
+      "step": 5027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.519733190536499,
+      "learning_rate": 1.5405701754385967e-05,
+      "loss": 1.1785,
+      "step": 5028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5422120094299316,
+      "learning_rate": 1.5404788011695908e-05,
+      "loss": 1.4287,
+      "step": 5029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.610335111618042,
+      "learning_rate": 1.540387426900585e-05,
+      "loss": 1.4214,
+      "step": 5030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8316690921783447,
+      "learning_rate": 1.5402960526315792e-05,
+      "loss": 1.4962,
+      "step": 5031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1987664699554443,
+      "learning_rate": 1.5402046783625733e-05,
+      "loss": 1.6789,
+      "step": 5032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.33236026763916,
+      "learning_rate": 1.5401133040935673e-05,
+      "loss": 1.4074,
+      "step": 5033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.021718978881836,
+      "learning_rate": 1.5400219298245614e-05,
+      "loss": 1.5597,
+      "step": 5034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9977402687072754,
+      "learning_rate": 1.5399305555555558e-05,
+      "loss": 1.7246,
+      "step": 5035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0386478900909424,
+      "learning_rate": 1.53983918128655e-05,
+      "loss": 1.6023,
+      "step": 5036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.190462827682495,
+      "learning_rate": 1.539747807017544e-05,
+      "loss": 1.8248,
+      "step": 5037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1262781620025635,
+      "learning_rate": 1.5396564327485383e-05,
+      "loss": 1.564,
+      "step": 5038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4598679542541504,
+      "learning_rate": 1.5395650584795323e-05,
+      "loss": 1.8341,
+      "step": 5039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.907538890838623,
+      "learning_rate": 1.5394736842105264e-05,
+      "loss": 1.2841,
+      "step": 5040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0947327613830566,
+      "learning_rate": 1.5393823099415204e-05,
+      "loss": 1.5546,
+      "step": 5041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2583820819854736,
+      "learning_rate": 1.539290935672515e-05,
+      "loss": 1.6499,
+      "step": 5042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2392172813415527,
+      "learning_rate": 1.539199561403509e-05,
+      "loss": 1.1219,
+      "step": 5043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.601600170135498,
+      "learning_rate": 1.539108187134503e-05,
+      "loss": 1.6591,
+      "step": 5044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1846365928649902,
+      "learning_rate": 1.5390168128654973e-05,
+      "loss": 1.9633,
+      "step": 5045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1527042388916016,
+      "learning_rate": 1.5389254385964914e-05,
+      "loss": 1.7787,
+      "step": 5046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.488776683807373,
+      "learning_rate": 1.5388340643274854e-05,
+      "loss": 1.4295,
+      "step": 5047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.56667160987854,
+      "learning_rate": 1.5387426900584795e-05,
+      "loss": 1.4897,
+      "step": 5048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.743242025375366,
+      "learning_rate": 1.538651315789474e-05,
+      "loss": 1.7697,
+      "step": 5049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.737020254135132,
+      "learning_rate": 1.538559941520468e-05,
+      "loss": 1.5915,
+      "step": 5050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.285635471343994,
+      "learning_rate": 1.538468567251462e-05,
+      "loss": 2.0293,
+      "step": 5051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.320972204208374,
+      "learning_rate": 1.5383771929824564e-05,
+      "loss": 2.1873,
+      "step": 5052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.007999420166016,
+      "learning_rate": 1.5382858187134504e-05,
+      "loss": 1.843,
+      "step": 5053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.605351209640503,
+      "learning_rate": 1.5381944444444445e-05,
+      "loss": 1.4525,
+      "step": 5054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.355543851852417,
+      "learning_rate": 1.5381030701754385e-05,
+      "loss": 1.3093,
+      "step": 5055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3030693531036377,
+      "learning_rate": 1.538011695906433e-05,
+      "loss": 1.2471,
+      "step": 5056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.451080799102783,
+      "learning_rate": 1.537920321637427e-05,
+      "loss": 1.6207,
+      "step": 5057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.087599992752075,
+      "learning_rate": 1.537828947368421e-05,
+      "loss": 1.735,
+      "step": 5058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.66434645652771,
+      "learning_rate": 1.5377375730994154e-05,
+      "loss": 1.3827,
+      "step": 5059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.200408935546875,
+      "learning_rate": 1.5376461988304095e-05,
+      "loss": 1.7608,
+      "step": 5060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.094129800796509,
+      "learning_rate": 1.5375548245614035e-05,
+      "loss": 1.116,
+      "step": 5061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4325151443481445,
+      "learning_rate": 1.5374634502923976e-05,
+      "loss": 1.6216,
+      "step": 5062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.770486831665039,
+      "learning_rate": 1.537372076023392e-05,
+      "loss": 1.3596,
+      "step": 5063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.819535732269287,
+      "learning_rate": 1.537280701754386e-05,
+      "loss": 1.7296,
+      "step": 5064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.00805401802063,
+      "learning_rate": 1.53718932748538e-05,
+      "loss": 1.5163,
+      "step": 5065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.674365758895874,
+      "learning_rate": 1.5370979532163745e-05,
+      "loss": 1.4388,
+      "step": 5066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.197394371032715,
+      "learning_rate": 1.5370065789473685e-05,
+      "loss": 1.6961,
+      "step": 5067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7054603099823,
+      "learning_rate": 1.5369152046783626e-05,
+      "loss": 1.5573,
+      "step": 5068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2922332286834717,
+      "learning_rate": 1.5368238304093566e-05,
+      "loss": 2.0494,
+      "step": 5069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.189697504043579,
+      "learning_rate": 1.536732456140351e-05,
+      "loss": 1.667,
+      "step": 5070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.399765968322754,
+      "learning_rate": 1.536641081871345e-05,
+      "loss": 1.3983,
+      "step": 5071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.523613452911377,
+      "learning_rate": 1.536549707602339e-05,
+      "loss": 1.5702,
+      "step": 5072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.673250675201416,
+      "learning_rate": 1.5364583333333335e-05,
+      "loss": 1.4462,
+      "step": 5073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.414132595062256,
+      "learning_rate": 1.5363669590643276e-05,
+      "loss": 1.9048,
+      "step": 5074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5999362468719482,
+      "learning_rate": 1.5362755847953216e-05,
+      "loss": 1.5743,
+      "step": 5075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3844661712646484,
+      "learning_rate": 1.5361842105263157e-05,
+      "loss": 1.4957,
+      "step": 5076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.824023723602295,
+      "learning_rate": 1.53609283625731e-05,
+      "loss": 1.9232,
+      "step": 5077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.802936553955078,
+      "learning_rate": 1.536001461988304e-05,
+      "loss": 1.5733,
+      "step": 5078
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9365129470825195,
+      "learning_rate": 1.5359100877192982e-05,
+      "loss": 1.7749,
+      "step": 5079
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6929311752319336,
+      "learning_rate": 1.5358187134502926e-05,
+      "loss": 1.7576,
+      "step": 5080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8094663619995117,
+      "learning_rate": 1.5357273391812866e-05,
+      "loss": 1.4388,
+      "step": 5081
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0485761165618896,
+      "learning_rate": 1.5356359649122807e-05,
+      "loss": 1.7314,
+      "step": 5082
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.506324529647827,
+      "learning_rate": 1.5355445906432747e-05,
+      "loss": 1.379,
+      "step": 5083
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5313587188720703,
+      "learning_rate": 1.535453216374269e-05,
+      "loss": 1.5219,
+      "step": 5084
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.827620267868042,
+      "learning_rate": 1.5353618421052632e-05,
+      "loss": 1.7695,
+      "step": 5085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.673233985900879,
+      "learning_rate": 1.5352704678362572e-05,
+      "loss": 1.7493,
+      "step": 5086
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5752718448638916,
+      "learning_rate": 1.5351790935672516e-05,
+      "loss": 1.3263,
+      "step": 5087
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7428340911865234,
+      "learning_rate": 1.535087719298246e-05,
+      "loss": 1.6439,
+      "step": 5088
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5426366329193115,
+      "learning_rate": 1.53499634502924e-05,
+      "loss": 1.3854,
+      "step": 5089
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1897130012512207,
+      "learning_rate": 1.534904970760234e-05,
+      "loss": 1.4448,
+      "step": 5090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4452781677246094,
+      "learning_rate": 1.534813596491228e-05,
+      "loss": 1.5454,
+      "step": 5091
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.183868885040283,
+      "learning_rate": 1.5347222222222226e-05,
+      "loss": 2.3553,
+      "step": 5092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8556067943573,
+      "learning_rate": 1.5346308479532166e-05,
+      "loss": 1.6704,
+      "step": 5093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7929630279541016,
+      "learning_rate": 1.5345394736842107e-05,
+      "loss": 1.5771,
+      "step": 5094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.832334518432617,
+      "learning_rate": 1.534448099415205e-05,
+      "loss": 1.5439,
+      "step": 5095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8950741291046143,
+      "learning_rate": 1.534356725146199e-05,
+      "loss": 1.8315,
+      "step": 5096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.530042886734009,
+      "learning_rate": 1.534265350877193e-05,
+      "loss": 1.7244,
+      "step": 5097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.803255558013916,
+      "learning_rate": 1.5341739766081872e-05,
+      "loss": 1.7829,
+      "step": 5098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0445098876953125,
+      "learning_rate": 1.5340826023391816e-05,
+      "loss": 1.224,
+      "step": 5099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.683999538421631,
+      "learning_rate": 1.5339912280701757e-05,
+      "loss": 1.7211,
+      "step": 5100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8864049911499023,
+      "learning_rate": 1.5338998538011697e-05,
+      "loss": 1.2097,
+      "step": 5101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6620230674743652,
+      "learning_rate": 1.533808479532164e-05,
+      "loss": 1.7178,
+      "step": 5102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0300004482269287,
+      "learning_rate": 1.533717105263158e-05,
+      "loss": 1.9832,
+      "step": 5103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1618645191192627,
+      "learning_rate": 1.5336257309941522e-05,
+      "loss": 1.9468,
+      "step": 5104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4745001792907715,
+      "learning_rate": 1.5335343567251463e-05,
+      "loss": 1.9629,
+      "step": 5105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.611889123916626,
+      "learning_rate": 1.5334429824561407e-05,
+      "loss": 1.573,
+      "step": 5106
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4063868522644043,
+      "learning_rate": 1.5333516081871347e-05,
+      "loss": 1.2857,
+      "step": 5107
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3694376945495605,
+      "learning_rate": 1.5332602339181288e-05,
+      "loss": 1.521,
+      "step": 5108
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8814008235931396,
+      "learning_rate": 1.533168859649123e-05,
+      "loss": 1.8309,
+      "step": 5109
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3768060207366943,
+      "learning_rate": 1.5330774853801172e-05,
+      "loss": 1.3136,
+      "step": 5110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.990063190460205,
+      "learning_rate": 1.5329861111111113e-05,
+      "loss": 1.7003,
+      "step": 5111
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.904848098754883,
+      "learning_rate": 1.5328947368421053e-05,
+      "loss": 1.4856,
+      "step": 5112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.843991279602051,
+      "learning_rate": 1.5328033625730997e-05,
+      "loss": 1.7477,
+      "step": 5113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9038851261138916,
+      "learning_rate": 1.5327119883040938e-05,
+      "loss": 1.7461,
+      "step": 5114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8249030113220215,
+      "learning_rate": 1.5326206140350878e-05,
+      "loss": 1.7342,
+      "step": 5115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.85182785987854,
+      "learning_rate": 1.5325292397660822e-05,
+      "loss": 1.9016,
+      "step": 5116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.307203769683838,
+      "learning_rate": 1.5324378654970762e-05,
+      "loss": 1.9799,
+      "step": 5117
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3922131061553955,
+      "learning_rate": 1.5323464912280703e-05,
+      "loss": 1.4249,
+      "step": 5118
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6613972187042236,
+      "learning_rate": 1.5322551169590644e-05,
+      "loss": 1.7137,
+      "step": 5119
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7195589542388916,
+      "learning_rate": 1.5321637426900587e-05,
+      "loss": 1.7003,
+      "step": 5120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.86264967918396,
+      "learning_rate": 1.5320723684210528e-05,
+      "loss": 1.7388,
+      "step": 5121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8895201683044434,
+      "learning_rate": 1.531980994152047e-05,
+      "loss": 1.4252,
+      "step": 5122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0583646297454834,
+      "learning_rate": 1.5318896198830412e-05,
+      "loss": 1.5473,
+      "step": 5123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.68776273727417,
+      "learning_rate": 1.5317982456140353e-05,
+      "loss": 1.6283,
+      "step": 5124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4445273876190186,
+      "learning_rate": 1.5317068713450293e-05,
+      "loss": 2.1711,
+      "step": 5125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.869662046432495,
+      "learning_rate": 1.5316154970760234e-05,
+      "loss": 1.6091,
+      "step": 5126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.023502349853516,
+      "learning_rate": 1.5315241228070178e-05,
+      "loss": 1.9893,
+      "step": 5127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.126680374145508,
+      "learning_rate": 1.531432748538012e-05,
+      "loss": 1.4053,
+      "step": 5128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.782155990600586,
+      "learning_rate": 1.531341374269006e-05,
+      "loss": 1.3466,
+      "step": 5129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9038116931915283,
+      "learning_rate": 1.5312500000000003e-05,
+      "loss": 1.2619,
+      "step": 5130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7416415214538574,
+      "learning_rate": 1.5311586257309943e-05,
+      "loss": 1.386,
+      "step": 5131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.632460117340088,
+      "learning_rate": 1.5310672514619884e-05,
+      "loss": 1.5815,
+      "step": 5132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0288538932800293,
+      "learning_rate": 1.5309758771929824e-05,
+      "loss": 1.373,
+      "step": 5133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.620884418487549,
+      "learning_rate": 1.530884502923977e-05,
+      "loss": 1.5438,
+      "step": 5134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.051438093185425,
+      "learning_rate": 1.530793128654971e-05,
+      "loss": 1.9753,
+      "step": 5135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.347494602203369,
+      "learning_rate": 1.530701754385965e-05,
+      "loss": 1.5879,
+      "step": 5136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8229174613952637,
+      "learning_rate": 1.5306103801169593e-05,
+      "loss": 1.4881,
+      "step": 5137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.336505889892578,
+      "learning_rate": 1.5305190058479534e-05,
+      "loss": 1.169,
+      "step": 5138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.887561798095703,
+      "learning_rate": 1.5304276315789474e-05,
+      "loss": 1.4203,
+      "step": 5139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.877474546432495,
+      "learning_rate": 1.5303362573099415e-05,
+      "loss": 1.5376,
+      "step": 5140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.804938793182373,
+      "learning_rate": 1.530244883040936e-05,
+      "loss": 1.8798,
+      "step": 5141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0168099403381348,
+      "learning_rate": 1.53015350877193e-05,
+      "loss": 1.673,
+      "step": 5142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0941004753112793,
+      "learning_rate": 1.530062134502924e-05,
+      "loss": 1.7151,
+      "step": 5143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7844817638397217,
+      "learning_rate": 1.5299707602339184e-05,
+      "loss": 1.7871,
+      "step": 5144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.186384677886963,
+      "learning_rate": 1.5298793859649124e-05,
+      "loss": 2.0634,
+      "step": 5145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.750068426132202,
+      "learning_rate": 1.5297880116959065e-05,
+      "loss": 1.5829,
+      "step": 5146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.220850944519043,
+      "learning_rate": 1.5296966374269005e-05,
+      "loss": 2.0378,
+      "step": 5147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.919755697250366,
+      "learning_rate": 1.529605263157895e-05,
+      "loss": 1.5098,
+      "step": 5148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.703308343887329,
+      "learning_rate": 1.529513888888889e-05,
+      "loss": 1.6692,
+      "step": 5149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1587302684783936,
+      "learning_rate": 1.529422514619883e-05,
+      "loss": 1.9232,
+      "step": 5150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.444929838180542,
+      "learning_rate": 1.5293311403508774e-05,
+      "loss": 1.7708,
+      "step": 5151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5738298892974854,
+      "learning_rate": 1.5292397660818715e-05,
+      "loss": 2.4767,
+      "step": 5152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.022622585296631,
+      "learning_rate": 1.5291483918128655e-05,
+      "loss": 2.0355,
+      "step": 5153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8236045837402344,
+      "learning_rate": 1.5290570175438596e-05,
+      "loss": 2.2349,
+      "step": 5154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.368727445602417,
+      "learning_rate": 1.528965643274854e-05,
+      "loss": 1.8839,
+      "step": 5155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.581260919570923,
+      "learning_rate": 1.528874269005848e-05,
+      "loss": 1.5544,
+      "step": 5156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9205269813537598,
+      "learning_rate": 1.528782894736842e-05,
+      "loss": 1.6797,
+      "step": 5157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.395899772644043,
+      "learning_rate": 1.5286915204678365e-05,
+      "loss": 1.2727,
+      "step": 5158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4911723136901855,
+      "learning_rate": 1.5286001461988305e-05,
+      "loss": 1.6341,
+      "step": 5159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.513371706008911,
+      "learning_rate": 1.5285087719298246e-05,
+      "loss": 1.3255,
+      "step": 5160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9203455448150635,
+      "learning_rate": 1.5284173976608186e-05,
+      "loss": 1.8696,
+      "step": 5161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.097069025039673,
+      "learning_rate": 1.528326023391813e-05,
+      "loss": 1.7979,
+      "step": 5162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0071628093719482,
+      "learning_rate": 1.528234649122807e-05,
+      "loss": 2.0617,
+      "step": 5163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.628363609313965,
+      "learning_rate": 1.528143274853801e-05,
+      "loss": 1.7464,
+      "step": 5164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.389277696609497,
+      "learning_rate": 1.5280519005847955e-05,
+      "loss": 2.1823,
+      "step": 5165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.080974578857422,
+      "learning_rate": 1.5279605263157896e-05,
+      "loss": 1.6084,
+      "step": 5166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1838760375976562,
+      "learning_rate": 1.5278691520467836e-05,
+      "loss": 1.5158,
+      "step": 5167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6009957790374756,
+      "learning_rate": 1.5277777777777777e-05,
+      "loss": 1.3984,
+      "step": 5168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8800036907196045,
+      "learning_rate": 1.527686403508772e-05,
+      "loss": 1.5335,
+      "step": 5169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0202455520629883,
+      "learning_rate": 1.527595029239766e-05,
+      "loss": 1.7404,
+      "step": 5170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.099362373352051,
+      "learning_rate": 1.5275036549707602e-05,
+      "loss": 2.0029,
+      "step": 5171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2616965770721436,
+      "learning_rate": 1.5274122807017546e-05,
+      "loss": 2.0579,
+      "step": 5172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7655832767486572,
+      "learning_rate": 1.5273209064327486e-05,
+      "loss": 1.6148,
+      "step": 5173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5112106800079346,
+      "learning_rate": 1.527229532163743e-05,
+      "loss": 1.7201,
+      "step": 5174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.432102918624878,
+      "learning_rate": 1.527138157894737e-05,
+      "loss": 1.4971,
+      "step": 5175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.764246940612793,
+      "learning_rate": 1.527046783625731e-05,
+      "loss": 1.3728,
+      "step": 5176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.998420000076294,
+      "learning_rate": 1.5269554093567255e-05,
+      "loss": 1.8116,
+      "step": 5177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.729032039642334,
+      "learning_rate": 1.5268640350877196e-05,
+      "loss": 1.4049,
+      "step": 5178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.248476982116699,
+      "learning_rate": 1.5267726608187136e-05,
+      "loss": 1.2058,
+      "step": 5179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3777408599853516,
+      "learning_rate": 1.5266812865497077e-05,
+      "loss": 1.2264,
+      "step": 5180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7080600261688232,
+      "learning_rate": 1.526589912280702e-05,
+      "loss": 1.353,
+      "step": 5181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.081751823425293,
+      "learning_rate": 1.526498538011696e-05,
+      "loss": 1.5948,
+      "step": 5182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2363100051879883,
+      "learning_rate": 1.52640716374269e-05,
+      "loss": 1.7468,
+      "step": 5183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4419853687286377,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 1.8299,
+      "step": 5184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5393307209014893,
+      "learning_rate": 1.5262244152046786e-05,
+      "loss": 1.327,
+      "step": 5185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1822280883789062,
+      "learning_rate": 1.5261330409356727e-05,
+      "loss": 1.5265,
+      "step": 5186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5299887657165527,
+      "learning_rate": 1.5260416666666667e-05,
+      "loss": 1.6511,
+      "step": 5187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0323076248168945,
+      "learning_rate": 1.525950292397661e-05,
+      "loss": 1.5186,
+      "step": 5188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.931586742401123,
+      "learning_rate": 1.5258589181286552e-05,
+      "loss": 1.6124,
+      "step": 5189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.066725015640259,
+      "learning_rate": 1.5257675438596494e-05,
+      "loss": 2.1591,
+      "step": 5190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3746583461761475,
+      "learning_rate": 1.5256761695906434e-05,
+      "loss": 1.6762,
+      "step": 5191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6801605224609375,
+      "learning_rate": 1.5255847953216377e-05,
+      "loss": 1.391,
+      "step": 5192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.238910675048828,
+      "learning_rate": 1.5254934210526317e-05,
+      "loss": 1.2696,
+      "step": 5193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.131720781326294,
+      "learning_rate": 1.525402046783626e-05,
+      "loss": 2.0235,
+      "step": 5194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1492903232574463,
+      "learning_rate": 1.52531067251462e-05,
+      "loss": 1.853,
+      "step": 5195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1968109607696533,
+      "learning_rate": 1.5252192982456142e-05,
+      "loss": 1.6489,
+      "step": 5196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6223273277282715,
+      "learning_rate": 1.5251279239766084e-05,
+      "loss": 1.6152,
+      "step": 5197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.846436023712158,
+      "learning_rate": 1.5250365497076025e-05,
+      "loss": 1.7309,
+      "step": 5198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1155290603637695,
+      "learning_rate": 1.5249451754385967e-05,
+      "loss": 1.6484,
+      "step": 5199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3424031734466553,
+      "learning_rate": 1.5248538011695908e-05,
+      "loss": 1.331,
+      "step": 5200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4714884757995605,
+      "learning_rate": 1.524762426900585e-05,
+      "loss": 1.3999,
+      "step": 5201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.147588014602661,
+      "learning_rate": 1.524671052631579e-05,
+      "loss": 1.8843,
+      "step": 5202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5500447750091553,
+      "learning_rate": 1.5245796783625733e-05,
+      "loss": 1.9282,
+      "step": 5203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8618125915527344,
+      "learning_rate": 1.5244883040935675e-05,
+      "loss": 1.5471,
+      "step": 5204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.653074026107788,
+      "learning_rate": 1.5243969298245615e-05,
+      "loss": 1.4226,
+      "step": 5205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4681828022003174,
+      "learning_rate": 1.5243055555555558e-05,
+      "loss": 1.8271,
+      "step": 5206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.571014642715454,
+      "learning_rate": 1.5242141812865498e-05,
+      "loss": 1.4835,
+      "step": 5207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7512848377227783,
+      "learning_rate": 1.524122807017544e-05,
+      "loss": 1.6084,
+      "step": 5208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.271217107772827,
+      "learning_rate": 1.524031432748538e-05,
+      "loss": 2.0998,
+      "step": 5209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.282477855682373,
+      "learning_rate": 1.5239400584795323e-05,
+      "loss": 1.223,
+      "step": 5210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.765202283859253,
+      "learning_rate": 1.5238486842105265e-05,
+      "loss": 1.4423,
+      "step": 5211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.464543342590332,
+      "learning_rate": 1.5237573099415206e-05,
+      "loss": 1.5176,
+      "step": 5212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.192629814147949,
+      "learning_rate": 1.5236659356725148e-05,
+      "loss": 1.2382,
+      "step": 5213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0126936435699463,
+      "learning_rate": 1.5235745614035089e-05,
+      "loss": 1.507,
+      "step": 5214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8031580448150635,
+      "learning_rate": 1.523483187134503e-05,
+      "loss": 1.6366,
+      "step": 5215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.802227735519409,
+      "learning_rate": 1.5233918128654971e-05,
+      "loss": 1.6455,
+      "step": 5216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6851961612701416,
+      "learning_rate": 1.5233004385964913e-05,
+      "loss": 1.6248,
+      "step": 5217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.630953073501587,
+      "learning_rate": 1.5232090643274856e-05,
+      "loss": 1.326,
+      "step": 5218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.098773956298828,
+      "learning_rate": 1.5231176900584796e-05,
+      "loss": 1.8283,
+      "step": 5219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6593260765075684,
+      "learning_rate": 1.5230263157894738e-05,
+      "loss": 1.3494,
+      "step": 5220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.385695219039917,
+      "learning_rate": 1.5229349415204679e-05,
+      "loss": 1.5325,
+      "step": 5221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9472169876098633,
+      "learning_rate": 1.5228435672514621e-05,
+      "loss": 1.4873,
+      "step": 5222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9596054553985596,
+      "learning_rate": 1.5227521929824562e-05,
+      "loss": 2.0079,
+      "step": 5223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8702526092529297,
+      "learning_rate": 1.5226608187134504e-05,
+      "loss": 1.5988,
+      "step": 5224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5931546688079834,
+      "learning_rate": 1.5225694444444446e-05,
+      "loss": 1.3609,
+      "step": 5225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.26560378074646,
+      "learning_rate": 1.5224780701754387e-05,
+      "loss": 1.354,
+      "step": 5226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.886354684829712,
+      "learning_rate": 1.5223866959064329e-05,
+      "loss": 1.643,
+      "step": 5227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.787118673324585,
+      "learning_rate": 1.522295321637427e-05,
+      "loss": 1.8524,
+      "step": 5228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5496580600738525,
+      "learning_rate": 1.5222039473684212e-05,
+      "loss": 1.6046,
+      "step": 5229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8164968490600586,
+      "learning_rate": 1.5221125730994152e-05,
+      "loss": 1.6125,
+      "step": 5230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4206199645996094,
+      "learning_rate": 1.5220211988304094e-05,
+      "loss": 1.8686,
+      "step": 5231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.618222951889038,
+      "learning_rate": 1.5219298245614037e-05,
+      "loss": 1.6897,
+      "step": 5232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.969547986984253,
+      "learning_rate": 1.5218384502923977e-05,
+      "loss": 1.775,
+      "step": 5233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2896487712860107,
+      "learning_rate": 1.521747076023392e-05,
+      "loss": 1.6831,
+      "step": 5234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8106813430786133,
+      "learning_rate": 1.521655701754386e-05,
+      "loss": 1.6467,
+      "step": 5235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8219106197357178,
+      "learning_rate": 1.5215643274853802e-05,
+      "loss": 1.4182,
+      "step": 5236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5506985187530518,
+      "learning_rate": 1.5214729532163743e-05,
+      "loss": 1.3092,
+      "step": 5237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8729827404022217,
+      "learning_rate": 1.5213815789473685e-05,
+      "loss": 1.5197,
+      "step": 5238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5159032344818115,
+      "learning_rate": 1.5212902046783627e-05,
+      "loss": 1.9315,
+      "step": 5239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.588418960571289,
+      "learning_rate": 1.5211988304093568e-05,
+      "loss": 1.5205,
+      "step": 5240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.512845993041992,
+      "learning_rate": 1.521107456140351e-05,
+      "loss": 1.5884,
+      "step": 5241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.381608724594116,
+      "learning_rate": 1.521016081871345e-05,
+      "loss": 2.1306,
+      "step": 5242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.584364414215088,
+      "learning_rate": 1.5209247076023393e-05,
+      "loss": 1.3822,
+      "step": 5243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.884676456451416,
+      "learning_rate": 1.5208333333333333e-05,
+      "loss": 1.7938,
+      "step": 5244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2425591945648193,
+      "learning_rate": 1.5207419590643275e-05,
+      "loss": 1.788,
+      "step": 5245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5464799404144287,
+      "learning_rate": 1.5206505847953218e-05,
+      "loss": 1.1916,
+      "step": 5246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7017338275909424,
+      "learning_rate": 1.5205592105263158e-05,
+      "loss": 1.6867,
+      "step": 5247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7917158603668213,
+      "learning_rate": 1.52046783625731e-05,
+      "loss": 1.5656,
+      "step": 5248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.030308246612549,
+      "learning_rate": 1.5203764619883041e-05,
+      "loss": 1.8385,
+      "step": 5249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.34956431388855,
+      "learning_rate": 1.5202850877192983e-05,
+      "loss": 2.0228,
+      "step": 5250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.614468812942505,
+      "learning_rate": 1.5201937134502924e-05,
+      "loss": 1.9718,
+      "step": 5251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1754367351531982,
+      "learning_rate": 1.5201023391812866e-05,
+      "loss": 1.6452,
+      "step": 5252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4806981086730957,
+      "learning_rate": 1.5200109649122808e-05,
+      "loss": 1.5159,
+      "step": 5253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.769440174102783,
+      "learning_rate": 1.5199195906432749e-05,
+      "loss": 1.8919,
+      "step": 5254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.924715518951416,
+      "learning_rate": 1.519828216374269e-05,
+      "loss": 1.601,
+      "step": 5255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.097238063812256,
+      "learning_rate": 1.5197368421052631e-05,
+      "loss": 1.8327,
+      "step": 5256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7830021381378174,
+      "learning_rate": 1.5196454678362574e-05,
+      "loss": 1.6441,
+      "step": 5257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.739079475402832,
+      "learning_rate": 1.5195540935672514e-05,
+      "loss": 1.3871,
+      "step": 5258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.811974287033081,
+      "learning_rate": 1.5194627192982458e-05,
+      "loss": 1.5288,
+      "step": 5259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.606001138687134,
+      "learning_rate": 1.51937134502924e-05,
+      "loss": 1.4302,
+      "step": 5260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.90110182762146,
+      "learning_rate": 1.519279970760234e-05,
+      "loss": 1.7224,
+      "step": 5261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3784334659576416,
+      "learning_rate": 1.5191885964912283e-05,
+      "loss": 1.4092,
+      "step": 5262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6230900287628174,
+      "learning_rate": 1.5190972222222225e-05,
+      "loss": 1.3965,
+      "step": 5263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3771140575408936,
+      "learning_rate": 1.5190058479532166e-05,
+      "loss": 1.7663,
+      "step": 5264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.553687334060669,
+      "learning_rate": 1.5189144736842108e-05,
+      "loss": 1.4147,
+      "step": 5265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9733879566192627,
+      "learning_rate": 1.5188230994152048e-05,
+      "loss": 1.6128,
+      "step": 5266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7094595432281494,
+      "learning_rate": 1.518731725146199e-05,
+      "loss": 1.4981,
+      "step": 5267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2698142528533936,
+      "learning_rate": 1.5186403508771931e-05,
+      "loss": 1.8227,
+      "step": 5268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4257726669311523,
+      "learning_rate": 1.5185489766081873e-05,
+      "loss": 2.0214,
+      "step": 5269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8808467388153076,
+      "learning_rate": 1.5184576023391816e-05,
+      "loss": 1.4632,
+      "step": 5270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9352922439575195,
+      "learning_rate": 1.5183662280701756e-05,
+      "loss": 1.5715,
+      "step": 5271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.156999349594116,
+      "learning_rate": 1.5182748538011698e-05,
+      "loss": 1.2861,
+      "step": 5272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6867315769195557,
+      "learning_rate": 1.5181834795321639e-05,
+      "loss": 1.7165,
+      "step": 5273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0766355991363525,
+      "learning_rate": 1.5180921052631581e-05,
+      "loss": 1.6245,
+      "step": 5274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.948227882385254,
+      "learning_rate": 1.5180007309941522e-05,
+      "loss": 1.7121,
+      "step": 5275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5939934253692627,
+      "learning_rate": 1.5179093567251464e-05,
+      "loss": 1.9497,
+      "step": 5276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9335176944732666,
+      "learning_rate": 1.5178179824561406e-05,
+      "loss": 1.7903,
+      "step": 5277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.991701364517212,
+      "learning_rate": 1.5177266081871347e-05,
+      "loss": 2.173,
+      "step": 5278
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0586068630218506,
+      "learning_rate": 1.5176352339181289e-05,
+      "loss": 2.0294,
+      "step": 5279
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.917569160461426,
+      "learning_rate": 1.517543859649123e-05,
+      "loss": 1.9783,
+      "step": 5280
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.308912754058838,
+      "learning_rate": 1.5174524853801172e-05,
+      "loss": 1.3615,
+      "step": 5281
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5405433177948,
+      "learning_rate": 1.5173611111111112e-05,
+      "loss": 1.5876,
+      "step": 5282
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3317089080810547,
+      "learning_rate": 1.5172697368421054e-05,
+      "loss": 1.7731,
+      "step": 5283
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.847451686859131,
+      "learning_rate": 1.5171783625730997e-05,
+      "loss": 1.4702,
+      "step": 5284
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9335782527923584,
+      "learning_rate": 1.5170869883040937e-05,
+      "loss": 1.4646,
+      "step": 5285
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6447551250457764,
+      "learning_rate": 1.516995614035088e-05,
+      "loss": 1.4665,
+      "step": 5286
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.918166160583496,
+      "learning_rate": 1.516904239766082e-05,
+      "loss": 1.6507,
+      "step": 5287
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0296220779418945,
+      "learning_rate": 1.5168128654970762e-05,
+      "loss": 1.8807,
+      "step": 5288
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.065075397491455,
+      "learning_rate": 1.5167214912280703e-05,
+      "loss": 1.8047,
+      "step": 5289
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3230960369110107,
+      "learning_rate": 1.5166301169590645e-05,
+      "loss": 1.5377,
+      "step": 5290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7462310791015625,
+      "learning_rate": 1.5165387426900587e-05,
+      "loss": 1.6086,
+      "step": 5291
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4814889430999756,
+      "learning_rate": 1.5164473684210528e-05,
+      "loss": 1.6442,
+      "step": 5292
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.423729658126831,
+      "learning_rate": 1.516355994152047e-05,
+      "loss": 1.9695,
+      "step": 5293
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0964770317077637,
+      "learning_rate": 1.516264619883041e-05,
+      "loss": 1.5205,
+      "step": 5294
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7944819927215576,
+      "learning_rate": 1.5161732456140353e-05,
+      "loss": 1.5082,
+      "step": 5295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.182724952697754,
+      "learning_rate": 1.5160818713450293e-05,
+      "loss": 2.1249,
+      "step": 5296
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5415639877319336,
+      "learning_rate": 1.5159904970760235e-05,
+      "loss": 1.9025,
+      "step": 5297
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.223878860473633,
+      "learning_rate": 1.5158991228070178e-05,
+      "loss": 1.719,
+      "step": 5298
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.214048385620117,
+      "learning_rate": 1.5158077485380118e-05,
+      "loss": 1.7952,
+      "step": 5299
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1354598999023438,
+      "learning_rate": 1.515716374269006e-05,
+      "loss": 1.8494,
+      "step": 5300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.464430570602417,
+      "learning_rate": 1.515625e-05,
+      "loss": 1.4908,
+      "step": 5301
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6034204959869385,
+      "learning_rate": 1.5155336257309943e-05,
+      "loss": 1.2785,
+      "step": 5302
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4378738403320312,
+      "learning_rate": 1.5154422514619884e-05,
+      "loss": 1.3763,
+      "step": 5303
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7222042083740234,
+      "learning_rate": 1.5153508771929826e-05,
+      "loss": 2.3311,
+      "step": 5304
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9816861152648926,
+      "learning_rate": 1.5152595029239768e-05,
+      "loss": 1.8067,
+      "step": 5305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.656503438949585,
+      "learning_rate": 1.5151681286549709e-05,
+      "loss": 1.5172,
+      "step": 5306
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.514878988265991,
+      "learning_rate": 1.515076754385965e-05,
+      "loss": 1.5167,
+      "step": 5307
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.488708257675171,
+      "learning_rate": 1.5149853801169591e-05,
+      "loss": 1.5866,
+      "step": 5308
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.86000919342041,
+      "learning_rate": 1.5148940058479533e-05,
+      "loss": 1.5683,
+      "step": 5309
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9699416160583496,
+      "learning_rate": 1.5148026315789474e-05,
+      "loss": 1.5428,
+      "step": 5310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.008214235305786,
+      "learning_rate": 1.5147112573099416e-05,
+      "loss": 1.5144,
+      "step": 5311
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3026318550109863,
+      "learning_rate": 1.5146198830409358e-05,
+      "loss": 2.2671,
+      "step": 5312
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9729549884796143,
+      "learning_rate": 1.5145285087719299e-05,
+      "loss": 1.9198,
+      "step": 5313
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7940919399261475,
+      "learning_rate": 1.5144371345029241e-05,
+      "loss": 1.5192,
+      "step": 5314
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.647005558013916,
+      "learning_rate": 1.5143457602339182e-05,
+      "loss": 1.6649,
+      "step": 5315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.239703893661499,
+      "learning_rate": 1.5142543859649124e-05,
+      "loss": 1.7782,
+      "step": 5316
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0126454830169678,
+      "learning_rate": 1.5141630116959064e-05,
+      "loss": 1.7348,
+      "step": 5317
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.855358123779297,
+      "learning_rate": 1.5140716374269007e-05,
+      "loss": 1.8793,
+      "step": 5318
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.403350353240967,
+      "learning_rate": 1.5139802631578949e-05,
+      "loss": 1.4549,
+      "step": 5319
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.850177526473999,
+      "learning_rate": 1.513888888888889e-05,
+      "loss": 1.6411,
+      "step": 5320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.242619037628174,
+      "learning_rate": 1.5137975146198832e-05,
+      "loss": 1.5925,
+      "step": 5321
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5162949562072754,
+      "learning_rate": 1.5137061403508772e-05,
+      "loss": 1.3694,
+      "step": 5322
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9238719940185547,
+      "learning_rate": 1.5136147660818714e-05,
+      "loss": 1.3215,
+      "step": 5323
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0199568271636963,
+      "learning_rate": 1.5135233918128655e-05,
+      "loss": 1.6377,
+      "step": 5324
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.397838592529297,
+      "learning_rate": 1.5134320175438597e-05,
+      "loss": 1.8026,
+      "step": 5325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.320891857147217,
+      "learning_rate": 1.513340643274854e-05,
+      "loss": 1.4422,
+      "step": 5326
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7445333003997803,
+      "learning_rate": 1.513249269005848e-05,
+      "loss": 1.7038,
+      "step": 5327
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1384594440460205,
+      "learning_rate": 1.5131578947368422e-05,
+      "loss": 1.7781,
+      "step": 5328
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0406744480133057,
+      "learning_rate": 1.5130665204678363e-05,
+      "loss": 2.044,
+      "step": 5329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.924121618270874,
+      "learning_rate": 1.5129751461988305e-05,
+      "loss": 1.5856,
+      "step": 5330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.650864362716675,
+      "learning_rate": 1.5128837719298245e-05,
+      "loss": 1.7875,
+      "step": 5331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.796257495880127,
+      "learning_rate": 1.5127923976608188e-05,
+      "loss": 1.7162,
+      "step": 5332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9502530097961426,
+      "learning_rate": 1.512701023391813e-05,
+      "loss": 1.4838,
+      "step": 5333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3219053745269775,
+      "learning_rate": 1.512609649122807e-05,
+      "loss": 2.1238,
+      "step": 5334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2937583923339844,
+      "learning_rate": 1.5125182748538013e-05,
+      "loss": 1.0586,
+      "step": 5335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9363181591033936,
+      "learning_rate": 1.5124269005847953e-05,
+      "loss": 1.374,
+      "step": 5336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7620794773101807,
+      "learning_rate": 1.5123355263157895e-05,
+      "loss": 1.5444,
+      "step": 5337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.727985382080078,
+      "learning_rate": 1.5122441520467836e-05,
+      "loss": 2.1501,
+      "step": 5338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.711470603942871,
+      "learning_rate": 1.5121527777777778e-05,
+      "loss": 1.3454,
+      "step": 5339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1144087314605713,
+      "learning_rate": 1.512061403508772e-05,
+      "loss": 1.7064,
+      "step": 5340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.479031801223755,
+      "learning_rate": 1.5119700292397661e-05,
+      "loss": 1.4551,
+      "step": 5341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7791097164154053,
+      "learning_rate": 1.5118786549707603e-05,
+      "loss": 1.9304,
+      "step": 5342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5897045135498047,
+      "learning_rate": 1.5117872807017544e-05,
+      "loss": 1.6453,
+      "step": 5343
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0074028968811035,
+      "learning_rate": 1.5116959064327488e-05,
+      "loss": 1.6343,
+      "step": 5344
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6195075511932373,
+      "learning_rate": 1.511604532163743e-05,
+      "loss": 1.7829,
+      "step": 5345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4696130752563477,
+      "learning_rate": 1.511513157894737e-05,
+      "loss": 1.7875,
+      "step": 5346
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7744181156158447,
+      "learning_rate": 1.5114217836257312e-05,
+      "loss": 1.5261,
+      "step": 5347
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3687636852264404,
+      "learning_rate": 1.5113304093567253e-05,
+      "loss": 1.2697,
+      "step": 5348
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2443478107452393,
+      "learning_rate": 1.5112390350877195e-05,
+      "loss": 1.8015,
+      "step": 5349
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5526483058929443,
+      "learning_rate": 1.5111476608187137e-05,
+      "loss": 1.6009,
+      "step": 5350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7361345291137695,
+      "learning_rate": 1.5110562865497078e-05,
+      "loss": 1.5616,
+      "step": 5351
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.968264579772949,
+      "learning_rate": 1.510964912280702e-05,
+      "loss": 1.8834,
+      "step": 5352
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2740581035614014,
+      "learning_rate": 1.510873538011696e-05,
+      "loss": 1.1283,
+      "step": 5353
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5104477405548096,
+      "learning_rate": 1.5107821637426903e-05,
+      "loss": 1.3631,
+      "step": 5354
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1101362705230713,
+      "learning_rate": 1.5106907894736843e-05,
+      "loss": 1.5244,
+      "step": 5355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9627621173858643,
+      "learning_rate": 1.5105994152046786e-05,
+      "loss": 1.7193,
+      "step": 5356
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.913769483566284,
+      "learning_rate": 1.5105080409356728e-05,
+      "loss": 1.6024,
+      "step": 5357
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.944683074951172,
+      "learning_rate": 1.5104166666666668e-05,
+      "loss": 1.6843,
+      "step": 5358
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9030189514160156,
+      "learning_rate": 1.510325292397661e-05,
+      "loss": 2.011,
+      "step": 5359
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.815169334411621,
+      "learning_rate": 1.5102339181286551e-05,
+      "loss": 1.677,
+      "step": 5360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.56204891204834,
+      "learning_rate": 1.5101425438596493e-05,
+      "loss": 1.1207,
+      "step": 5361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6993796825408936,
+      "learning_rate": 1.5100511695906434e-05,
+      "loss": 1.5874,
+      "step": 5362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5775344371795654,
+      "learning_rate": 1.5099597953216376e-05,
+      "loss": 1.7878,
+      "step": 5363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9817309379577637,
+      "learning_rate": 1.5098684210526318e-05,
+      "loss": 1.8083,
+      "step": 5364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3597593307495117,
+      "learning_rate": 1.5097770467836259e-05,
+      "loss": 1.5467,
+      "step": 5365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.670233964920044,
+      "learning_rate": 1.5096856725146201e-05,
+      "loss": 1.4596,
+      "step": 5366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3324124813079834,
+      "learning_rate": 1.5095942982456142e-05,
+      "loss": 1.428,
+      "step": 5367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.493959426879883,
+      "learning_rate": 1.5095029239766084e-05,
+      "loss": 1.4276,
+      "step": 5368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8257298469543457,
+      "learning_rate": 1.5094115497076024e-05,
+      "loss": 1.6823,
+      "step": 5369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6310689449310303,
+      "learning_rate": 1.5093201754385967e-05,
+      "loss": 1.7009,
+      "step": 5370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.873237133026123,
+      "learning_rate": 1.5092288011695909e-05,
+      "loss": 1.6171,
+      "step": 5371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.600733518600464,
+      "learning_rate": 1.509137426900585e-05,
+      "loss": 1.4056,
+      "step": 5372
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.718358278274536,
+      "learning_rate": 1.5090460526315792e-05,
+      "loss": 1.3088,
+      "step": 5373
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1416800022125244,
+      "learning_rate": 1.5089546783625732e-05,
+      "loss": 1.668,
+      "step": 5374
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9101874828338623,
+      "learning_rate": 1.5088633040935674e-05,
+      "loss": 1.6497,
+      "step": 5375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2441320419311523,
+      "learning_rate": 1.5087719298245615e-05,
+      "loss": 1.8863,
+      "step": 5376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0083539485931396,
+      "learning_rate": 1.5086805555555557e-05,
+      "loss": 1.953,
+      "step": 5377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.787747859954834,
+      "learning_rate": 1.50858918128655e-05,
+      "loss": 1.5817,
+      "step": 5378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.726381540298462,
+      "learning_rate": 1.508497807017544e-05,
+      "loss": 1.6805,
+      "step": 5379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.52118182182312,
+      "learning_rate": 1.5084064327485382e-05,
+      "loss": 1.4505,
+      "step": 5380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.537226915359497,
+      "learning_rate": 1.5083150584795323e-05,
+      "loss": 1.4596,
+      "step": 5381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.778538703918457,
+      "learning_rate": 1.5082236842105265e-05,
+      "loss": 1.6574,
+      "step": 5382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7989518642425537,
+      "learning_rate": 1.5081323099415205e-05,
+      "loss": 1.4805,
+      "step": 5383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1168055534362793,
+      "learning_rate": 1.5080409356725148e-05,
+      "loss": 1.4042,
+      "step": 5384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.64422607421875,
+      "learning_rate": 1.507949561403509e-05,
+      "loss": 1.7478,
+      "step": 5385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6141958236694336,
+      "learning_rate": 1.507858187134503e-05,
+      "loss": 1.3245,
+      "step": 5386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.881850004196167,
+      "learning_rate": 1.5077668128654973e-05,
+      "loss": 1.9834,
+      "step": 5387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8354222774505615,
+      "learning_rate": 1.5076754385964913e-05,
+      "loss": 1.4253,
+      "step": 5388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6070196628570557,
+      "learning_rate": 1.5075840643274855e-05,
+      "loss": 1.3519,
+      "step": 5389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.269648551940918,
+      "learning_rate": 1.5074926900584796e-05,
+      "loss": 1.456,
+      "step": 5390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9346442222595215,
+      "learning_rate": 1.5074013157894738e-05,
+      "loss": 1.797,
+      "step": 5391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1879823207855225,
+      "learning_rate": 1.507309941520468e-05,
+      "loss": 2.0645,
+      "step": 5392
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6811771392822266,
+      "learning_rate": 1.507218567251462e-05,
+      "loss": 1.8109,
+      "step": 5393
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5171055793762207,
+      "learning_rate": 1.5071271929824563e-05,
+      "loss": 1.6403,
+      "step": 5394
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.904700517654419,
+      "learning_rate": 1.5070358187134504e-05,
+      "loss": 2.0184,
+      "step": 5395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8544116020202637,
+      "learning_rate": 1.5069444444444446e-05,
+      "loss": 1.5191,
+      "step": 5396
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.525392532348633,
+      "learning_rate": 1.5068530701754386e-05,
+      "loss": 1.6498,
+      "step": 5397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3645365238189697,
+      "learning_rate": 1.5067616959064329e-05,
+      "loss": 1.3679,
+      "step": 5398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2622275352478027,
+      "learning_rate": 1.506670321637427e-05,
+      "loss": 1.4891,
+      "step": 5399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.079468011856079,
+      "learning_rate": 1.5065789473684211e-05,
+      "loss": 1.609,
+      "step": 5400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8794472217559814,
+      "learning_rate": 1.5064875730994153e-05,
+      "loss": 1.6333,
+      "step": 5401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0312252044677734,
+      "learning_rate": 1.5063961988304094e-05,
+      "loss": 1.6842,
+      "step": 5402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6398940086364746,
+      "learning_rate": 1.5063048245614036e-05,
+      "loss": 1.4107,
+      "step": 5403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.877465009689331,
+      "learning_rate": 1.5062134502923977e-05,
+      "loss": 1.4935,
+      "step": 5404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0569276809692383,
+      "learning_rate": 1.5061220760233919e-05,
+      "loss": 1.7288,
+      "step": 5405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9427459239959717,
+      "learning_rate": 1.5060307017543861e-05,
+      "loss": 1.8437,
+      "step": 5406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4701523780822754,
+      "learning_rate": 1.5059393274853802e-05,
+      "loss": 1.8065,
+      "step": 5407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7035999298095703,
+      "learning_rate": 1.5058479532163744e-05,
+      "loss": 1.4458,
+      "step": 5408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.118039608001709,
+      "learning_rate": 1.5057565789473684e-05,
+      "loss": 1.747,
+      "step": 5409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7576324939727783,
+      "learning_rate": 1.5056652046783627e-05,
+      "loss": 1.8507,
+      "step": 5410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1882894039154053,
+      "learning_rate": 1.5055738304093567e-05,
+      "loss": 1.8477,
+      "step": 5411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.67954158782959,
+      "learning_rate": 1.505482456140351e-05,
+      "loss": 1.6226,
+      "step": 5412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.99478816986084,
+      "learning_rate": 1.5053910818713452e-05,
+      "loss": 1.4977,
+      "step": 5413
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1819839477539062,
+      "learning_rate": 1.5052997076023392e-05,
+      "loss": 1.832,
+      "step": 5414
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.657222032546997,
+      "learning_rate": 1.5052083333333334e-05,
+      "loss": 2.071,
+      "step": 5415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0127973556518555,
+      "learning_rate": 1.5051169590643275e-05,
+      "loss": 1.4421,
+      "step": 5416
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.065359354019165,
+      "learning_rate": 1.5050255847953217e-05,
+      "loss": 1.9371,
+      "step": 5417
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5998408794403076,
+      "learning_rate": 1.5049342105263158e-05,
+      "loss": 1.4809,
+      "step": 5418
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.374769687652588,
+      "learning_rate": 1.50484283625731e-05,
+      "loss": 1.3316,
+      "step": 5419
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.644575834274292,
+      "learning_rate": 1.5047514619883042e-05,
+      "loss": 1.8949,
+      "step": 5420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.85657000541687,
+      "learning_rate": 1.5046600877192983e-05,
+      "loss": 1.7903,
+      "step": 5421
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.61415958404541,
+      "learning_rate": 1.5045687134502925e-05,
+      "loss": 1.4869,
+      "step": 5422
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5943639278411865,
+      "learning_rate": 1.5044773391812865e-05,
+      "loss": 1.6114,
+      "step": 5423
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1320252418518066,
+      "learning_rate": 1.5043859649122808e-05,
+      "loss": 1.7063,
+      "step": 5424
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.090930938720703,
+      "learning_rate": 1.5042945906432748e-05,
+      "loss": 1.5353,
+      "step": 5425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2246830463409424,
+      "learning_rate": 1.504203216374269e-05,
+      "loss": 1.8118,
+      "step": 5426
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2321767807006836,
+      "learning_rate": 1.5041118421052633e-05,
+      "loss": 1.6203,
+      "step": 5427
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2091782093048096,
+      "learning_rate": 1.5040204678362573e-05,
+      "loss": 1.5745,
+      "step": 5428
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.184080123901367,
+      "learning_rate": 1.5039290935672515e-05,
+      "loss": 1.979,
+      "step": 5429
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2902989387512207,
+      "learning_rate": 1.503837719298246e-05,
+      "loss": 1.3378,
+      "step": 5430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2200701236724854,
+      "learning_rate": 1.50374634502924e-05,
+      "loss": 1.4582,
+      "step": 5431
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7021162509918213,
+      "learning_rate": 1.5036549707602342e-05,
+      "loss": 1.6986,
+      "step": 5432
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9548957347869873,
+      "learning_rate": 1.5035635964912283e-05,
+      "loss": 1.8723,
+      "step": 5433
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.538895130157471,
+      "learning_rate": 1.5034722222222225e-05,
+      "loss": 1.8748,
+      "step": 5434
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.939911365509033,
+      "learning_rate": 1.5033808479532165e-05,
+      "loss": 1.6411,
+      "step": 5435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.468402862548828,
+      "learning_rate": 1.5032894736842108e-05,
+      "loss": 1.4187,
+      "step": 5436
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9620296955108643,
+      "learning_rate": 1.503198099415205e-05,
+      "loss": 1.6016,
+      "step": 5437
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7310919761657715,
+      "learning_rate": 1.503106725146199e-05,
+      "loss": 1.5763,
+      "step": 5438
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.06659197807312,
+      "learning_rate": 1.5030153508771932e-05,
+      "loss": 1.5306,
+      "step": 5439
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.538907289505005,
+      "learning_rate": 1.5029239766081873e-05,
+      "loss": 1.6667,
+      "step": 5440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6860849857330322,
+      "learning_rate": 1.5028326023391815e-05,
+      "loss": 1.565,
+      "step": 5441
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1550259590148926,
+      "learning_rate": 1.5027412280701756e-05,
+      "loss": 1.5602,
+      "step": 5442
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.262770175933838,
+      "learning_rate": 1.5026498538011698e-05,
+      "loss": 1.5478,
+      "step": 5443
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.088977813720703,
+      "learning_rate": 1.502558479532164e-05,
+      "loss": 1.461,
+      "step": 5444
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6134426593780518,
+      "learning_rate": 1.502467105263158e-05,
+      "loss": 1.9706,
+      "step": 5445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6689388751983643,
+      "learning_rate": 1.5023757309941523e-05,
+      "loss": 1.5199,
+      "step": 5446
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.84354305267334,
+      "learning_rate": 1.5022843567251463e-05,
+      "loss": 1.656,
+      "step": 5447
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.595001697540283,
+      "learning_rate": 1.5021929824561406e-05,
+      "loss": 1.3481,
+      "step": 5448
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.870365619659424,
+      "learning_rate": 1.5021016081871346e-05,
+      "loss": 1.7516,
+      "step": 5449
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.67360258102417,
+      "learning_rate": 1.5020102339181288e-05,
+      "loss": 1.2919,
+      "step": 5450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.611398696899414,
+      "learning_rate": 1.501918859649123e-05,
+      "loss": 1.3046,
+      "step": 5451
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.09470796585083,
+      "learning_rate": 1.5018274853801171e-05,
+      "loss": 1.8971,
+      "step": 5452
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5623748302459717,
+      "learning_rate": 1.5017361111111113e-05,
+      "loss": 1.3915,
+      "step": 5453
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8750412464141846,
+      "learning_rate": 1.5016447368421054e-05,
+      "loss": 1.6748,
+      "step": 5454
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0203351974487305,
+      "learning_rate": 1.5015533625730996e-05,
+      "loss": 1.8289,
+      "step": 5455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.755481719970703,
+      "learning_rate": 1.5014619883040937e-05,
+      "loss": 1.5842,
+      "step": 5456
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8144285678863525,
+      "learning_rate": 1.5013706140350879e-05,
+      "loss": 1.6742,
+      "step": 5457
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.641339063644409,
+      "learning_rate": 1.5012792397660821e-05,
+      "loss": 1.3057,
+      "step": 5458
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5496153831481934,
+      "learning_rate": 1.5011878654970762e-05,
+      "loss": 1.57,
+      "step": 5459
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0114030838012695,
+      "learning_rate": 1.5010964912280704e-05,
+      "loss": 1.693,
+      "step": 5460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5039117336273193,
+      "learning_rate": 1.5010051169590644e-05,
+      "loss": 1.9586,
+      "step": 5461
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1851484775543213,
+      "learning_rate": 1.5009137426900587e-05,
+      "loss": 1.7062,
+      "step": 5462
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7166268825531006,
+      "learning_rate": 1.5008223684210527e-05,
+      "loss": 1.5753,
+      "step": 5463
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.888059616088867,
+      "learning_rate": 1.500730994152047e-05,
+      "loss": 1.765,
+      "step": 5464
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.20979380607605,
+      "learning_rate": 1.5006396198830412e-05,
+      "loss": 1.4195,
+      "step": 5465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7742481231689453,
+      "learning_rate": 1.5005482456140352e-05,
+      "loss": 1.6249,
+      "step": 5466
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.404872179031372,
+      "learning_rate": 1.5004568713450294e-05,
+      "loss": 1.554,
+      "step": 5467
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.393521785736084,
+      "learning_rate": 1.5003654970760235e-05,
+      "loss": 1.6011,
+      "step": 5468
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.987621307373047,
+      "learning_rate": 1.5002741228070177e-05,
+      "loss": 1.8116,
+      "step": 5469
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.22434401512146,
+      "learning_rate": 1.5001827485380118e-05,
+      "loss": 1.672,
+      "step": 5470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0003983974456787,
+      "learning_rate": 1.500091374269006e-05,
+      "loss": 1.6309,
+      "step": 5471
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.368828535079956,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.197,
+      "step": 5472
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.027845621109009,
+      "learning_rate": 1.4999086257309943e-05,
+      "loss": 1.7958,
+      "step": 5473
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.504915714263916,
+      "learning_rate": 1.4998172514619885e-05,
+      "loss": 1.9032,
+      "step": 5474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8876278400421143,
+      "learning_rate": 1.4997258771929825e-05,
+      "loss": 1.5995,
+      "step": 5475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7173264026641846,
+      "learning_rate": 1.4996345029239768e-05,
+      "loss": 1.6147,
+      "step": 5476
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.459378719329834,
+      "learning_rate": 1.4995431286549708e-05,
+      "loss": 1.6068,
+      "step": 5477
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.826458692550659,
+      "learning_rate": 1.499451754385965e-05,
+      "loss": 1.7144,
+      "step": 5478
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7944324016571045,
+      "learning_rate": 1.4993603801169591e-05,
+      "loss": 1.4423,
+      "step": 5479
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.808997392654419,
+      "learning_rate": 1.4992690058479533e-05,
+      "loss": 1.7739,
+      "step": 5480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6634225845336914,
+      "learning_rate": 1.4991776315789475e-05,
+      "loss": 1.6826,
+      "step": 5481
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8349125385284424,
+      "learning_rate": 1.4990862573099416e-05,
+      "loss": 1.6803,
+      "step": 5482
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.690931558609009,
+      "learning_rate": 1.4989948830409358e-05,
+      "loss": 1.4844,
+      "step": 5483
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.096341609954834,
+      "learning_rate": 1.4989035087719299e-05,
+      "loss": 1.7009,
+      "step": 5484
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.075101852416992,
+      "learning_rate": 1.498812134502924e-05,
+      "loss": 1.4785,
+      "step": 5485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.804126501083374,
+      "learning_rate": 1.4987207602339181e-05,
+      "loss": 1.5921,
+      "step": 5486
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.681056976318359,
+      "learning_rate": 1.4986293859649124e-05,
+      "loss": 1.5722,
+      "step": 5487
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8788108825683594,
+      "learning_rate": 1.4985380116959066e-05,
+      "loss": 1.6741,
+      "step": 5488
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7365365028381348,
+      "learning_rate": 1.4984466374269006e-05,
+      "loss": 1.4961,
+      "step": 5489
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4637274742126465,
+      "learning_rate": 1.4983552631578949e-05,
+      "loss": 1.4528,
+      "step": 5490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.636704444885254,
+      "learning_rate": 1.4982638888888889e-05,
+      "loss": 1.5652,
+      "step": 5491
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7449471950531006,
+      "learning_rate": 1.4981725146198831e-05,
+      "loss": 1.5116,
+      "step": 5492
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5366370677948,
+      "learning_rate": 1.4980811403508772e-05,
+      "loss": 1.583,
+      "step": 5493
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.948221206665039,
+      "learning_rate": 1.4979897660818714e-05,
+      "loss": 1.7252,
+      "step": 5494
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7452123165130615,
+      "learning_rate": 1.4978983918128656e-05,
+      "loss": 1.5627,
+      "step": 5495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.87701153755188,
+      "learning_rate": 1.4978070175438597e-05,
+      "loss": 1.5856,
+      "step": 5496
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9436452388763428,
+      "learning_rate": 1.4977156432748539e-05,
+      "loss": 1.8427,
+      "step": 5497
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.961819648742676,
+      "learning_rate": 1.497624269005848e-05,
+      "loss": 1.9551,
+      "step": 5498
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5894954204559326,
+      "learning_rate": 1.4975328947368422e-05,
+      "loss": 1.3973,
+      "step": 5499
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4696779251098633,
+      "learning_rate": 1.4974415204678362e-05,
+      "loss": 1.301,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.478508710861206,
+      "learning_rate": 1.4973501461988304e-05,
+      "loss": 2.1569,
+      "step": 5501
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.267916440963745,
+      "learning_rate": 1.4972587719298247e-05,
+      "loss": 1.2006,
+      "step": 5502
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3237557411193848,
+      "learning_rate": 1.4971673976608187e-05,
+      "loss": 1.6308,
+      "step": 5503
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7791759967803955,
+      "learning_rate": 1.497076023391813e-05,
+      "loss": 1.7333,
+      "step": 5504
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.020115852355957,
+      "learning_rate": 1.496984649122807e-05,
+      "loss": 1.8942,
+      "step": 5505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.046222686767578,
+      "learning_rate": 1.4968932748538012e-05,
+      "loss": 1.6258,
+      "step": 5506
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.631427764892578,
+      "learning_rate": 1.4968019005847953e-05,
+      "loss": 1.5426,
+      "step": 5507
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.353494167327881,
+      "learning_rate": 1.4967105263157895e-05,
+      "loss": 1.1966,
+      "step": 5508
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5932717323303223,
+      "learning_rate": 1.4966191520467837e-05,
+      "loss": 1.4228,
+      "step": 5509
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.334470272064209,
+      "learning_rate": 1.4965277777777778e-05,
+      "loss": 1.1969,
+      "step": 5510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0311636924743652,
+      "learning_rate": 1.496436403508772e-05,
+      "loss": 1.3274,
+      "step": 5511
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.924107313156128,
+      "learning_rate": 1.496345029239766e-05,
+      "loss": 1.2773,
+      "step": 5512
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.573669195175171,
+      "learning_rate": 1.4962536549707603e-05,
+      "loss": 1.2814,
+      "step": 5513
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2067391872406006,
+      "learning_rate": 1.4961622807017543e-05,
+      "loss": 1.1133,
+      "step": 5514
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.282874822616577,
+      "learning_rate": 1.4960709064327487e-05,
+      "loss": 1.6012,
+      "step": 5515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.070775032043457,
+      "learning_rate": 1.495979532163743e-05,
+      "loss": 2.0339,
+      "step": 5516
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1155178546905518,
+      "learning_rate": 1.495888157894737e-05,
+      "loss": 1.9243,
+      "step": 5517
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5077884197235107,
+      "learning_rate": 1.4957967836257312e-05,
+      "loss": 1.6648,
+      "step": 5518
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0621538162231445,
+      "learning_rate": 1.4957054093567254e-05,
+      "loss": 1.5311,
+      "step": 5519
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.578559160232544,
+      "learning_rate": 1.4956140350877195e-05,
+      "loss": 1.8286,
+      "step": 5520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.145209312438965,
+      "learning_rate": 1.4955226608187137e-05,
+      "loss": 1.3973,
+      "step": 5521
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7377259731292725,
+      "learning_rate": 1.4954312865497078e-05,
+      "loss": 1.4209,
+      "step": 5522
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2092409133911133,
+      "learning_rate": 1.495339912280702e-05,
+      "loss": 1.5744,
+      "step": 5523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.936176300048828,
+      "learning_rate": 1.495248538011696e-05,
+      "loss": 1.4779,
+      "step": 5524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.853040933609009,
+      "learning_rate": 1.4951571637426903e-05,
+      "loss": 1.8,
+      "step": 5525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.420564889907837,
+      "learning_rate": 1.4950657894736845e-05,
+      "loss": 1.3705,
+      "step": 5526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0745224952697754,
+      "learning_rate": 1.4949744152046785e-05,
+      "loss": 1.3989,
+      "step": 5527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.980050563812256,
+      "learning_rate": 1.4948830409356728e-05,
+      "loss": 1.8716,
+      "step": 5528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2973692417144775,
+      "learning_rate": 1.4947916666666668e-05,
+      "loss": 2.0426,
+      "step": 5529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.820119619369507,
+      "learning_rate": 1.494700292397661e-05,
+      "loss": 1.4537,
+      "step": 5530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8900339603424072,
+      "learning_rate": 1.494608918128655e-05,
+      "loss": 1.9444,
+      "step": 5531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5959506034851074,
+      "learning_rate": 1.4945175438596493e-05,
+      "loss": 1.441,
+      "step": 5532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1496195793151855,
+      "learning_rate": 1.4944261695906435e-05,
+      "loss": 1.6248,
+      "step": 5533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.637450695037842,
+      "learning_rate": 1.4943347953216376e-05,
+      "loss": 1.5114,
+      "step": 5534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.013136625289917,
+      "learning_rate": 1.4942434210526318e-05,
+      "loss": 1.781,
+      "step": 5535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.896873950958252,
+      "learning_rate": 1.4941520467836259e-05,
+      "loss": 1.3418,
+      "step": 5536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.921611785888672,
+      "learning_rate": 1.49406067251462e-05,
+      "loss": 1.5541,
+      "step": 5537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6361491680145264,
+      "learning_rate": 1.4939692982456141e-05,
+      "loss": 1.5214,
+      "step": 5538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9508848190307617,
+      "learning_rate": 1.4938779239766083e-05,
+      "loss": 1.7929,
+      "step": 5539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1182963848114014,
+      "learning_rate": 1.4937865497076026e-05,
+      "loss": 1.5745,
+      "step": 5540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0115654468536377,
+      "learning_rate": 1.4936951754385966e-05,
+      "loss": 1.2296,
+      "step": 5541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.836763858795166,
+      "learning_rate": 1.4936038011695908e-05,
+      "loss": 1.7658,
+      "step": 5542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.303478717803955,
+      "learning_rate": 1.4935124269005849e-05,
+      "loss": 1.4528,
+      "step": 5543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6803390979766846,
+      "learning_rate": 1.4934210526315791e-05,
+      "loss": 1.3794,
+      "step": 5544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4512834548950195,
+      "learning_rate": 1.4933296783625732e-05,
+      "loss": 1.7844,
+      "step": 5545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7442471981048584,
+      "learning_rate": 1.4932383040935674e-05,
+      "loss": 1.5558,
+      "step": 5546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1639974117279053,
+      "learning_rate": 1.4931469298245616e-05,
+      "loss": 1.5257,
+      "step": 5547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0716238021850586,
+      "learning_rate": 1.4930555555555557e-05,
+      "loss": 1.5099,
+      "step": 5548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.723548650741577,
+      "learning_rate": 1.4929641812865499e-05,
+      "loss": 1.6859,
+      "step": 5549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.995598077774048,
+      "learning_rate": 1.492872807017544e-05,
+      "loss": 1.4986,
+      "step": 5550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3137195110321045,
+      "learning_rate": 1.4927814327485382e-05,
+      "loss": 1.406,
+      "step": 5551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8074755668640137,
+      "learning_rate": 1.4926900584795322e-05,
+      "loss": 1.6173,
+      "step": 5552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.486809730529785,
+      "learning_rate": 1.4925986842105264e-05,
+      "loss": 1.9884,
+      "step": 5553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9670307636260986,
+      "learning_rate": 1.4925073099415207e-05,
+      "loss": 1.5014,
+      "step": 5554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.727348804473877,
+      "learning_rate": 1.4924159356725147e-05,
+      "loss": 1.5444,
+      "step": 5555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5942792892456055,
+      "learning_rate": 1.492324561403509e-05,
+      "loss": 1.5732,
+      "step": 5556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.873307704925537,
+      "learning_rate": 1.492233187134503e-05,
+      "loss": 1.7844,
+      "step": 5557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6642866134643555,
+      "learning_rate": 1.4921418128654972e-05,
+      "loss": 1.3803,
+      "step": 5558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.316070556640625,
+      "learning_rate": 1.4920504385964913e-05,
+      "loss": 2.1391,
+      "step": 5559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.223665714263916,
+      "learning_rate": 1.4919590643274855e-05,
+      "loss": 1.9395,
+      "step": 5560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4633047580718994,
+      "learning_rate": 1.4918676900584797e-05,
+      "loss": 1.3582,
+      "step": 5561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.861968755722046,
+      "learning_rate": 1.4917763157894738e-05,
+      "loss": 1.9639,
+      "step": 5562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5937275886535645,
+      "learning_rate": 1.491684941520468e-05,
+      "loss": 1.8037,
+      "step": 5563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6083528995513916,
+      "learning_rate": 1.491593567251462e-05,
+      "loss": 1.4621,
+      "step": 5564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.959719181060791,
+      "learning_rate": 1.4915021929824563e-05,
+      "loss": 1.7457,
+      "step": 5565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.589911460876465,
+      "learning_rate": 1.4914108187134503e-05,
+      "loss": 1.728,
+      "step": 5566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5451252460479736,
+      "learning_rate": 1.4913194444444445e-05,
+      "loss": 1.543,
+      "step": 5567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7330901622772217,
+      "learning_rate": 1.4912280701754388e-05,
+      "loss": 1.7097,
+      "step": 5568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2214620113372803,
+      "learning_rate": 1.4911366959064328e-05,
+      "loss": 1.7113,
+      "step": 5569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.835425853729248,
+      "learning_rate": 1.491045321637427e-05,
+      "loss": 1.5746,
+      "step": 5570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9426746368408203,
+      "learning_rate": 1.4909539473684211e-05,
+      "loss": 1.1267,
+      "step": 5571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9818637371063232,
+      "learning_rate": 1.4908625730994153e-05,
+      "loss": 1.7274,
+      "step": 5572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5145697593688965,
+      "learning_rate": 1.4907711988304094e-05,
+      "loss": 1.6689,
+      "step": 5573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7020723819732666,
+      "learning_rate": 1.4906798245614036e-05,
+      "loss": 1.5939,
+      "step": 5574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5245909690856934,
+      "learning_rate": 1.4905884502923978e-05,
+      "loss": 2.2595,
+      "step": 5575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.087373733520508,
+      "learning_rate": 1.4904970760233919e-05,
+      "loss": 1.6836,
+      "step": 5576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.812155246734619,
+      "learning_rate": 1.490405701754386e-05,
+      "loss": 1.4931,
+      "step": 5577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.849919557571411,
+      "learning_rate": 1.4903143274853801e-05,
+      "loss": 1.5557,
+      "step": 5578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.793048858642578,
+      "learning_rate": 1.4902229532163744e-05,
+      "loss": 1.7961,
+      "step": 5579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.105710983276367,
+      "learning_rate": 1.4901315789473684e-05,
+      "loss": 1.5993,
+      "step": 5580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3361406326293945,
+      "learning_rate": 1.4900402046783626e-05,
+      "loss": 1.4365,
+      "step": 5581
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.149625778198242,
+      "learning_rate": 1.4899488304093569e-05,
+      "loss": 1.6735,
+      "step": 5582
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0126419067382812,
+      "learning_rate": 1.4898574561403509e-05,
+      "loss": 1.8528,
+      "step": 5583
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8103244304656982,
+      "learning_rate": 1.4897660818713451e-05,
+      "loss": 1.438,
+      "step": 5584
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0769360065460205,
+      "learning_rate": 1.4896747076023392e-05,
+      "loss": 1.8673,
+      "step": 5585
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.895702600479126,
+      "learning_rate": 1.4895833333333334e-05,
+      "loss": 1.628,
+      "step": 5586
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2331972122192383,
+      "learning_rate": 1.4894919590643275e-05,
+      "loss": 1.9351,
+      "step": 5587
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2316811084747314,
+      "learning_rate": 1.4894005847953217e-05,
+      "loss": 1.146,
+      "step": 5588
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0501303672790527,
+      "learning_rate": 1.4893092105263159e-05,
+      "loss": 1.5792,
+      "step": 5589
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2693629264831543,
+      "learning_rate": 1.48921783625731e-05,
+      "loss": 2.1567,
+      "step": 5590
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.490464210510254,
+      "learning_rate": 1.4891264619883042e-05,
+      "loss": 1.9465,
+      "step": 5591
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6036572456359863,
+      "learning_rate": 1.4890350877192982e-05,
+      "loss": 1.6643,
+      "step": 5592
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.694124460220337,
+      "learning_rate": 1.4889437134502924e-05,
+      "loss": 1.3814,
+      "step": 5593
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1966543197631836,
+      "learning_rate": 1.4888523391812865e-05,
+      "loss": 1.7624,
+      "step": 5594
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2650792598724365,
+      "learning_rate": 1.4887609649122807e-05,
+      "loss": 1.4236,
+      "step": 5595
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.372316598892212,
+      "learning_rate": 1.488669590643275e-05,
+      "loss": 1.3675,
+      "step": 5596
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.029040336608887,
+      "learning_rate": 1.488578216374269e-05,
+      "loss": 1.8525,
+      "step": 5597
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.555039644241333,
+      "learning_rate": 1.4884868421052632e-05,
+      "loss": 1.3436,
+      "step": 5598
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6200499534606934,
+      "learning_rate": 1.4883954678362573e-05,
+      "loss": 1.4389,
+      "step": 5599
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.512524366378784,
+      "learning_rate": 1.4883040935672515e-05,
+      "loss": 1.2774,
+      "step": 5600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.352337598800659,
+      "learning_rate": 1.4882127192982459e-05,
+      "loss": 1.556,
+      "step": 5601
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.532153367996216,
+      "learning_rate": 1.48812134502924e-05,
+      "loss": 1.4292,
+      "step": 5602
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5138051509857178,
+      "learning_rate": 1.4880299707602342e-05,
+      "loss": 1.5313,
+      "step": 5603
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5059757232666016,
+      "learning_rate": 1.4879385964912282e-05,
+      "loss": 1.9036,
+      "step": 5604
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8638339042663574,
+      "learning_rate": 1.4878472222222224e-05,
+      "loss": 1.5628,
+      "step": 5605
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.140212297439575,
+      "learning_rate": 1.4877558479532167e-05,
+      "loss": 1.4716,
+      "step": 5606
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5187346935272217,
+      "learning_rate": 1.4876644736842107e-05,
+      "loss": 1.7058,
+      "step": 5607
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.98895525932312,
+      "learning_rate": 1.487573099415205e-05,
+      "loss": 1.9055,
+      "step": 5608
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.571481704711914,
+      "learning_rate": 1.487481725146199e-05,
+      "loss": 1.9369,
+      "step": 5609
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7280869483947754,
+      "learning_rate": 1.4873903508771932e-05,
+      "loss": 1.4648,
+      "step": 5610
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7624998092651367,
+      "learning_rate": 1.4872989766081873e-05,
+      "loss": 1.5656,
+      "step": 5611
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.760974407196045,
+      "learning_rate": 1.4872076023391815e-05,
+      "loss": 1.3686,
+      "step": 5612
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4408137798309326,
+      "learning_rate": 1.4871162280701757e-05,
+      "loss": 1.2189,
+      "step": 5613
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.80330753326416,
+      "learning_rate": 1.4870248538011698e-05,
+      "loss": 1.4572,
+      "step": 5614
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.794252872467041,
+      "learning_rate": 1.486933479532164e-05,
+      "loss": 1.5225,
+      "step": 5615
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7453453540802,
+      "learning_rate": 1.486842105263158e-05,
+      "loss": 1.6836,
+      "step": 5616
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.253161907196045,
+      "learning_rate": 1.4867507309941523e-05,
+      "loss": 1.6139,
+      "step": 5617
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.048281669616699,
+      "learning_rate": 1.4866593567251463e-05,
+      "loss": 1.8389,
+      "step": 5618
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2888894081115723,
+      "learning_rate": 1.4865679824561405e-05,
+      "loss": 1.4132,
+      "step": 5619
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.62660551071167,
+      "learning_rate": 1.4864766081871348e-05,
+      "loss": 1.5214,
+      "step": 5620
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9050798416137695,
+      "learning_rate": 1.4863852339181288e-05,
+      "loss": 1.5324,
+      "step": 5621
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9226438999176025,
+      "learning_rate": 1.486293859649123e-05,
+      "loss": 1.5772,
+      "step": 5622
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6755869388580322,
+      "learning_rate": 1.486202485380117e-05,
+      "loss": 1.6317,
+      "step": 5623
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9253408908843994,
+      "learning_rate": 1.4861111111111113e-05,
+      "loss": 1.5377,
+      "step": 5624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.110107660293579,
+      "learning_rate": 1.4860197368421054e-05,
+      "loss": 1.8703,
+      "step": 5625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.161278009414673,
+      "learning_rate": 1.4859283625730996e-05,
+      "loss": 1.2307,
+      "step": 5626
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.1612396240234375,
+      "learning_rate": 1.4858369883040938e-05,
+      "loss": 1.5734,
+      "step": 5627
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1614134311676025,
+      "learning_rate": 1.4857456140350879e-05,
+      "loss": 1.5533,
+      "step": 5628
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7784883975982666,
+      "learning_rate": 1.485654239766082e-05,
+      "loss": 1.5823,
+      "step": 5629
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.386467218399048,
+      "learning_rate": 1.4855628654970761e-05,
+      "loss": 2.1657,
+      "step": 5630
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.477149248123169,
+      "learning_rate": 1.4854714912280703e-05,
+      "loss": 1.5387,
+      "step": 5631
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.752094030380249,
+      "learning_rate": 1.4853801169590644e-05,
+      "loss": 1.7172,
+      "step": 5632
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8255438804626465,
+      "learning_rate": 1.4852887426900586e-05,
+      "loss": 1.6654,
+      "step": 5633
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.694406032562256,
+      "learning_rate": 1.4851973684210528e-05,
+      "loss": 1.4389,
+      "step": 5634
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.140787124633789,
+      "learning_rate": 1.4851059941520469e-05,
+      "loss": 1.811,
+      "step": 5635
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3379592895507812,
+      "learning_rate": 1.4850146198830411e-05,
+      "loss": 1.6895,
+      "step": 5636
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0030384063720703,
+      "learning_rate": 1.4849232456140352e-05,
+      "loss": 1.3925,
+      "step": 5637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.807201862335205,
+      "learning_rate": 1.4848318713450294e-05,
+      "loss": 1.5663,
+      "step": 5638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.591416597366333,
+      "learning_rate": 1.4847404970760234e-05,
+      "loss": 1.3557,
+      "step": 5639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2541465759277344,
+      "learning_rate": 1.4846491228070177e-05,
+      "loss": 1.665,
+      "step": 5640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0590405464172363,
+      "learning_rate": 1.4845577485380119e-05,
+      "loss": 1.479,
+      "step": 5641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.74165678024292,
+      "learning_rate": 1.484466374269006e-05,
+      "loss": 1.4109,
+      "step": 5642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7414562702178955,
+      "learning_rate": 1.4843750000000002e-05,
+      "loss": 1.3035,
+      "step": 5643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.384528160095215,
+      "learning_rate": 1.4842836257309942e-05,
+      "loss": 1.9191,
+      "step": 5644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0507242679595947,
+      "learning_rate": 1.4841922514619884e-05,
+      "loss": 1.7354,
+      "step": 5645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4804434776306152,
+      "learning_rate": 1.4841008771929825e-05,
+      "loss": 1.5251,
+      "step": 5646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1170992851257324,
+      "learning_rate": 1.4840095029239767e-05,
+      "loss": 1.7633,
+      "step": 5647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8639323711395264,
+      "learning_rate": 1.483918128654971e-05,
+      "loss": 1.4055,
+      "step": 5648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.888040065765381,
+      "learning_rate": 1.483826754385965e-05,
+      "loss": 1.4658,
+      "step": 5649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4653706550598145,
+      "learning_rate": 1.4837353801169592e-05,
+      "loss": 1.2931,
+      "step": 5650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9399640560150146,
+      "learning_rate": 1.4836440058479533e-05,
+      "loss": 1.6892,
+      "step": 5651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4541773796081543,
+      "learning_rate": 1.4835526315789475e-05,
+      "loss": 1.9073,
+      "step": 5652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.732346296310425,
+      "learning_rate": 1.4834612573099415e-05,
+      "loss": 1.6755,
+      "step": 5653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6887853145599365,
+      "learning_rate": 1.4833698830409358e-05,
+      "loss": 1.6105,
+      "step": 5654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5298397541046143,
+      "learning_rate": 1.48327850877193e-05,
+      "loss": 1.6606,
+      "step": 5655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1255452632904053,
+      "learning_rate": 1.483187134502924e-05,
+      "loss": 1.8084,
+      "step": 5656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5303783416748047,
+      "learning_rate": 1.4830957602339183e-05,
+      "loss": 1.2356,
+      "step": 5657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7006566524505615,
+      "learning_rate": 1.4830043859649123e-05,
+      "loss": 2.0095,
+      "step": 5658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.066999912261963,
+      "learning_rate": 1.4829130116959065e-05,
+      "loss": 1.7316,
+      "step": 5659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8281075954437256,
+      "learning_rate": 1.4828216374269006e-05,
+      "loss": 1.3406,
+      "step": 5660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7140848636627197,
+      "learning_rate": 1.4827302631578948e-05,
+      "loss": 1.4926,
+      "step": 5661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3178844451904297,
+      "learning_rate": 1.482638888888889e-05,
+      "loss": 1.7586,
+      "step": 5662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5683653354644775,
+      "learning_rate": 1.482547514619883e-05,
+      "loss": 1.4317,
+      "step": 5663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1002581119537354,
+      "learning_rate": 1.4824561403508773e-05,
+      "loss": 1.731,
+      "step": 5664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1873939037323,
+      "learning_rate": 1.4823647660818714e-05,
+      "loss": 1.8755,
+      "step": 5665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.266352415084839,
+      "learning_rate": 1.4822733918128656e-05,
+      "loss": 1.7391,
+      "step": 5666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0014030933380127,
+      "learning_rate": 1.4821820175438596e-05,
+      "loss": 1.9139,
+      "step": 5667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.959256172180176,
+      "learning_rate": 1.4820906432748539e-05,
+      "loss": 1.6309,
+      "step": 5668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.566417932510376,
+      "learning_rate": 1.481999269005848e-05,
+      "loss": 1.5581,
+      "step": 5669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.446230411529541,
+      "learning_rate": 1.4819078947368421e-05,
+      "loss": 1.9381,
+      "step": 5670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4730892181396484,
+      "learning_rate": 1.4818165204678364e-05,
+      "loss": 1.3295,
+      "step": 5671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.394057273864746,
+      "learning_rate": 1.4817251461988304e-05,
+      "loss": 1.4805,
+      "step": 5672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1220264434814453,
+      "learning_rate": 1.4816337719298246e-05,
+      "loss": 1.8436,
+      "step": 5673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7104814052581787,
+      "learning_rate": 1.4815423976608187e-05,
+      "loss": 1.2767,
+      "step": 5674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1759798526763916,
+      "learning_rate": 1.4814510233918129e-05,
+      "loss": 1.7222,
+      "step": 5675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6498312950134277,
+      "learning_rate": 1.4813596491228071e-05,
+      "loss": 1.8777,
+      "step": 5676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.906663417816162,
+      "learning_rate": 1.4812682748538012e-05,
+      "loss": 2.0992,
+      "step": 5677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4352707862854004,
+      "learning_rate": 1.4811769005847954e-05,
+      "loss": 1.9489,
+      "step": 5678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0477211475372314,
+      "learning_rate": 1.4810855263157895e-05,
+      "loss": 1.8359,
+      "step": 5679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.521376132965088,
+      "learning_rate": 1.4809941520467837e-05,
+      "loss": 2.1062,
+      "step": 5680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.900352716445923,
+      "learning_rate": 1.4809027777777777e-05,
+      "loss": 1.6684,
+      "step": 5681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4893088340759277,
+      "learning_rate": 1.480811403508772e-05,
+      "loss": 1.395,
+      "step": 5682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5580127239227295,
+      "learning_rate": 1.4807200292397662e-05,
+      "loss": 1.1745,
+      "step": 5683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3150267601013184,
+      "learning_rate": 1.4806286549707602e-05,
+      "loss": 1.4147,
+      "step": 5684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3019256591796875,
+      "learning_rate": 1.4805372807017544e-05,
+      "loss": 1.5749,
+      "step": 5685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.363223075866699,
+      "learning_rate": 1.4804459064327488e-05,
+      "loss": 1.8824,
+      "step": 5686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7575559616088867,
+      "learning_rate": 1.4803545321637429e-05,
+      "loss": 1.7003,
+      "step": 5687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8367364406585693,
+      "learning_rate": 1.4802631578947371e-05,
+      "loss": 1.6718,
+      "step": 5688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.589792251586914,
+      "learning_rate": 1.4801717836257312e-05,
+      "loss": 1.434,
+      "step": 5689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4093570709228516,
+      "learning_rate": 1.4800804093567254e-05,
+      "loss": 1.3902,
+      "step": 5690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.661247730255127,
+      "learning_rate": 1.4799890350877194e-05,
+      "loss": 1.5621,
+      "step": 5691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8733325004577637,
+      "learning_rate": 1.4798976608187137e-05,
+      "loss": 1.4897,
+      "step": 5692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.229257822036743,
+      "learning_rate": 1.4798062865497079e-05,
+      "loss": 1.7848,
+      "step": 5693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.747816324234009,
+      "learning_rate": 1.479714912280702e-05,
+      "loss": 1.4127,
+      "step": 5694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.817887544631958,
+      "learning_rate": 1.4796235380116962e-05,
+      "loss": 1.5793,
+      "step": 5695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.793529510498047,
+      "learning_rate": 1.4795321637426902e-05,
+      "loss": 1.8898,
+      "step": 5696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1879260540008545,
+      "learning_rate": 1.4794407894736844e-05,
+      "loss": 1.8028,
+      "step": 5697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.516667127609253,
+      "learning_rate": 1.4793494152046785e-05,
+      "loss": 1.7095,
+      "step": 5698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9769411087036133,
+      "learning_rate": 1.4792580409356727e-05,
+      "loss": 1.6573,
+      "step": 5699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1160082817077637,
+      "learning_rate": 1.479166666666667e-05,
+      "loss": 1.8403,
+      "step": 5700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6052091121673584,
+      "learning_rate": 1.479075292397661e-05,
+      "loss": 1.5444,
+      "step": 5701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1233742237091064,
+      "learning_rate": 1.4789839181286552e-05,
+      "loss": 1.2415,
+      "step": 5702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6020641326904297,
+      "learning_rate": 1.4788925438596493e-05,
+      "loss": 1.5859,
+      "step": 5703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6485092639923096,
+      "learning_rate": 1.4788011695906435e-05,
+      "loss": 1.3385,
+      "step": 5704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.685985565185547,
+      "learning_rate": 1.4787097953216375e-05,
+      "loss": 1.8232,
+      "step": 5705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6381030082702637,
+      "learning_rate": 1.4786184210526318e-05,
+      "loss": 1.5303,
+      "step": 5706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3262441158294678,
+      "learning_rate": 1.478527046783626e-05,
+      "loss": 1.9107,
+      "step": 5707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.537007808685303,
+      "learning_rate": 1.47843567251462e-05,
+      "loss": 1.7574,
+      "step": 5708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.516822338104248,
+      "learning_rate": 1.4783442982456143e-05,
+      "loss": 1.3957,
+      "step": 5709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.082767963409424,
+      "learning_rate": 1.4782529239766083e-05,
+      "loss": 1.7587,
+      "step": 5710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.241929054260254,
+      "learning_rate": 1.4781615497076025e-05,
+      "loss": 1.6541,
+      "step": 5711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2985665798187256,
+      "learning_rate": 1.4780701754385966e-05,
+      "loss": 1.3229,
+      "step": 5712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.963270664215088,
+      "learning_rate": 1.4779788011695908e-05,
+      "loss": 1.6549,
+      "step": 5713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1179263591766357,
+      "learning_rate": 1.477887426900585e-05,
+      "loss": 1.4204,
+      "step": 5714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3099594116210938,
+      "learning_rate": 1.477796052631579e-05,
+      "loss": 1.7878,
+      "step": 5715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9363133907318115,
+      "learning_rate": 1.4777046783625733e-05,
+      "loss": 1.6994,
+      "step": 5716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5893032550811768,
+      "learning_rate": 1.4776133040935674e-05,
+      "loss": 1.6887,
+      "step": 5717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.075063705444336,
+      "learning_rate": 1.4775219298245616e-05,
+      "loss": 2.1131,
+      "step": 5718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2414491176605225,
+      "learning_rate": 1.4774305555555556e-05,
+      "loss": 1.0862,
+      "step": 5719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.705289125442505,
+      "learning_rate": 1.4773391812865499e-05,
+      "loss": 1.947,
+      "step": 5720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9311156272888184,
+      "learning_rate": 1.477247807017544e-05,
+      "loss": 1.8695,
+      "step": 5721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3410284519195557,
+      "learning_rate": 1.4771564327485381e-05,
+      "loss": 1.6413,
+      "step": 5722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7206737995147705,
+      "learning_rate": 1.4770650584795323e-05,
+      "loss": 1.3854,
+      "step": 5723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8286750316619873,
+      "learning_rate": 1.4769736842105264e-05,
+      "loss": 1.5563,
+      "step": 5724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0064053535461426,
+      "learning_rate": 1.4768823099415206e-05,
+      "loss": 1.5958,
+      "step": 5725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4097740650177,
+      "learning_rate": 1.4767909356725147e-05,
+      "loss": 1.5171,
+      "step": 5726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2647652626037598,
+      "learning_rate": 1.4766995614035089e-05,
+      "loss": 1.1133,
+      "step": 5727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8302903175354004,
+      "learning_rate": 1.4766081871345031e-05,
+      "loss": 1.6928,
+      "step": 5728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.115468978881836,
+      "learning_rate": 1.4765168128654972e-05,
+      "loss": 1.6376,
+      "step": 5729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.967442274093628,
+      "learning_rate": 1.4764254385964914e-05,
+      "loss": 1.4391,
+      "step": 5730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.150604248046875,
+      "learning_rate": 1.4763340643274854e-05,
+      "loss": 1.9256,
+      "step": 5731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.507420301437378,
+      "learning_rate": 1.4762426900584797e-05,
+      "loss": 1.3971,
+      "step": 5732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.271456718444824,
+      "learning_rate": 1.4761513157894737e-05,
+      "loss": 1.9031,
+      "step": 5733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8767404556274414,
+      "learning_rate": 1.476059941520468e-05,
+      "loss": 2.0089,
+      "step": 5734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3295483589172363,
+      "learning_rate": 1.4759685672514622e-05,
+      "loss": 1.517,
+      "step": 5735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.056319236755371,
+      "learning_rate": 1.4758771929824562e-05,
+      "loss": 1.543,
+      "step": 5736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.681877374649048,
+      "learning_rate": 1.4757858187134504e-05,
+      "loss": 1.4789,
+      "step": 5737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.404294967651367,
+      "learning_rate": 1.4756944444444445e-05,
+      "loss": 1.521,
+      "step": 5738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.233461618423462,
+      "learning_rate": 1.4756030701754387e-05,
+      "loss": 1.8653,
+      "step": 5739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4229588508605957,
+      "learning_rate": 1.4755116959064328e-05,
+      "loss": 1.3835,
+      "step": 5740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7559568881988525,
+      "learning_rate": 1.475420321637427e-05,
+      "loss": 1.1955,
+      "step": 5741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7358429431915283,
+      "learning_rate": 1.4753289473684212e-05,
+      "loss": 1.5493,
+      "step": 5742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5114266872406006,
+      "learning_rate": 1.4752375730994153e-05,
+      "loss": 1.4882,
+      "step": 5743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5204102993011475,
+      "learning_rate": 1.4751461988304095e-05,
+      "loss": 1.4812,
+      "step": 5744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4242963790893555,
+      "learning_rate": 1.4750548245614035e-05,
+      "loss": 1.4413,
+      "step": 5745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.618577718734741,
+      "learning_rate": 1.4749634502923978e-05,
+      "loss": 1.5702,
+      "step": 5746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3432729244232178,
+      "learning_rate": 1.4748720760233918e-05,
+      "loss": 1.5832,
+      "step": 5747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6617112159729004,
+      "learning_rate": 1.474780701754386e-05,
+      "loss": 1.4367,
+      "step": 5748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4027044773101807,
+      "learning_rate": 1.4746893274853803e-05,
+      "loss": 1.374,
+      "step": 5749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1237378120422363,
+      "learning_rate": 1.4745979532163743e-05,
+      "loss": 1.6604,
+      "step": 5750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.235267162322998,
+      "learning_rate": 1.4745065789473685e-05,
+      "loss": 1.5977,
+      "step": 5751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.400432586669922,
+      "learning_rate": 1.4744152046783626e-05,
+      "loss": 1.4003,
+      "step": 5752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8832080364227295,
+      "learning_rate": 1.4743238304093568e-05,
+      "loss": 1.5523,
+      "step": 5753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9589438438415527,
+      "learning_rate": 1.4742324561403509e-05,
+      "loss": 1.381,
+      "step": 5754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7701003551483154,
+      "learning_rate": 1.474141081871345e-05,
+      "loss": 1.8541,
+      "step": 5755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3099193572998047,
+      "learning_rate": 1.4740497076023393e-05,
+      "loss": 1.3434,
+      "step": 5756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.515643358230591,
+      "learning_rate": 1.4739583333333334e-05,
+      "loss": 2.2214,
+      "step": 5757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7608370780944824,
+      "learning_rate": 1.4738669590643276e-05,
+      "loss": 1.4328,
+      "step": 5758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.392058849334717,
+      "learning_rate": 1.4737755847953216e-05,
+      "loss": 2.0399,
+      "step": 5759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.991332769393921,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 1.8144,
+      "step": 5760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.343219518661499,
+      "learning_rate": 1.4735928362573099e-05,
+      "loss": 1.8674,
+      "step": 5761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.585369348526001,
+      "learning_rate": 1.4735014619883041e-05,
+      "loss": 1.5087,
+      "step": 5762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2572336196899414,
+      "learning_rate": 1.4734100877192984e-05,
+      "loss": 1.7101,
+      "step": 5763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.986802101135254,
+      "learning_rate": 1.4733187134502924e-05,
+      "loss": 1.6456,
+      "step": 5764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.054372787475586,
+      "learning_rate": 1.4732273391812866e-05,
+      "loss": 1.5907,
+      "step": 5765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.806117534637451,
+      "learning_rate": 1.4731359649122807e-05,
+      "loss": 1.5371,
+      "step": 5766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.944645643234253,
+      "learning_rate": 1.4730445906432749e-05,
+      "loss": 1.4261,
+      "step": 5767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9071643352508545,
+      "learning_rate": 1.472953216374269e-05,
+      "loss": 1.9303,
+      "step": 5768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5426220893859863,
+      "learning_rate": 1.4728618421052632e-05,
+      "loss": 1.9416,
+      "step": 5769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9660260677337646,
+      "learning_rate": 1.4727704678362574e-05,
+      "loss": 1.5633,
+      "step": 5770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.90264892578125,
+      "learning_rate": 1.4726790935672515e-05,
+      "loss": 1.5477,
+      "step": 5771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0560805797576904,
+      "learning_rate": 1.4725877192982458e-05,
+      "loss": 1.6972,
+      "step": 5772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.549204111099243,
+      "learning_rate": 1.47249634502924e-05,
+      "loss": 1.484,
+      "step": 5773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6263949871063232,
+      "learning_rate": 1.4724049707602341e-05,
+      "loss": 1.4704,
+      "step": 5774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.609750986099243,
+      "learning_rate": 1.4723135964912283e-05,
+      "loss": 1.5219,
+      "step": 5775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.458449602127075,
+      "learning_rate": 1.4722222222222224e-05,
+      "loss": 1.5349,
+      "step": 5776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.50100040435791,
+      "learning_rate": 1.4721308479532166e-05,
+      "loss": 1.6161,
+      "step": 5777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.710205554962158,
+      "learning_rate": 1.4720394736842107e-05,
+      "loss": 1.2811,
+      "step": 5778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.830606698989868,
+      "learning_rate": 1.4719480994152049e-05,
+      "loss": 1.7084,
+      "step": 5779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6115896701812744,
+      "learning_rate": 1.4718567251461991e-05,
+      "loss": 1.7504,
+      "step": 5780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.831085205078125,
+      "learning_rate": 1.4717653508771932e-05,
+      "loss": 1.653,
+      "step": 5781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0029938220977783,
+      "learning_rate": 1.4716739766081874e-05,
+      "loss": 1.382,
+      "step": 5782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2420413494110107,
+      "learning_rate": 1.4715826023391814e-05,
+      "loss": 1.6595,
+      "step": 5783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0971693992614746,
+      "learning_rate": 1.4714912280701757e-05,
+      "loss": 1.3365,
+      "step": 5784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.187941074371338,
+      "learning_rate": 1.4713998538011697e-05,
+      "loss": 1.7237,
+      "step": 5785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7180991172790527,
+      "learning_rate": 1.471308479532164e-05,
+      "loss": 1.8253,
+      "step": 5786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7386951446533203,
+      "learning_rate": 1.4712171052631582e-05,
+      "loss": 1.9687,
+      "step": 5787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8679027557373047,
+      "learning_rate": 1.4711257309941522e-05,
+      "loss": 1.7626,
+      "step": 5788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6824140548706055,
+      "learning_rate": 1.4710343567251464e-05,
+      "loss": 1.5229,
+      "step": 5789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.251384735107422,
+      "learning_rate": 1.4709429824561405e-05,
+      "loss": 1.2769,
+      "step": 5790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.928380250930786,
+      "learning_rate": 1.4708516081871347e-05,
+      "loss": 1.5914,
+      "step": 5791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6165544986724854,
+      "learning_rate": 1.4707602339181288e-05,
+      "loss": 1.8545,
+      "step": 5792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.47328519821167,
+      "learning_rate": 1.470668859649123e-05,
+      "loss": 1.2929,
+      "step": 5793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.763871908187866,
+      "learning_rate": 1.4705774853801172e-05,
+      "loss": 1.6531,
+      "step": 5794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.956865072250366,
+      "learning_rate": 1.4704861111111113e-05,
+      "loss": 1.5645,
+      "step": 5795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0908162593841553,
+      "learning_rate": 1.4703947368421055e-05,
+      "loss": 1.6827,
+      "step": 5796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8034610748291016,
+      "learning_rate": 1.4703033625730995e-05,
+      "loss": 1.4378,
+      "step": 5797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8329455852508545,
+      "learning_rate": 1.4702119883040938e-05,
+      "loss": 1.4735,
+      "step": 5798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2718820571899414,
+      "learning_rate": 1.4701206140350878e-05,
+      "loss": 1.7999,
+      "step": 5799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9084625244140625,
+      "learning_rate": 1.470029239766082e-05,
+      "loss": 1.9043,
+      "step": 5800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.912153482437134,
+      "learning_rate": 1.4699378654970763e-05,
+      "loss": 1.7758,
+      "step": 5801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7825911045074463,
+      "learning_rate": 1.4698464912280703e-05,
+      "loss": 1.4711,
+      "step": 5802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5069262981414795,
+      "learning_rate": 1.4697551169590645e-05,
+      "loss": 1.5244,
+      "step": 5803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.300192356109619,
+      "learning_rate": 1.4696637426900586e-05,
+      "loss": 1.1933,
+      "step": 5804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3269011974334717,
+      "learning_rate": 1.4695723684210528e-05,
+      "loss": 1.2539,
+      "step": 5805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.732903480529785,
+      "learning_rate": 1.4694809941520469e-05,
+      "loss": 1.5864,
+      "step": 5806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2626636028289795,
+      "learning_rate": 1.469389619883041e-05,
+      "loss": 1.6483,
+      "step": 5807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0904462337493896,
+      "learning_rate": 1.4692982456140353e-05,
+      "loss": 1.8587,
+      "step": 5808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8161580562591553,
+      "learning_rate": 1.4692068713450294e-05,
+      "loss": 1.4167,
+      "step": 5809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.309831380844116,
+      "learning_rate": 1.4691154970760236e-05,
+      "loss": 1.9166,
+      "step": 5810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.259278297424316,
+      "learning_rate": 1.4690241228070176e-05,
+      "loss": 2.6205,
+      "step": 5811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9132187366485596,
+      "learning_rate": 1.4689327485380119e-05,
+      "loss": 1.7951,
+      "step": 5812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5825328826904297,
+      "learning_rate": 1.4688413742690059e-05,
+      "loss": 1.4994,
+      "step": 5813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5312790870666504,
+      "learning_rate": 1.4687500000000001e-05,
+      "loss": 1.9933,
+      "step": 5814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.127131223678589,
+      "learning_rate": 1.4686586257309943e-05,
+      "loss": 2.1898,
+      "step": 5815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.742070436477661,
+      "learning_rate": 1.4685672514619884e-05,
+      "loss": 1.5841,
+      "step": 5816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2973685264587402,
+      "learning_rate": 1.4684758771929826e-05,
+      "loss": 1.602,
+      "step": 5817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1042702198028564,
+      "learning_rate": 1.4683845029239767e-05,
+      "loss": 1.6944,
+      "step": 5818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.87386155128479,
+      "learning_rate": 1.4682931286549709e-05,
+      "loss": 1.7132,
+      "step": 5819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2957570552825928,
+      "learning_rate": 1.468201754385965e-05,
+      "loss": 1.1936,
+      "step": 5820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5999038219451904,
+      "learning_rate": 1.4681103801169592e-05,
+      "loss": 1.8806,
+      "step": 5821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0192863941192627,
+      "learning_rate": 1.4680190058479534e-05,
+      "loss": 1.9151,
+      "step": 5822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.821892023086548,
+      "learning_rate": 1.4679276315789474e-05,
+      "loss": 1.5395,
+      "step": 5823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.971987009048462,
+      "learning_rate": 1.4678362573099417e-05,
+      "loss": 1.7281,
+      "step": 5824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.820354461669922,
+      "learning_rate": 1.4677448830409357e-05,
+      "loss": 1.4953,
+      "step": 5825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7722666263580322,
+      "learning_rate": 1.46765350877193e-05,
+      "loss": 1.5143,
+      "step": 5826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.091463804244995,
+      "learning_rate": 1.467562134502924e-05,
+      "loss": 1.7972,
+      "step": 5827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.756298065185547,
+      "learning_rate": 1.4674707602339182e-05,
+      "loss": 1.6,
+      "step": 5828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3228495121002197,
+      "learning_rate": 1.4673793859649124e-05,
+      "loss": 2.524,
+      "step": 5829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8614113330841064,
+      "learning_rate": 1.4672880116959065e-05,
+      "loss": 1.6198,
+      "step": 5830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.716980218887329,
+      "learning_rate": 1.4671966374269007e-05,
+      "loss": 1.3517,
+      "step": 5831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.550435781478882,
+      "learning_rate": 1.4671052631578948e-05,
+      "loss": 1.1824,
+      "step": 5832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6499760150909424,
+      "learning_rate": 1.467013888888889e-05,
+      "loss": 2.1573,
+      "step": 5833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9054131507873535,
+      "learning_rate": 1.466922514619883e-05,
+      "loss": 1.6357,
+      "step": 5834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.489424228668213,
+      "learning_rate": 1.4668311403508773e-05,
+      "loss": 1.376,
+      "step": 5835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5647456645965576,
+      "learning_rate": 1.4667397660818715e-05,
+      "loss": 1.517,
+      "step": 5836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8966212272644043,
+      "learning_rate": 1.4666483918128655e-05,
+      "loss": 1.8523,
+      "step": 5837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0857150554656982,
+      "learning_rate": 1.4665570175438598e-05,
+      "loss": 1.4084,
+      "step": 5838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6631107330322266,
+      "learning_rate": 1.4664656432748538e-05,
+      "loss": 1.4795,
+      "step": 5839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.856520175933838,
+      "learning_rate": 1.466374269005848e-05,
+      "loss": 1.3568,
+      "step": 5840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6788978576660156,
+      "learning_rate": 1.4662828947368421e-05,
+      "loss": 1.3928,
+      "step": 5841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4227867126464844,
+      "learning_rate": 1.4661915204678363e-05,
+      "loss": 1.3013,
+      "step": 5842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.715524911880493,
+      "learning_rate": 1.4661001461988305e-05,
+      "loss": 1.8937,
+      "step": 5843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5370383262634277,
+      "learning_rate": 1.4660087719298246e-05,
+      "loss": 1.6005,
+      "step": 5844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9719762802124023,
+      "learning_rate": 1.4659173976608188e-05,
+      "loss": 1.5854,
+      "step": 5845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.396662712097168,
+      "learning_rate": 1.4658260233918129e-05,
+      "loss": 1.7084,
+      "step": 5846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.487699031829834,
+      "learning_rate": 1.465734649122807e-05,
+      "loss": 1.3949,
+      "step": 5847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1156554222106934,
+      "learning_rate": 1.4656432748538011e-05,
+      "loss": 1.9121,
+      "step": 5848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.114844560623169,
+      "learning_rate": 1.4655519005847954e-05,
+      "loss": 1.8286,
+      "step": 5849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.364377021789551,
+      "learning_rate": 1.4654605263157896e-05,
+      "loss": 1.7125,
+      "step": 5850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.919464111328125,
+      "learning_rate": 1.4653691520467836e-05,
+      "loss": 1.6704,
+      "step": 5851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.563403606414795,
+      "learning_rate": 1.4652777777777779e-05,
+      "loss": 1.7678,
+      "step": 5852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.938206434249878,
+      "learning_rate": 1.4651864035087719e-05,
+      "loss": 1.4566,
+      "step": 5853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9833240509033203,
+      "learning_rate": 1.4650950292397661e-05,
+      "loss": 1.6496,
+      "step": 5854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3597002029418945,
+      "learning_rate": 1.4650036549707602e-05,
+      "loss": 1.9654,
+      "step": 5855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.80698561668396,
+      "learning_rate": 1.4649122807017544e-05,
+      "loss": 1.776,
+      "step": 5856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.735919952392578,
+      "learning_rate": 1.4648209064327488e-05,
+      "loss": 1.7065,
+      "step": 5857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4958510398864746,
+      "learning_rate": 1.4647295321637429e-05,
+      "loss": 1.5882,
+      "step": 5858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.625136137008667,
+      "learning_rate": 1.464638157894737e-05,
+      "loss": 1.4896,
+      "step": 5859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5734639167785645,
+      "learning_rate": 1.4645467836257313e-05,
+      "loss": 1.4066,
+      "step": 5860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.814484119415283,
+      "learning_rate": 1.4644554093567253e-05,
+      "loss": 1.401,
+      "step": 5861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7059121131896973,
+      "learning_rate": 1.4643640350877196e-05,
+      "loss": 1.4833,
+      "step": 5862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.138436794281006,
+      "learning_rate": 1.4642726608187136e-05,
+      "loss": 1.6218,
+      "step": 5863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.334958791732788,
+      "learning_rate": 1.4641812865497078e-05,
+      "loss": 1.2626,
+      "step": 5864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.791773796081543,
+      "learning_rate": 1.4640899122807019e-05,
+      "loss": 1.5282,
+      "step": 5865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.167588949203491,
+      "learning_rate": 1.4639985380116961e-05,
+      "loss": 1.2905,
+      "step": 5866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9320507049560547,
+      "learning_rate": 1.4639071637426903e-05,
+      "loss": 1.7591,
+      "step": 5867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.113870859146118,
+      "learning_rate": 1.4638157894736844e-05,
+      "loss": 1.2666,
+      "step": 5868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.088139772415161,
+      "learning_rate": 1.4637244152046786e-05,
+      "loss": 1.4169,
+      "step": 5869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.075969934463501,
+      "learning_rate": 1.4636330409356727e-05,
+      "loss": 1.7395,
+      "step": 5870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6303999423980713,
+      "learning_rate": 1.4635416666666669e-05,
+      "loss": 1.7401,
+      "step": 5871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.081054925918579,
+      "learning_rate": 1.463450292397661e-05,
+      "loss": 1.6838,
+      "step": 5872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9479572772979736,
+      "learning_rate": 1.4633589181286552e-05,
+      "loss": 1.5587,
+      "step": 5873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6542515754699707,
+      "learning_rate": 1.4632675438596494e-05,
+      "loss": 1.5984,
+      "step": 5874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1300392150878906,
+      "learning_rate": 1.4631761695906434e-05,
+      "loss": 1.7665,
+      "step": 5875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.528092861175537,
+      "learning_rate": 1.4630847953216377e-05,
+      "loss": 1.7997,
+      "step": 5876
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6133952140808105,
+      "learning_rate": 1.4629934210526317e-05,
+      "loss": 1.5449,
+      "step": 5877
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9255125522613525,
+      "learning_rate": 1.462902046783626e-05,
+      "loss": 1.8793,
+      "step": 5878
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4479053020477295,
+      "learning_rate": 1.46281067251462e-05,
+      "loss": 1.1937,
+      "step": 5879
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4907755851745605,
+      "learning_rate": 1.4627192982456142e-05,
+      "loss": 1.4254,
+      "step": 5880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.672088384628296,
+      "learning_rate": 1.4626279239766084e-05,
+      "loss": 1.5175,
+      "step": 5881
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.794405937194824,
+      "learning_rate": 1.4625365497076025e-05,
+      "loss": 1.8296,
+      "step": 5882
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3727011680603027,
+      "learning_rate": 1.4624451754385967e-05,
+      "loss": 1.3084,
+      "step": 5883
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.733166217803955,
+      "learning_rate": 1.4623538011695908e-05,
+      "loss": 1.6637,
+      "step": 5884
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9796035289764404,
+      "learning_rate": 1.462262426900585e-05,
+      "loss": 1.8958,
+      "step": 5885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5115864276885986,
+      "learning_rate": 1.462171052631579e-05,
+      "loss": 1.6975,
+      "step": 5886
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9991202354431152,
+      "learning_rate": 1.4620796783625733e-05,
+      "loss": 1.5583,
+      "step": 5887
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8584647178649902,
+      "learning_rate": 1.4619883040935675e-05,
+      "loss": 1.8529,
+      "step": 5888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6540369987487793,
+      "learning_rate": 1.4618969298245615e-05,
+      "loss": 1.2205,
+      "step": 5889
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.629321575164795,
+      "learning_rate": 1.4618055555555558e-05,
+      "loss": 1.5927,
+      "step": 5890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.077730178833008,
+      "learning_rate": 1.4617141812865498e-05,
+      "loss": 1.6454,
+      "step": 5891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3398807048797607,
+      "learning_rate": 1.461622807017544e-05,
+      "loss": 1.5698,
+      "step": 5892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0422439575195312,
+      "learning_rate": 1.461531432748538e-05,
+      "loss": 1.6019,
+      "step": 5893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9087281227111816,
+      "learning_rate": 1.4614400584795323e-05,
+      "loss": 1.5102,
+      "step": 5894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0295188426971436,
+      "learning_rate": 1.4613486842105265e-05,
+      "loss": 1.7122,
+      "step": 5895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.671928644180298,
+      "learning_rate": 1.4612573099415206e-05,
+      "loss": 1.6244,
+      "step": 5896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6958844661712646,
+      "learning_rate": 1.4611659356725148e-05,
+      "loss": 1.6517,
+      "step": 5897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7896697521209717,
+      "learning_rate": 1.4610745614035089e-05,
+      "loss": 1.4698,
+      "step": 5898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6604201793670654,
+      "learning_rate": 1.460983187134503e-05,
+      "loss": 1.6399,
+      "step": 5899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8382248878479004,
+      "learning_rate": 1.4608918128654971e-05,
+      "loss": 1.4841,
+      "step": 5900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7099761962890625,
+      "learning_rate": 1.4608004385964914e-05,
+      "loss": 1.8349,
+      "step": 5901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.583824396133423,
+      "learning_rate": 1.4607090643274856e-05,
+      "loss": 1.5677,
+      "step": 5902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5692994594573975,
+      "learning_rate": 1.4606176900584796e-05,
+      "loss": 1.4008,
+      "step": 5903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3688924312591553,
+      "learning_rate": 1.4605263157894739e-05,
+      "loss": 1.4361,
+      "step": 5904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7114980220794678,
+      "learning_rate": 1.4604349415204679e-05,
+      "loss": 1.7517,
+      "step": 5905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.363939046859741,
+      "learning_rate": 1.4603435672514621e-05,
+      "loss": 2.2764,
+      "step": 5906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9956538677215576,
+      "learning_rate": 1.4602521929824562e-05,
+      "loss": 2.1188,
+      "step": 5907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.851583242416382,
+      "learning_rate": 1.4601608187134504e-05,
+      "loss": 1.5489,
+      "step": 5908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.041266441345215,
+      "learning_rate": 1.4600694444444446e-05,
+      "loss": 1.6063,
+      "step": 5909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5575883388519287,
+      "learning_rate": 1.4599780701754387e-05,
+      "loss": 2.2366,
+      "step": 5910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7608280181884766,
+      "learning_rate": 1.4598866959064329e-05,
+      "loss": 1.3724,
+      "step": 5911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9013400077819824,
+      "learning_rate": 1.459795321637427e-05,
+      "loss": 2.0321,
+      "step": 5912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5818636417388916,
+      "learning_rate": 1.4597039473684212e-05,
+      "loss": 1.5704,
+      "step": 5913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3436458110809326,
+      "learning_rate": 1.4596125730994152e-05,
+      "loss": 1.5794,
+      "step": 5914
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1017329692840576,
+      "learning_rate": 1.4595211988304094e-05,
+      "loss": 1.5567,
+      "step": 5915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.64974308013916,
+      "learning_rate": 1.4594298245614037e-05,
+      "loss": 1.5475,
+      "step": 5916
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.619108200073242,
+      "learning_rate": 1.4593384502923977e-05,
+      "loss": 1.5554,
+      "step": 5917
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1413516998291016,
+      "learning_rate": 1.459247076023392e-05,
+      "loss": 1.512,
+      "step": 5918
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.983091354370117,
+      "learning_rate": 1.459155701754386e-05,
+      "loss": 1.6217,
+      "step": 5919
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.079110622406006,
+      "learning_rate": 1.4590643274853802e-05,
+      "loss": 2.2027,
+      "step": 5920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.21122670173645,
+      "learning_rate": 1.4589729532163743e-05,
+      "loss": 1.9936,
+      "step": 5921
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.336442232131958,
+      "learning_rate": 1.4588815789473685e-05,
+      "loss": 1.3044,
+      "step": 5922
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.084099292755127,
+      "learning_rate": 1.4587902046783627e-05,
+      "loss": 1.7278,
+      "step": 5923
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.739835262298584,
+      "learning_rate": 1.4586988304093568e-05,
+      "loss": 1.3351,
+      "step": 5924
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.70570707321167,
+      "learning_rate": 1.458607456140351e-05,
+      "loss": 1.4058,
+      "step": 5925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.70192813873291,
+      "learning_rate": 1.458516081871345e-05,
+      "loss": 1.5929,
+      "step": 5926
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1741957664489746,
+      "learning_rate": 1.4584247076023393e-05,
+      "loss": 1.558,
+      "step": 5927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.601177930831909,
+      "learning_rate": 1.4583333333333333e-05,
+      "loss": 1.5007,
+      "step": 5928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.56325626373291,
+      "learning_rate": 1.4582419590643275e-05,
+      "loss": 1.6765,
+      "step": 5929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6439638137817383,
+      "learning_rate": 1.4581505847953216e-05,
+      "loss": 1.8799,
+      "step": 5930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5316946506500244,
+      "learning_rate": 1.4580592105263158e-05,
+      "loss": 1.5465,
+      "step": 5931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.562526226043701,
+      "learning_rate": 1.45796783625731e-05,
+      "loss": 1.541,
+      "step": 5932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4060449600219727,
+      "learning_rate": 1.4578764619883041e-05,
+      "loss": 1.3695,
+      "step": 5933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3137283325195312,
+      "learning_rate": 1.4577850877192983e-05,
+      "loss": 1.8904,
+      "step": 5934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.596637725830078,
+      "learning_rate": 1.4576937134502924e-05,
+      "loss": 1.5968,
+      "step": 5935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7412476539611816,
+      "learning_rate": 1.4576023391812866e-05,
+      "loss": 1.8319,
+      "step": 5936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6828455924987793,
+      "learning_rate": 1.4575109649122806e-05,
+      "loss": 1.7946,
+      "step": 5937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9733545780181885,
+      "learning_rate": 1.4574195906432749e-05,
+      "loss": 1.7374,
+      "step": 5938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9035613536834717,
+      "learning_rate": 1.457328216374269e-05,
+      "loss": 1.5504,
+      "step": 5939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.842705011367798,
+      "learning_rate": 1.4572368421052631e-05,
+      "loss": 1.7534,
+      "step": 5940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.232166051864624,
+      "learning_rate": 1.4571454678362574e-05,
+      "loss": 2.0523,
+      "step": 5941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9375743865966797,
+      "learning_rate": 1.4570540935672514e-05,
+      "loss": 1.8647,
+      "step": 5942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.963496446609497,
+      "learning_rate": 1.4569627192982458e-05,
+      "loss": 1.4751,
+      "step": 5943
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5632827281951904,
+      "learning_rate": 1.45687134502924e-05,
+      "loss": 1.7754,
+      "step": 5944
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.591334581375122,
+      "learning_rate": 1.456779970760234e-05,
+      "loss": 1.6637,
+      "step": 5945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9550061225891113,
+      "learning_rate": 1.4566885964912283e-05,
+      "loss": 1.4679,
+      "step": 5946
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.41903018951416,
+      "learning_rate": 1.4565972222222225e-05,
+      "loss": 1.3853,
+      "step": 5947
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.824232816696167,
+      "learning_rate": 1.4565058479532166e-05,
+      "loss": 1.8392,
+      "step": 5948
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.858515739440918,
+      "learning_rate": 1.4564144736842108e-05,
+      "loss": 2.0427,
+      "step": 5949
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.93123722076416,
+      "learning_rate": 1.4563230994152049e-05,
+      "loss": 1.9304,
+      "step": 5950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8151824474334717,
+      "learning_rate": 1.456231725146199e-05,
+      "loss": 1.7641,
+      "step": 5951
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9867546558380127,
+      "learning_rate": 1.4561403508771931e-05,
+      "loss": 1.7782,
+      "step": 5952
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4467623233795166,
+      "learning_rate": 1.4560489766081873e-05,
+      "loss": 1.301,
+      "step": 5953
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6052134037017822,
+      "learning_rate": 1.4559576023391816e-05,
+      "loss": 1.2596,
+      "step": 5954
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7298977375030518,
+      "learning_rate": 1.4558662280701756e-05,
+      "loss": 1.9233,
+      "step": 5955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8644521236419678,
+      "learning_rate": 1.4557748538011698e-05,
+      "loss": 1.8348,
+      "step": 5956
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.605602264404297,
+      "learning_rate": 1.4556834795321639e-05,
+      "loss": 1.8011,
+      "step": 5957
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.09980845451355,
+      "learning_rate": 1.4555921052631581e-05,
+      "loss": 1.6779,
+      "step": 5958
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0169687271118164,
+      "learning_rate": 1.4555007309941522e-05,
+      "loss": 1.7431,
+      "step": 5959
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.845794916152954,
+      "learning_rate": 1.4554093567251464e-05,
+      "loss": 1.9797,
+      "step": 5960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.038846969604492,
+      "learning_rate": 1.4553179824561406e-05,
+      "loss": 1.5316,
+      "step": 5961
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4156105518341064,
+      "learning_rate": 1.4552266081871347e-05,
+      "loss": 2.23,
+      "step": 5962
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2726924419403076,
+      "learning_rate": 1.4551352339181289e-05,
+      "loss": 1.9213,
+      "step": 5963
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.842355966567993,
+      "learning_rate": 1.455043859649123e-05,
+      "loss": 1.3573,
+      "step": 5964
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.104804039001465,
+      "learning_rate": 1.4549524853801172e-05,
+      "loss": 1.7771,
+      "step": 5965
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4046783447265625,
+      "learning_rate": 1.4548611111111112e-05,
+      "loss": 1.5829,
+      "step": 5966
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6179778575897217,
+      "learning_rate": 1.4547697368421054e-05,
+      "loss": 1.6343,
+      "step": 5967
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4033713340759277,
+      "learning_rate": 1.4546783625730995e-05,
+      "loss": 1.5523,
+      "step": 5968
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2680158615112305,
+      "learning_rate": 1.4545869883040937e-05,
+      "loss": 2.05,
+      "step": 5969
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0633113384246826,
+      "learning_rate": 1.454495614035088e-05,
+      "loss": 1.858,
+      "step": 5970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.228233814239502,
+      "learning_rate": 1.454404239766082e-05,
+      "loss": 1.2602,
+      "step": 5971
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0465056896209717,
+      "learning_rate": 1.4543128654970762e-05,
+      "loss": 1.6617,
+      "step": 5972
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4226107597351074,
+      "learning_rate": 1.4542214912280703e-05,
+      "loss": 1.4619,
+      "step": 5973
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.628303289413452,
+      "learning_rate": 1.4541301169590645e-05,
+      "loss": 1.5809,
+      "step": 5974
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9745523929595947,
+      "learning_rate": 1.4540387426900585e-05,
+      "loss": 1.5416,
+      "step": 5975
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.289069890975952,
+      "learning_rate": 1.4539473684210528e-05,
+      "loss": 1.2543,
+      "step": 5976
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.044921636581421,
+      "learning_rate": 1.453855994152047e-05,
+      "loss": 1.5654,
+      "step": 5977
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.754159688949585,
+      "learning_rate": 1.453764619883041e-05,
+      "loss": 1.6721,
+      "step": 5978
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.951925754547119,
+      "learning_rate": 1.4536732456140353e-05,
+      "loss": 1.6128,
+      "step": 5979
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8806395530700684,
+      "learning_rate": 1.4535818713450293e-05,
+      "loss": 1.5686,
+      "step": 5980
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0607967376708984,
+      "learning_rate": 1.4534904970760235e-05,
+      "loss": 1.7086,
+      "step": 5981
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.601270914077759,
+      "learning_rate": 1.4533991228070176e-05,
+      "loss": 1.5688,
+      "step": 5982
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.460113286972046,
+      "learning_rate": 1.4533077485380118e-05,
+      "loss": 1.2282,
+      "step": 5983
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8689236640930176,
+      "learning_rate": 1.453216374269006e-05,
+      "loss": 1.735,
+      "step": 5984
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7621796131134033,
+      "learning_rate": 1.453125e-05,
+      "loss": 1.3467,
+      "step": 5985
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.727064609527588,
+      "learning_rate": 1.4530336257309943e-05,
+      "loss": 1.9285,
+      "step": 5986
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.33074951171875,
+      "learning_rate": 1.4529422514619884e-05,
+      "loss": 2.4833,
+      "step": 5987
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8611831665039062,
+      "learning_rate": 1.4528508771929826e-05,
+      "loss": 1.439,
+      "step": 5988
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.814073324203491,
+      "learning_rate": 1.4527595029239766e-05,
+      "loss": 1.5324,
+      "step": 5989
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.436880350112915,
+      "learning_rate": 1.4526681286549709e-05,
+      "loss": 1.9607,
+      "step": 5990
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2651407718658447,
+      "learning_rate": 1.452576754385965e-05,
+      "loss": 1.6363,
+      "step": 5991
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.137441635131836,
+      "learning_rate": 1.4524853801169591e-05,
+      "loss": 1.8637,
+      "step": 5992
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.292184829711914,
+      "learning_rate": 1.4523940058479534e-05,
+      "loss": 2.1784,
+      "step": 5993
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.003828525543213,
+      "learning_rate": 1.4523026315789474e-05,
+      "loss": 1.8675,
+      "step": 5994
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8830151557922363,
+      "learning_rate": 1.4522112573099416e-05,
+      "loss": 1.4225,
+      "step": 5995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2144429683685303,
+      "learning_rate": 1.4521198830409357e-05,
+      "loss": 2.0655,
+      "step": 5996
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6404013633728027,
+      "learning_rate": 1.4520285087719299e-05,
+      "loss": 1.5056,
+      "step": 5997
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.896794557571411,
+      "learning_rate": 1.4519371345029241e-05,
+      "loss": 1.7894,
+      "step": 5998
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.734067440032959,
+      "learning_rate": 1.4518457602339182e-05,
+      "loss": 1.3696,
+      "step": 5999
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8753740787506104,
+      "learning_rate": 1.4517543859649124e-05,
+      "loss": 1.5816,
+      "step": 6000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3809432983398438,
+      "learning_rate": 1.4516630116959065e-05,
+      "loss": 1.6242,
+      "step": 6001
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.277284860610962,
+      "learning_rate": 1.4515716374269007e-05,
+      "loss": 1.6238,
+      "step": 6002
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.962249994277954,
+      "learning_rate": 1.4514802631578947e-05,
+      "loss": 1.6193,
+      "step": 6003
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.94559907913208,
+      "learning_rate": 1.451388888888889e-05,
+      "loss": 1.6335,
+      "step": 6004
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0985333919525146,
+      "learning_rate": 1.4512975146198832e-05,
+      "loss": 1.6586,
+      "step": 6005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4186460971832275,
+      "learning_rate": 1.4512061403508772e-05,
+      "loss": 1.6119,
+      "step": 6006
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.811779260635376,
+      "learning_rate": 1.4511147660818714e-05,
+      "loss": 1.6353,
+      "step": 6007
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.893935441970825,
+      "learning_rate": 1.4510233918128655e-05,
+      "loss": 1.5264,
+      "step": 6008
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4571926593780518,
+      "learning_rate": 1.4509320175438597e-05,
+      "loss": 1.4172,
+      "step": 6009
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.875267744064331,
+      "learning_rate": 1.4508406432748538e-05,
+      "loss": 1.4991,
+      "step": 6010
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.602731227874756,
+      "learning_rate": 1.450749269005848e-05,
+      "loss": 1.5279,
+      "step": 6011
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.812995672225952,
+      "learning_rate": 1.4506578947368422e-05,
+      "loss": 1.7555,
+      "step": 6012
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.777416944503784,
+      "learning_rate": 1.4505665204678363e-05,
+      "loss": 1.4379,
+      "step": 6013
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2011992931365967,
+      "learning_rate": 1.4504751461988305e-05,
+      "loss": 1.7215,
+      "step": 6014
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9363794326782227,
+      "learning_rate": 1.4503837719298245e-05,
+      "loss": 1.4667,
+      "step": 6015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.464723825454712,
+      "learning_rate": 1.4502923976608188e-05,
+      "loss": 1.3249,
+      "step": 6016
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7844278812408447,
+      "learning_rate": 1.4502010233918128e-05,
+      "loss": 1.5672,
+      "step": 6017
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7188074588775635,
+      "learning_rate": 1.450109649122807e-05,
+      "loss": 1.46,
+      "step": 6018
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1549253463745117,
+      "learning_rate": 1.4500182748538013e-05,
+      "loss": 1.3217,
+      "step": 6019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.225611686706543,
+      "learning_rate": 1.4499269005847953e-05,
+      "loss": 1.2116,
+      "step": 6020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.614959478378296,
+      "learning_rate": 1.4498355263157895e-05,
+      "loss": 1.5696,
+      "step": 6021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.973224639892578,
+      "learning_rate": 1.4497441520467836e-05,
+      "loss": 1.5869,
+      "step": 6022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8882744312286377,
+      "learning_rate": 1.4496527777777778e-05,
+      "loss": 1.5942,
+      "step": 6023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9685745239257812,
+      "learning_rate": 1.4495614035087719e-05,
+      "loss": 1.5587,
+      "step": 6024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.854285478591919,
+      "learning_rate": 1.4494700292397661e-05,
+      "loss": 1.8004,
+      "step": 6025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7747089862823486,
+      "learning_rate": 1.4493786549707603e-05,
+      "loss": 1.4959,
+      "step": 6026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.67633318901062,
+      "learning_rate": 1.4492872807017544e-05,
+      "loss": 1.6821,
+      "step": 6027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.545884847640991,
+      "learning_rate": 1.4491959064327488e-05,
+      "loss": 1.3398,
+      "step": 6028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1791863441467285,
+      "learning_rate": 1.449104532163743e-05,
+      "loss": 1.6208,
+      "step": 6029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9377827644348145,
+      "learning_rate": 1.449013157894737e-05,
+      "loss": 1.5092,
+      "step": 6030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.650735855102539,
+      "learning_rate": 1.4489217836257313e-05,
+      "loss": 1.4594,
+      "step": 6031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.280400514602661,
+      "learning_rate": 1.4488304093567253e-05,
+      "loss": 2.0525,
+      "step": 6032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.853447437286377,
+      "learning_rate": 1.4487390350877195e-05,
+      "loss": 1.8238,
+      "step": 6033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.580256938934326,
+      "learning_rate": 1.4486476608187136e-05,
+      "loss": 1.567,
+      "step": 6034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9518887996673584,
+      "learning_rate": 1.4485562865497078e-05,
+      "loss": 1.6064,
+      "step": 6035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0347700119018555,
+      "learning_rate": 1.448464912280702e-05,
+      "loss": 1.8126,
+      "step": 6036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.175438404083252,
+      "learning_rate": 1.448373538011696e-05,
+      "loss": 1.7076,
+      "step": 6037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.75510311126709,
+      "learning_rate": 1.4482821637426903e-05,
+      "loss": 1.4805,
+      "step": 6038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2249863147735596,
+      "learning_rate": 1.4481907894736844e-05,
+      "loss": 1.4201,
+      "step": 6039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5971338748931885,
+      "learning_rate": 1.4480994152046786e-05,
+      "loss": 1.327,
+      "step": 6040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5815749168395996,
+      "learning_rate": 1.4480080409356726e-05,
+      "loss": 1.3324,
+      "step": 6041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5038607120513916,
+      "learning_rate": 1.4479166666666669e-05,
+      "loss": 2.0612,
+      "step": 6042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.232444763183594,
+      "learning_rate": 1.447825292397661e-05,
+      "loss": 2.1582,
+      "step": 6043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9119057655334473,
+      "learning_rate": 1.4477339181286551e-05,
+      "loss": 1.8417,
+      "step": 6044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6334590911865234,
+      "learning_rate": 1.4476425438596493e-05,
+      "loss": 1.6256,
+      "step": 6045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7140681743621826,
+      "learning_rate": 1.4475511695906434e-05,
+      "loss": 1.2708,
+      "step": 6046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.717118501663208,
+      "learning_rate": 1.4474597953216376e-05,
+      "loss": 1.2563,
+      "step": 6047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3865554332733154,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 1.6027,
+      "step": 6048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7314512729644775,
+      "learning_rate": 1.4472770467836259e-05,
+      "loss": 1.6932,
+      "step": 6049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.202327013015747,
+      "learning_rate": 1.4471856725146201e-05,
+      "loss": 1.9092,
+      "step": 6050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.700655460357666,
+      "learning_rate": 1.4470942982456142e-05,
+      "loss": 1.4996,
+      "step": 6051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.284148693084717,
+      "learning_rate": 1.4470029239766084e-05,
+      "loss": 1.8311,
+      "step": 6052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0791690349578857,
+      "learning_rate": 1.4469115497076024e-05,
+      "loss": 1.7191,
+      "step": 6053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.03112530708313,
+      "learning_rate": 1.4468201754385967e-05,
+      "loss": 1.6545,
+      "step": 6054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.954132080078125,
+      "learning_rate": 1.4467288011695907e-05,
+      "loss": 1.5525,
+      "step": 6055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.555636167526245,
+      "learning_rate": 1.446637426900585e-05,
+      "loss": 1.3863,
+      "step": 6056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.953294277191162,
+      "learning_rate": 1.4465460526315792e-05,
+      "loss": 1.5189,
+      "step": 6057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6616337299346924,
+      "learning_rate": 1.4464546783625732e-05,
+      "loss": 1.577,
+      "step": 6058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5735652446746826,
+      "learning_rate": 1.4463633040935674e-05,
+      "loss": 1.4082,
+      "step": 6059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7695353031158447,
+      "learning_rate": 1.4462719298245615e-05,
+      "loss": 1.3785,
+      "step": 6060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6942732334136963,
+      "learning_rate": 1.4461805555555557e-05,
+      "loss": 1.4205,
+      "step": 6061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.597210168838501,
+      "learning_rate": 1.4460891812865498e-05,
+      "loss": 2.1424,
+      "step": 6062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.006986141204834,
+      "learning_rate": 1.445997807017544e-05,
+      "loss": 1.8443,
+      "step": 6063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.143012046813965,
+      "learning_rate": 1.4459064327485382e-05,
+      "loss": 1.6767,
+      "step": 6064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.347276210784912,
+      "learning_rate": 1.4458150584795323e-05,
+      "loss": 1.2332,
+      "step": 6065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.49528169631958,
+      "learning_rate": 1.4457236842105265e-05,
+      "loss": 1.4331,
+      "step": 6066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4634358882904053,
+      "learning_rate": 1.4456323099415205e-05,
+      "loss": 1.2647,
+      "step": 6067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2653651237487793,
+      "learning_rate": 1.4455409356725148e-05,
+      "loss": 1.8234,
+      "step": 6068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.792555093765259,
+      "learning_rate": 1.4454495614035088e-05,
+      "loss": 1.717,
+      "step": 6069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4230151176452637,
+      "learning_rate": 1.445358187134503e-05,
+      "loss": 1.7814,
+      "step": 6070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4929087162017822,
+      "learning_rate": 1.4452668128654973e-05,
+      "loss": 1.3382,
+      "step": 6071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6424014568328857,
+      "learning_rate": 1.4451754385964913e-05,
+      "loss": 1.4335,
+      "step": 6072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6878256797790527,
+      "learning_rate": 1.4450840643274855e-05,
+      "loss": 1.5616,
+      "step": 6073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5574560165405273,
+      "learning_rate": 1.4449926900584796e-05,
+      "loss": 1.9099,
+      "step": 6074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8690943717956543,
+      "learning_rate": 1.4449013157894738e-05,
+      "loss": 1.5792,
+      "step": 6075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6598665714263916,
+      "learning_rate": 1.4448099415204679e-05,
+      "loss": 1.6472,
+      "step": 6076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.498290777206421,
+      "learning_rate": 1.444718567251462e-05,
+      "loss": 1.9826,
+      "step": 6077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4373931884765625,
+      "learning_rate": 1.4446271929824563e-05,
+      "loss": 1.4977,
+      "step": 6078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.777115821838379,
+      "learning_rate": 1.4445358187134504e-05,
+      "loss": 1.6776,
+      "step": 6079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6561052799224854,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.4504,
+      "step": 6080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3166825771331787,
+      "learning_rate": 1.4443530701754386e-05,
+      "loss": 1.8892,
+      "step": 6081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3527822494506836,
+      "learning_rate": 1.4442616959064329e-05,
+      "loss": 1.3483,
+      "step": 6082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5714104175567627,
+      "learning_rate": 1.4441703216374269e-05,
+      "loss": 1.4855,
+      "step": 6083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.776005268096924,
+      "learning_rate": 1.4440789473684211e-05,
+      "loss": 1.5392,
+      "step": 6084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3701331615448,
+      "learning_rate": 1.4439875730994154e-05,
+      "loss": 2.062,
+      "step": 6085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2470614910125732,
+      "learning_rate": 1.4438961988304094e-05,
+      "loss": 1.5034,
+      "step": 6086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.414957046508789,
+      "learning_rate": 1.4438048245614036e-05,
+      "loss": 1.827,
+      "step": 6087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9345970153808594,
+      "learning_rate": 1.4437134502923977e-05,
+      "loss": 1.5059,
+      "step": 6088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8616392612457275,
+      "learning_rate": 1.4436220760233919e-05,
+      "loss": 1.8729,
+      "step": 6089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.395181655883789,
+      "learning_rate": 1.443530701754386e-05,
+      "loss": 1.4719,
+      "step": 6090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.423164129257202,
+      "learning_rate": 1.4434393274853802e-05,
+      "loss": 1.2675,
+      "step": 6091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8860466480255127,
+      "learning_rate": 1.4433479532163744e-05,
+      "loss": 1.7609,
+      "step": 6092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.072852849960327,
+      "learning_rate": 1.4432565789473685e-05,
+      "loss": 1.8348,
+      "step": 6093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.677781343460083,
+      "learning_rate": 1.4431652046783627e-05,
+      "loss": 1.9003,
+      "step": 6094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9101622104644775,
+      "learning_rate": 1.4430738304093567e-05,
+      "loss": 1.6382,
+      "step": 6095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1903140544891357,
+      "learning_rate": 1.442982456140351e-05,
+      "loss": 1.914,
+      "step": 6096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1604957580566406,
+      "learning_rate": 1.442891081871345e-05,
+      "loss": 1.521,
+      "step": 6097
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9826254844665527,
+      "learning_rate": 1.4427997076023392e-05,
+      "loss": 1.4175,
+      "step": 6098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.696336030960083,
+      "learning_rate": 1.4427083333333334e-05,
+      "loss": 1.6168,
+      "step": 6099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8699300289154053,
+      "learning_rate": 1.4426169590643275e-05,
+      "loss": 1.6144,
+      "step": 6100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9071407318115234,
+      "learning_rate": 1.4425255847953217e-05,
+      "loss": 1.8785,
+      "step": 6101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.405797004699707,
+      "learning_rate": 1.4424342105263158e-05,
+      "loss": 1.9375,
+      "step": 6102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9757766723632812,
+      "learning_rate": 1.44234283625731e-05,
+      "loss": 1.5948,
+      "step": 6103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.543975591659546,
+      "learning_rate": 1.442251461988304e-05,
+      "loss": 1.3765,
+      "step": 6104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.926697015762329,
+      "learning_rate": 1.4421600877192983e-05,
+      "loss": 1.7181,
+      "step": 6105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1665608882904053,
+      "learning_rate": 1.4420687134502925e-05,
+      "loss": 1.9124,
+      "step": 6106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7936298847198486,
+      "learning_rate": 1.4419773391812865e-05,
+      "loss": 1.6656,
+      "step": 6107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4678008556365967,
+      "learning_rate": 1.4418859649122808e-05,
+      "loss": 1.5002,
+      "step": 6108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4992010593414307,
+      "learning_rate": 1.4417945906432748e-05,
+      "loss": 1.2862,
+      "step": 6109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3025248050689697,
+      "learning_rate": 1.441703216374269e-05,
+      "loss": 1.2543,
+      "step": 6110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8868491649627686,
+      "learning_rate": 1.4416118421052631e-05,
+      "loss": 1.5528,
+      "step": 6111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9319119453430176,
+      "learning_rate": 1.4415204678362573e-05,
+      "loss": 1.806,
+      "step": 6112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5528435707092285,
+      "learning_rate": 1.4414290935672515e-05,
+      "loss": 1.3579,
+      "step": 6113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6291682720184326,
+      "learning_rate": 1.4413377192982458e-05,
+      "loss": 1.6342,
+      "step": 6114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.955476999282837,
+      "learning_rate": 1.44124634502924e-05,
+      "loss": 1.5608,
+      "step": 6115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.882033586502075,
+      "learning_rate": 1.4411549707602342e-05,
+      "loss": 1.4911,
+      "step": 6116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.116405963897705,
+      "learning_rate": 1.4410635964912283e-05,
+      "loss": 1.6397,
+      "step": 6117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.51706600189209,
+      "learning_rate": 1.4409722222222225e-05,
+      "loss": 2.2471,
+      "step": 6118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.496499538421631,
+      "learning_rate": 1.4408808479532165e-05,
+      "loss": 2.0941,
+      "step": 6119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.858982801437378,
+      "learning_rate": 1.4407894736842108e-05,
+      "loss": 1.4298,
+      "step": 6120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4623496532440186,
+      "learning_rate": 1.4406980994152048e-05,
+      "loss": 1.5986,
+      "step": 6121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.199610710144043,
+      "learning_rate": 1.440606725146199e-05,
+      "loss": 1.8813,
+      "step": 6122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9358315467834473,
+      "learning_rate": 1.4405153508771933e-05,
+      "loss": 1.5661,
+      "step": 6123
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7667691707611084,
+      "learning_rate": 1.4404239766081873e-05,
+      "loss": 1.5106,
+      "step": 6124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7642405033111572,
+      "learning_rate": 1.4403326023391815e-05,
+      "loss": 1.5646,
+      "step": 6125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.529820680618286,
+      "learning_rate": 1.4402412280701756e-05,
+      "loss": 1.4412,
+      "step": 6126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4430389404296875,
+      "learning_rate": 1.4401498538011698e-05,
+      "loss": 1.4948,
+      "step": 6127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.290400505065918,
+      "learning_rate": 1.4400584795321639e-05,
+      "loss": 1.1979,
+      "step": 6128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.400066375732422,
+      "learning_rate": 1.439967105263158e-05,
+      "loss": 1.4189,
+      "step": 6129
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1012489795684814,
+      "learning_rate": 1.4398757309941523e-05,
+      "loss": 1.6404,
+      "step": 6130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9976539611816406,
+      "learning_rate": 1.4397843567251464e-05,
+      "loss": 1.7242,
+      "step": 6131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7076661586761475,
+      "learning_rate": 1.4396929824561406e-05,
+      "loss": 1.9688,
+      "step": 6132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.882317304611206,
+      "learning_rate": 1.4396016081871346e-05,
+      "loss": 1.7758,
+      "step": 6133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.545844316482544,
+      "learning_rate": 1.4395102339181289e-05,
+      "loss": 1.3529,
+      "step": 6134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.106701135635376,
+      "learning_rate": 1.4394188596491229e-05,
+      "loss": 1.4409,
+      "step": 6135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.385845184326172,
+      "learning_rate": 1.4393274853801171e-05,
+      "loss": 1.417,
+      "step": 6136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4549081325531006,
+      "learning_rate": 1.4392361111111113e-05,
+      "loss": 1.4044,
+      "step": 6137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9485924243927,
+      "learning_rate": 1.4391447368421054e-05,
+      "loss": 1.5995,
+      "step": 6138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0309839248657227,
+      "learning_rate": 1.4390533625730996e-05,
+      "loss": 1.6783,
+      "step": 6139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.50504994392395,
+      "learning_rate": 1.4389619883040937e-05,
+      "loss": 1.2882,
+      "step": 6140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.2589802742004395,
+      "learning_rate": 1.4388706140350879e-05,
+      "loss": 1.6372,
+      "step": 6141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4645192623138428,
+      "learning_rate": 1.438779239766082e-05,
+      "loss": 1.2425,
+      "step": 6142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.369917869567871,
+      "learning_rate": 1.4386878654970762e-05,
+      "loss": 1.8155,
+      "step": 6143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.921518325805664,
+      "learning_rate": 1.4385964912280704e-05,
+      "loss": 1.6212,
+      "step": 6144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8581175804138184,
+      "learning_rate": 1.4385051169590644e-05,
+      "loss": 1.7284,
+      "step": 6145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8997042179107666,
+      "learning_rate": 1.4384137426900587e-05,
+      "loss": 1.6579,
+      "step": 6146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.554689645767212,
+      "learning_rate": 1.4383223684210527e-05,
+      "loss": 1.6084,
+      "step": 6147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.045029878616333,
+      "learning_rate": 1.438230994152047e-05,
+      "loss": 2.1566,
+      "step": 6148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9879472255706787,
+      "learning_rate": 1.438139619883041e-05,
+      "loss": 2.0883,
+      "step": 6149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.308776378631592,
+      "learning_rate": 1.4380482456140352e-05,
+      "loss": 2.2584,
+      "step": 6150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6938955783843994,
+      "learning_rate": 1.4379568713450294e-05,
+      "loss": 1.5153,
+      "step": 6151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.147540807723999,
+      "learning_rate": 1.4378654970760235e-05,
+      "loss": 1.8002,
+      "step": 6152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0560734272003174,
+      "learning_rate": 1.4377741228070177e-05,
+      "loss": 2.0778,
+      "step": 6153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.103811740875244,
+      "learning_rate": 1.4376827485380118e-05,
+      "loss": 1.529,
+      "step": 6154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.067946195602417,
+      "learning_rate": 1.437591374269006e-05,
+      "loss": 1.6686,
+      "step": 6155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1082730293273926,
+      "learning_rate": 1.4375e-05,
+      "loss": 1.7862,
+      "step": 6156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5992584228515625,
+      "learning_rate": 1.4374086257309943e-05,
+      "loss": 1.2669,
+      "step": 6157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.017165422439575,
+      "learning_rate": 1.4373172514619885e-05,
+      "loss": 1.7217,
+      "step": 6158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9711694717407227,
+      "learning_rate": 1.4372258771929825e-05,
+      "loss": 1.3297,
+      "step": 6159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0650362968444824,
+      "learning_rate": 1.4371345029239768e-05,
+      "loss": 1.8762,
+      "step": 6160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.746899366378784,
+      "learning_rate": 1.4370431286549708e-05,
+      "loss": 1.7231,
+      "step": 6161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.722261905670166,
+      "learning_rate": 1.436951754385965e-05,
+      "loss": 1.7369,
+      "step": 6162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2638590335845947,
+      "learning_rate": 1.4368603801169591e-05,
+      "loss": 1.2856,
+      "step": 6163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7845165729522705,
+      "learning_rate": 1.4367690058479533e-05,
+      "loss": 1.5846,
+      "step": 6164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.336266040802002,
+      "learning_rate": 1.4366776315789475e-05,
+      "loss": 1.523,
+      "step": 6165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.605271339416504,
+      "learning_rate": 1.4365862573099416e-05,
+      "loss": 1.5469,
+      "step": 6166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7573788166046143,
+      "learning_rate": 1.4364948830409358e-05,
+      "loss": 1.6051,
+      "step": 6167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0555062294006348,
+      "learning_rate": 1.4364035087719299e-05,
+      "loss": 1.8063,
+      "step": 6168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.297200918197632,
+      "learning_rate": 1.436312134502924e-05,
+      "loss": 1.7631,
+      "step": 6169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7880468368530273,
+      "learning_rate": 1.4362207602339181e-05,
+      "loss": 1.8164,
+      "step": 6170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8410394191741943,
+      "learning_rate": 1.4361293859649124e-05,
+      "loss": 1.7486,
+      "step": 6171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.774146795272827,
+      "learning_rate": 1.4360380116959066e-05,
+      "loss": 1.8644,
+      "step": 6172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7902204990386963,
+      "learning_rate": 1.4359466374269006e-05,
+      "loss": 1.5891,
+      "step": 6173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.870190143585205,
+      "learning_rate": 1.4358552631578949e-05,
+      "loss": 1.667,
+      "step": 6174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9788804054260254,
+      "learning_rate": 1.4357638888888889e-05,
+      "loss": 1.7728,
+      "step": 6175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.69157075881958,
+      "learning_rate": 1.4356725146198831e-05,
+      "loss": 1.4934,
+      "step": 6176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6537303924560547,
+      "learning_rate": 1.4355811403508772e-05,
+      "loss": 1.5956,
+      "step": 6177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.955042600631714,
+      "learning_rate": 1.4354897660818714e-05,
+      "loss": 1.7694,
+      "step": 6178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1274540424346924,
+      "learning_rate": 1.4353983918128656e-05,
+      "loss": 1.4973,
+      "step": 6179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1963765621185303,
+      "learning_rate": 1.4353070175438597e-05,
+      "loss": 1.5578,
+      "step": 6180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.199460983276367,
+      "learning_rate": 1.4352156432748539e-05,
+      "loss": 1.7962,
+      "step": 6181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7636122703552246,
+      "learning_rate": 1.435124269005848e-05,
+      "loss": 1.6701,
+      "step": 6182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.589531898498535,
+      "learning_rate": 1.4350328947368422e-05,
+      "loss": 1.5956,
+      "step": 6183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7828214168548584,
+      "learning_rate": 1.4349415204678362e-05,
+      "loss": 1.5268,
+      "step": 6184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.074368953704834,
+      "learning_rate": 1.4348501461988305e-05,
+      "loss": 1.8976,
+      "step": 6185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7741315364837646,
+      "learning_rate": 1.4347587719298247e-05,
+      "loss": 1.7816,
+      "step": 6186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.332521438598633,
+      "learning_rate": 1.4346673976608187e-05,
+      "loss": 1.6056,
+      "step": 6187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1962485313415527,
+      "learning_rate": 1.434576023391813e-05,
+      "loss": 1.7875,
+      "step": 6188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.439981698989868,
+      "learning_rate": 1.434484649122807e-05,
+      "loss": 1.5621,
+      "step": 6189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.966439723968506,
+      "learning_rate": 1.4343932748538012e-05,
+      "loss": 1.7692,
+      "step": 6190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1327109336853027,
+      "learning_rate": 1.4343019005847953e-05,
+      "loss": 1.9323,
+      "step": 6191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1701467037200928,
+      "learning_rate": 1.4342105263157895e-05,
+      "loss": 1.9427,
+      "step": 6192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.611905813217163,
+      "learning_rate": 1.4341191520467837e-05,
+      "loss": 1.6186,
+      "step": 6193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2597804069519043,
+      "learning_rate": 1.4340277777777778e-05,
+      "loss": 1.269,
+      "step": 6194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.189011573791504,
+      "learning_rate": 1.433936403508772e-05,
+      "loss": 1.8143,
+      "step": 6195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0528297424316406,
+      "learning_rate": 1.433845029239766e-05,
+      "loss": 1.5915,
+      "step": 6196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5809154510498047,
+      "learning_rate": 1.4337536549707603e-05,
+      "loss": 1.4749,
+      "step": 6197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9437763690948486,
+      "learning_rate": 1.4336622807017543e-05,
+      "loss": 1.6645,
+      "step": 6198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4407122135162354,
+      "learning_rate": 1.4335709064327487e-05,
+      "loss": 1.5147,
+      "step": 6199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.334672212600708,
+      "learning_rate": 1.433479532163743e-05,
+      "loss": 1.5006,
+      "step": 6200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9288976192474365,
+      "learning_rate": 1.433388157894737e-05,
+      "loss": 1.8388,
+      "step": 6201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5579822063446045,
+      "learning_rate": 1.4332967836257312e-05,
+      "loss": 1.7602,
+      "step": 6202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.16534161567688,
+      "learning_rate": 1.4332054093567254e-05,
+      "loss": 1.8644,
+      "step": 6203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9347972869873047,
+      "learning_rate": 1.4331140350877195e-05,
+      "loss": 1.4789,
+      "step": 6204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0288748741149902,
+      "learning_rate": 1.4330226608187137e-05,
+      "loss": 1.4861,
+      "step": 6205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1680400371551514,
+      "learning_rate": 1.4329312865497078e-05,
+      "loss": 2.132,
+      "step": 6206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7086799144744873,
+      "learning_rate": 1.432839912280702e-05,
+      "loss": 1.4771,
+      "step": 6207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.699035406112671,
+      "learning_rate": 1.432748538011696e-05,
+      "loss": 1.5576,
+      "step": 6208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3820676803588867,
+      "learning_rate": 1.4326571637426903e-05,
+      "loss": 2.1701,
+      "step": 6209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.992457628250122,
+      "learning_rate": 1.4325657894736845e-05,
+      "loss": 1.5041,
+      "step": 6210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4689714908599854,
+      "learning_rate": 1.4324744152046785e-05,
+      "loss": 1.4185,
+      "step": 6211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8401341438293457,
+      "learning_rate": 1.4323830409356728e-05,
+      "loss": 1.7034,
+      "step": 6212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4897897243499756,
+      "learning_rate": 1.4322916666666668e-05,
+      "loss": 1.3238,
+      "step": 6213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8985607624053955,
+      "learning_rate": 1.432200292397661e-05,
+      "loss": 1.5612,
+      "step": 6214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5364670753479004,
+      "learning_rate": 1.432108918128655e-05,
+      "loss": 1.5174,
+      "step": 6215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1704413890838623,
+      "learning_rate": 1.4320175438596493e-05,
+      "loss": 1.6037,
+      "step": 6216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.81911563873291,
+      "learning_rate": 1.4319261695906435e-05,
+      "loss": 1.7474,
+      "step": 6217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8527755737304688,
+      "learning_rate": 1.4318347953216376e-05,
+      "loss": 1.6773,
+      "step": 6218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3698208332061768,
+      "learning_rate": 1.4317434210526318e-05,
+      "loss": 1.5732,
+      "step": 6219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9983980655670166,
+      "learning_rate": 1.4316520467836259e-05,
+      "loss": 1.2354,
+      "step": 6220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0781142711639404,
+      "learning_rate": 1.43156067251462e-05,
+      "loss": 1.5881,
+      "step": 6221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9810991287231445,
+      "learning_rate": 1.4314692982456141e-05,
+      "loss": 1.6407,
+      "step": 6222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6579577922821045,
+      "learning_rate": 1.4313779239766084e-05,
+      "loss": 1.3733,
+      "step": 6223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.719204902648926,
+      "learning_rate": 1.4312865497076026e-05,
+      "loss": 1.5449,
+      "step": 6224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0873770713806152,
+      "learning_rate": 1.4311951754385966e-05,
+      "loss": 2.0945,
+      "step": 6225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.486833333969116,
+      "learning_rate": 1.4311038011695909e-05,
+      "loss": 1.5778,
+      "step": 6226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.988175630569458,
+      "learning_rate": 1.4310124269005849e-05,
+      "loss": 1.5642,
+      "step": 6227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.852717161178589,
+      "learning_rate": 1.4309210526315791e-05,
+      "loss": 1.5763,
+      "step": 6228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4314281940460205,
+      "learning_rate": 1.4308296783625732e-05,
+      "loss": 1.3278,
+      "step": 6229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5612239837646484,
+      "learning_rate": 1.4307383040935674e-05,
+      "loss": 1.2558,
+      "step": 6230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.106841564178467,
+      "learning_rate": 1.4306469298245616e-05,
+      "loss": 1.8211,
+      "step": 6231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.991957426071167,
+      "learning_rate": 1.4305555555555557e-05,
+      "loss": 2.0015,
+      "step": 6232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9597606658935547,
+      "learning_rate": 1.4304641812865499e-05,
+      "loss": 1.8334,
+      "step": 6233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.331620931625366,
+      "learning_rate": 1.430372807017544e-05,
+      "loss": 1.8079,
+      "step": 6234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.801621675491333,
+      "learning_rate": 1.4302814327485382e-05,
+      "loss": 1.5083,
+      "step": 6235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8606245517730713,
+      "learning_rate": 1.4301900584795322e-05,
+      "loss": 1.7705,
+      "step": 6236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9378561973571777,
+      "learning_rate": 1.4300986842105264e-05,
+      "loss": 1.7605,
+      "step": 6237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.927901029586792,
+      "learning_rate": 1.4300073099415207e-05,
+      "loss": 1.6878,
+      "step": 6238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3448123931884766,
+      "learning_rate": 1.4299159356725147e-05,
+      "loss": 1.3195,
+      "step": 6239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7057251930236816,
+      "learning_rate": 1.429824561403509e-05,
+      "loss": 1.7925,
+      "step": 6240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4283459186553955,
+      "learning_rate": 1.429733187134503e-05,
+      "loss": 1.4831,
+      "step": 6241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.515089988708496,
+      "learning_rate": 1.4296418128654972e-05,
+      "loss": 1.4249,
+      "step": 6242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7757232189178467,
+      "learning_rate": 1.4295504385964913e-05,
+      "loss": 1.4268,
+      "step": 6243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.777630090713501,
+      "learning_rate": 1.4294590643274855e-05,
+      "loss": 1.4604,
+      "step": 6244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5960140228271484,
+      "learning_rate": 1.4293676900584797e-05,
+      "loss": 1.5877,
+      "step": 6245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.912156105041504,
+      "learning_rate": 1.4292763157894738e-05,
+      "loss": 1.4558,
+      "step": 6246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9040043354034424,
+      "learning_rate": 1.429184941520468e-05,
+      "loss": 1.538,
+      "step": 6247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7824459075927734,
+      "learning_rate": 1.429093567251462e-05,
+      "loss": 1.4646,
+      "step": 6248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.643807888031006,
+      "learning_rate": 1.4290021929824563e-05,
+      "loss": 1.5421,
+      "step": 6249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.978203773498535,
+      "learning_rate": 1.4289108187134503e-05,
+      "loss": 1.7401,
+      "step": 6250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.084531307220459,
+      "learning_rate": 1.4288194444444445e-05,
+      "loss": 1.5192,
+      "step": 6251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8327763080596924,
+      "learning_rate": 1.4287280701754388e-05,
+      "loss": 1.5474,
+      "step": 6252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0905067920684814,
+      "learning_rate": 1.4286366959064328e-05,
+      "loss": 1.8948,
+      "step": 6253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.688356399536133,
+      "learning_rate": 1.428545321637427e-05,
+      "loss": 1.3828,
+      "step": 6254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8284194469451904,
+      "learning_rate": 1.4284539473684211e-05,
+      "loss": 1.4189,
+      "step": 6255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7264740467071533,
+      "learning_rate": 1.4283625730994153e-05,
+      "loss": 1.9523,
+      "step": 6256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.709906816482544,
+      "learning_rate": 1.4282711988304094e-05,
+      "loss": 1.6248,
+      "step": 6257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4344425201416016,
+      "learning_rate": 1.4281798245614036e-05,
+      "loss": 1.2876,
+      "step": 6258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.363780975341797,
+      "learning_rate": 1.4280884502923978e-05,
+      "loss": 1.4108,
+      "step": 6259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5051777362823486,
+      "learning_rate": 1.4279970760233919e-05,
+      "loss": 1.4877,
+      "step": 6260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.437601089477539,
+      "learning_rate": 1.427905701754386e-05,
+      "loss": 1.4133,
+      "step": 6261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7204208374023438,
+      "learning_rate": 1.4278143274853801e-05,
+      "loss": 1.8476,
+      "step": 6262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8706953525543213,
+      "learning_rate": 1.4277229532163744e-05,
+      "loss": 1.3222,
+      "step": 6263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.655426025390625,
+      "learning_rate": 1.4276315789473684e-05,
+      "loss": 1.9697,
+      "step": 6264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9867918491363525,
+      "learning_rate": 1.4275402046783626e-05,
+      "loss": 1.8523,
+      "step": 6265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8634555339813232,
+      "learning_rate": 1.4274488304093569e-05,
+      "loss": 1.6837,
+      "step": 6266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.730639696121216,
+      "learning_rate": 1.4273574561403509e-05,
+      "loss": 1.4722,
+      "step": 6267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.201263427734375,
+      "learning_rate": 1.4272660818713451e-05,
+      "loss": 1.8124,
+      "step": 6268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.52866530418396,
+      "learning_rate": 1.4271747076023392e-05,
+      "loss": 1.5231,
+      "step": 6269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7556471824645996,
+      "learning_rate": 1.4270833333333334e-05,
+      "loss": 1.5869,
+      "step": 6270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3915393352508545,
+      "learning_rate": 1.4269919590643275e-05,
+      "loss": 1.356,
+      "step": 6271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.185034990310669,
+      "learning_rate": 1.4269005847953217e-05,
+      "loss": 1.6129,
+      "step": 6272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.297053575515747,
+      "learning_rate": 1.4268092105263159e-05,
+      "loss": 1.816,
+      "step": 6273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3084418773651123,
+      "learning_rate": 1.42671783625731e-05,
+      "loss": 1.2325,
+      "step": 6274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.014009475708008,
+      "learning_rate": 1.4266264619883042e-05,
+      "loss": 1.6592,
+      "step": 6275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.906694173812866,
+      "learning_rate": 1.4265350877192982e-05,
+      "loss": 1.98,
+      "step": 6276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6331393718719482,
+      "learning_rate": 1.4264437134502925e-05,
+      "loss": 1.5834,
+      "step": 6277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.902911424636841,
+      "learning_rate": 1.4263523391812865e-05,
+      "loss": 1.9213,
+      "step": 6278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.030024528503418,
+      "learning_rate": 1.4262609649122807e-05,
+      "loss": 2.3887,
+      "step": 6279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5229263305664062,
+      "learning_rate": 1.426169590643275e-05,
+      "loss": 1.3701,
+      "step": 6280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5832571983337402,
+      "learning_rate": 1.426078216374269e-05,
+      "loss": 1.6076,
+      "step": 6281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.783708095550537,
+      "learning_rate": 1.4259868421052632e-05,
+      "loss": 1.8452,
+      "step": 6282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.915347099304199,
+      "learning_rate": 1.4258954678362573e-05,
+      "loss": 1.9042,
+      "step": 6283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.556532859802246,
+      "learning_rate": 1.4258040935672515e-05,
+      "loss": 1.7123,
+      "step": 6284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6039116382598877,
+      "learning_rate": 1.4257127192982459e-05,
+      "loss": 1.5339,
+      "step": 6285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5599732398986816,
+      "learning_rate": 1.42562134502924e-05,
+      "loss": 1.703,
+      "step": 6286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.248213529586792,
+      "learning_rate": 1.4255299707602342e-05,
+      "loss": 1.258,
+      "step": 6287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.913771867752075,
+      "learning_rate": 1.4254385964912282e-05,
+      "loss": 1.7321,
+      "step": 6288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.897608518600464,
+      "learning_rate": 1.4253472222222224e-05,
+      "loss": 1.6981,
+      "step": 6289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5797924995422363,
+      "learning_rate": 1.4252558479532167e-05,
+      "loss": 1.4338,
+      "step": 6290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5122976303100586,
+      "learning_rate": 1.4251644736842107e-05,
+      "loss": 1.674,
+      "step": 6291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6975467205047607,
+      "learning_rate": 1.425073099415205e-05,
+      "loss": 1.6067,
+      "step": 6292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8117334842681885,
+      "learning_rate": 1.424981725146199e-05,
+      "loss": 1.6097,
+      "step": 6293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7704339027404785,
+      "learning_rate": 1.4248903508771932e-05,
+      "loss": 1.6148,
+      "step": 6294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.768749475479126,
+      "learning_rate": 1.4247989766081873e-05,
+      "loss": 1.8251,
+      "step": 6295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.179579019546509,
+      "learning_rate": 1.4247076023391815e-05,
+      "loss": 1.1289,
+      "step": 6296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.965731382369995,
+      "learning_rate": 1.4246162280701757e-05,
+      "loss": 1.7598,
+      "step": 6297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.872081756591797,
+      "learning_rate": 1.4245248538011698e-05,
+      "loss": 1.7588,
+      "step": 6298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8844141960144043,
+      "learning_rate": 1.424433479532164e-05,
+      "loss": 1.3903,
+      "step": 6299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.450079917907715,
+      "learning_rate": 1.424342105263158e-05,
+      "loss": 1.4314,
+      "step": 6300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.065153121948242,
+      "learning_rate": 1.4242507309941523e-05,
+      "loss": 1.5226,
+      "step": 6301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7822868824005127,
+      "learning_rate": 1.4241593567251463e-05,
+      "loss": 1.4696,
+      "step": 6302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.513371229171753,
+      "learning_rate": 1.4240679824561405e-05,
+      "loss": 1.8084,
+      "step": 6303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.778446674346924,
+      "learning_rate": 1.4239766081871348e-05,
+      "loss": 1.4144,
+      "step": 6304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.947099208831787,
+      "learning_rate": 1.4238852339181288e-05,
+      "loss": 1.9501,
+      "step": 6305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.094726800918579,
+      "learning_rate": 1.423793859649123e-05,
+      "loss": 2.0001,
+      "step": 6306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.234607219696045,
+      "learning_rate": 1.423702485380117e-05,
+      "loss": 1.5182,
+      "step": 6307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8322553634643555,
+      "learning_rate": 1.4236111111111113e-05,
+      "loss": 1.4591,
+      "step": 6308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9387927055358887,
+      "learning_rate": 1.4235197368421054e-05,
+      "loss": 1.5224,
+      "step": 6309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7885425090789795,
+      "learning_rate": 1.4234283625730996e-05,
+      "loss": 1.4821,
+      "step": 6310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.752751588821411,
+      "learning_rate": 1.4233369883040938e-05,
+      "loss": 1.4823,
+      "step": 6311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4250941276550293,
+      "learning_rate": 1.4232456140350879e-05,
+      "loss": 1.3099,
+      "step": 6312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0732038021087646,
+      "learning_rate": 1.423154239766082e-05,
+      "loss": 1.9045,
+      "step": 6313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1303000450134277,
+      "learning_rate": 1.4230628654970761e-05,
+      "loss": 2.0087,
+      "step": 6314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9449782371520996,
+      "learning_rate": 1.4229714912280704e-05,
+      "loss": 1.7156,
+      "step": 6315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8256900310516357,
+      "learning_rate": 1.4228801169590644e-05,
+      "loss": 1.8844,
+      "step": 6316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.658348321914673,
+      "learning_rate": 1.4227887426900586e-05,
+      "loss": 1.6717,
+      "step": 6317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6275475025177,
+      "learning_rate": 1.4226973684210529e-05,
+      "loss": 1.4582,
+      "step": 6318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6187238693237305,
+      "learning_rate": 1.4226059941520469e-05,
+      "loss": 1.4576,
+      "step": 6319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.74564266204834,
+      "learning_rate": 1.4225146198830411e-05,
+      "loss": 1.5506,
+      "step": 6320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.17985463142395,
+      "learning_rate": 1.4224232456140352e-05,
+      "loss": 1.9277,
+      "step": 6321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.59576678276062,
+      "learning_rate": 1.4223318713450294e-05,
+      "loss": 1.5722,
+      "step": 6322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.725637674331665,
+      "learning_rate": 1.4222404970760235e-05,
+      "loss": 1.4532,
+      "step": 6323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8644859790802,
+      "learning_rate": 1.4221491228070177e-05,
+      "loss": 1.5913,
+      "step": 6324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8821871280670166,
+      "learning_rate": 1.4220577485380119e-05,
+      "loss": 1.7017,
+      "step": 6325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4533820152282715,
+      "learning_rate": 1.421966374269006e-05,
+      "loss": 1.4438,
+      "step": 6326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7265563011169434,
+      "learning_rate": 1.4218750000000002e-05,
+      "loss": 1.4242,
+      "step": 6327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.583104133605957,
+      "learning_rate": 1.4217836257309942e-05,
+      "loss": 1.5393,
+      "step": 6328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.416771411895752,
+      "learning_rate": 1.4216922514619884e-05,
+      "loss": 1.4612,
+      "step": 6329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1025514602661133,
+      "learning_rate": 1.4216008771929825e-05,
+      "loss": 1.8809,
+      "step": 6330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9791505336761475,
+      "learning_rate": 1.4215095029239767e-05,
+      "loss": 1.4587,
+      "step": 6331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3024871349334717,
+      "learning_rate": 1.421418128654971e-05,
+      "loss": 1.3722,
+      "step": 6332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0241267681121826,
+      "learning_rate": 1.421326754385965e-05,
+      "loss": 1.8054,
+      "step": 6333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.48203706741333,
+      "learning_rate": 1.4212353801169592e-05,
+      "loss": 1.8573,
+      "step": 6334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.140866756439209,
+      "learning_rate": 1.4211440058479533e-05,
+      "loss": 1.6652,
+      "step": 6335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.768934965133667,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 1.3835,
+      "step": 6336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2749710083007812,
+      "learning_rate": 1.4209612573099415e-05,
+      "loss": 1.2469,
+      "step": 6337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1706223487854004,
+      "learning_rate": 1.4208698830409358e-05,
+      "loss": 1.878,
+      "step": 6338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.980213165283203,
+      "learning_rate": 1.42077850877193e-05,
+      "loss": 1.6918,
+      "step": 6339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.008521556854248,
+      "learning_rate": 1.420687134502924e-05,
+      "loss": 1.9563,
+      "step": 6340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0189626216888428,
+      "learning_rate": 1.4205957602339183e-05,
+      "loss": 1.5091,
+      "step": 6341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.810006618499756,
+      "learning_rate": 1.4205043859649123e-05,
+      "loss": 1.6115,
+      "step": 6342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8252995014190674,
+      "learning_rate": 1.4204130116959065e-05,
+      "loss": 1.5679,
+      "step": 6343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9882891178131104,
+      "learning_rate": 1.4203216374269006e-05,
+      "loss": 1.9824,
+      "step": 6344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2605884075164795,
+      "learning_rate": 1.4202302631578948e-05,
+      "loss": 1.2201,
+      "step": 6345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.975250005722046,
+      "learning_rate": 1.420138888888889e-05,
+      "loss": 1.7541,
+      "step": 6346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8488168716430664,
+      "learning_rate": 1.4200475146198831e-05,
+      "loss": 1.4183,
+      "step": 6347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9823222160339355,
+      "learning_rate": 1.4199561403508773e-05,
+      "loss": 1.5312,
+      "step": 6348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8372857570648193,
+      "learning_rate": 1.4198647660818714e-05,
+      "loss": 1.7495,
+      "step": 6349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4781408309936523,
+      "learning_rate": 1.4197733918128656e-05,
+      "loss": 1.7558,
+      "step": 6350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9618563652038574,
+      "learning_rate": 1.4196820175438596e-05,
+      "loss": 1.5799,
+      "step": 6351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0699076652526855,
+      "learning_rate": 1.4195906432748539e-05,
+      "loss": 1.5252,
+      "step": 6352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.362518072128296,
+      "learning_rate": 1.419499269005848e-05,
+      "loss": 1.3494,
+      "step": 6353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7171804904937744,
+      "learning_rate": 1.4194078947368421e-05,
+      "loss": 1.3174,
+      "step": 6354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.765826463699341,
+      "learning_rate": 1.4193165204678364e-05,
+      "loss": 1.4274,
+      "step": 6355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1121222972869873,
+      "learning_rate": 1.4192251461988304e-05,
+      "loss": 1.7053,
+      "step": 6356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8994154930114746,
+      "learning_rate": 1.4191337719298246e-05,
+      "loss": 1.821,
+      "step": 6357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1763439178466797,
+      "learning_rate": 1.4190423976608187e-05,
+      "loss": 1.7451,
+      "step": 6358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5662617683410645,
+      "learning_rate": 1.4189510233918129e-05,
+      "loss": 1.3961,
+      "step": 6359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8651084899902344,
+      "learning_rate": 1.4188596491228071e-05,
+      "loss": 1.8999,
+      "step": 6360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8460376262664795,
+      "learning_rate": 1.4187682748538012e-05,
+      "loss": 1.813,
+      "step": 6361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5823731422424316,
+      "learning_rate": 1.4186769005847954e-05,
+      "loss": 2.3306,
+      "step": 6362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8832883834838867,
+      "learning_rate": 1.4185855263157895e-05,
+      "loss": 1.5834,
+      "step": 6363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7257518768310547,
+      "learning_rate": 1.4184941520467837e-05,
+      "loss": 1.7329,
+      "step": 6364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.117851734161377,
+      "learning_rate": 1.4184027777777777e-05,
+      "loss": 1.7149,
+      "step": 6365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7355618476867676,
+      "learning_rate": 1.418311403508772e-05,
+      "loss": 1.6159,
+      "step": 6366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4496867656707764,
+      "learning_rate": 1.4182200292397662e-05,
+      "loss": 1.408,
+      "step": 6367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.031886100769043,
+      "learning_rate": 1.4181286549707602e-05,
+      "loss": 1.5576,
+      "step": 6368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1930623054504395,
+      "learning_rate": 1.4180372807017545e-05,
+      "loss": 1.9053,
+      "step": 6369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.206192970275879,
+      "learning_rate": 1.4179459064327488e-05,
+      "loss": 1.3252,
+      "step": 6370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1693365573883057,
+      "learning_rate": 1.4178545321637429e-05,
+      "loss": 1.8355,
+      "step": 6371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.452277898788452,
+      "learning_rate": 1.4177631578947371e-05,
+      "loss": 1.6991,
+      "step": 6372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5833446979522705,
+      "learning_rate": 1.4176717836257312e-05,
+      "loss": 1.5315,
+      "step": 6373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.556114435195923,
+      "learning_rate": 1.4175804093567254e-05,
+      "loss": 1.4217,
+      "step": 6374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1550981998443604,
+      "learning_rate": 1.4174890350877194e-05,
+      "loss": 1.6673,
+      "step": 6375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3193905353546143,
+      "learning_rate": 1.4173976608187137e-05,
+      "loss": 1.4475,
+      "step": 6376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2272210121154785,
+      "learning_rate": 1.4173062865497079e-05,
+      "loss": 1.3907,
+      "step": 6377
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.551363229751587,
+      "learning_rate": 1.417214912280702e-05,
+      "loss": 1.4939,
+      "step": 6378
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8100340366363525,
+      "learning_rate": 1.4171235380116962e-05,
+      "loss": 1.4831,
+      "step": 6379
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5979254245758057,
+      "learning_rate": 1.4170321637426902e-05,
+      "loss": 1.5463,
+      "step": 6380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5098907947540283,
+      "learning_rate": 1.4169407894736844e-05,
+      "loss": 1.7511,
+      "step": 6381
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.105774164199829,
+      "learning_rate": 1.4168494152046785e-05,
+      "loss": 1.6499,
+      "step": 6382
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6001229286193848,
+      "learning_rate": 1.4167580409356727e-05,
+      "loss": 1.4999,
+      "step": 6383
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.726980686187744,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 1.4126,
+      "step": 6384
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.992363691329956,
+      "learning_rate": 1.416575292397661e-05,
+      "loss": 1.5673,
+      "step": 6385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5639994144439697,
+      "learning_rate": 1.4164839181286552e-05,
+      "loss": 1.8696,
+      "step": 6386
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8829410076141357,
+      "learning_rate": 1.4163925438596493e-05,
+      "loss": 1.4148,
+      "step": 6387
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6861348152160645,
+      "learning_rate": 1.4163011695906435e-05,
+      "loss": 1.3607,
+      "step": 6388
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1692731380462646,
+      "learning_rate": 1.4162097953216375e-05,
+      "loss": 1.0958,
+      "step": 6389
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7581207752227783,
+      "learning_rate": 1.4161184210526318e-05,
+      "loss": 1.8876,
+      "step": 6390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1072824001312256,
+      "learning_rate": 1.416027046783626e-05,
+      "loss": 1.6951,
+      "step": 6391
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7985408306121826,
+      "learning_rate": 1.41593567251462e-05,
+      "loss": 1.7553,
+      "step": 6392
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3482718467712402,
+      "learning_rate": 1.4158442982456143e-05,
+      "loss": 1.3484,
+      "step": 6393
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.464664936065674,
+      "learning_rate": 1.4157529239766083e-05,
+      "loss": 1.0497,
+      "step": 6394
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.801393747329712,
+      "learning_rate": 1.4156615497076025e-05,
+      "loss": 1.598,
+      "step": 6395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.675733804702759,
+      "learning_rate": 1.4155701754385966e-05,
+      "loss": 1.5132,
+      "step": 6396
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7683310508728027,
+      "learning_rate": 1.4154788011695908e-05,
+      "loss": 1.3859,
+      "step": 6397
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1282899379730225,
+      "learning_rate": 1.415387426900585e-05,
+      "loss": 1.6261,
+      "step": 6398
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.028093099594116,
+      "learning_rate": 1.415296052631579e-05,
+      "loss": 1.9605,
+      "step": 6399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1086857318878174,
+      "learning_rate": 1.4152046783625733e-05,
+      "loss": 1.6269,
+      "step": 6400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.534221887588501,
+      "learning_rate": 1.4151133040935674e-05,
+      "loss": 1.3977,
+      "step": 6401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.13651180267334,
+      "learning_rate": 1.4150219298245616e-05,
+      "loss": 1.7272,
+      "step": 6402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6142282485961914,
+      "learning_rate": 1.4149305555555556e-05,
+      "loss": 1.5926,
+      "step": 6403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.459512233734131,
+      "learning_rate": 1.4148391812865499e-05,
+      "loss": 1.4345,
+      "step": 6404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.210611343383789,
+      "learning_rate": 1.414747807017544e-05,
+      "loss": 1.7369,
+      "step": 6405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8072237968444824,
+      "learning_rate": 1.4146564327485381e-05,
+      "loss": 1.2972,
+      "step": 6406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6587483882904053,
+      "learning_rate": 1.4145650584795324e-05,
+      "loss": 1.6858,
+      "step": 6407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1364822387695312,
+      "learning_rate": 1.4144736842105264e-05,
+      "loss": 1.9212,
+      "step": 6408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.436258316040039,
+      "learning_rate": 1.4143823099415206e-05,
+      "loss": 1.3461,
+      "step": 6409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.528174877166748,
+      "learning_rate": 1.4142909356725147e-05,
+      "loss": 1.3989,
+      "step": 6410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9916300773620605,
+      "learning_rate": 1.4141995614035089e-05,
+      "loss": 1.4861,
+      "step": 6411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.40586519241333,
+      "learning_rate": 1.4141081871345031e-05,
+      "loss": 1.4055,
+      "step": 6412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.495177745819092,
+      "learning_rate": 1.4140168128654972e-05,
+      "loss": 1.6784,
+      "step": 6413
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8554465770721436,
+      "learning_rate": 1.4139254385964914e-05,
+      "loss": 1.6526,
+      "step": 6414
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9476187229156494,
+      "learning_rate": 1.4138340643274855e-05,
+      "loss": 1.4472,
+      "step": 6415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6055514812469482,
+      "learning_rate": 1.4137426900584797e-05,
+      "loss": 1.5271,
+      "step": 6416
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.619173288345337,
+      "learning_rate": 1.4136513157894737e-05,
+      "loss": 1.3941,
+      "step": 6417
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2618048191070557,
+      "learning_rate": 1.413559941520468e-05,
+      "loss": 1.8201,
+      "step": 6418
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6465272903442383,
+      "learning_rate": 1.413468567251462e-05,
+      "loss": 1.4191,
+      "step": 6419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.916733980178833,
+      "learning_rate": 1.4133771929824562e-05,
+      "loss": 1.6946,
+      "step": 6420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.269259214401245,
+      "learning_rate": 1.4132858187134504e-05,
+      "loss": 1.6438,
+      "step": 6421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7389638423919678,
+      "learning_rate": 1.4131944444444445e-05,
+      "loss": 1.7919,
+      "step": 6422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0411293506622314,
+      "learning_rate": 1.4131030701754387e-05,
+      "loss": 1.3671,
+      "step": 6423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3793041706085205,
+      "learning_rate": 1.4130116959064328e-05,
+      "loss": 1.737,
+      "step": 6424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.294391393661499,
+      "learning_rate": 1.412920321637427e-05,
+      "loss": 1.3648,
+      "step": 6425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5707743167877197,
+      "learning_rate": 1.412828947368421e-05,
+      "loss": 1.6527,
+      "step": 6426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5793755054473877,
+      "learning_rate": 1.4127375730994153e-05,
+      "loss": 1.7774,
+      "step": 6427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8430569171905518,
+      "learning_rate": 1.4126461988304095e-05,
+      "loss": 1.3463,
+      "step": 6428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.738232135772705,
+      "learning_rate": 1.4125548245614035e-05,
+      "loss": 1.6075,
+      "step": 6429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5916006565093994,
+      "learning_rate": 1.4124634502923978e-05,
+      "loss": 1.351,
+      "step": 6430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.729311227798462,
+      "learning_rate": 1.4123720760233918e-05,
+      "loss": 1.6581,
+      "step": 6431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.455439567565918,
+      "learning_rate": 1.412280701754386e-05,
+      "loss": 1.5291,
+      "step": 6432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9219658374786377,
+      "learning_rate": 1.4121893274853801e-05,
+      "loss": 1.6367,
+      "step": 6433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5283617973327637,
+      "learning_rate": 1.4120979532163743e-05,
+      "loss": 1.3938,
+      "step": 6434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.909128189086914,
+      "learning_rate": 1.4120065789473685e-05,
+      "loss": 1.5099,
+      "step": 6435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.713322401046753,
+      "learning_rate": 1.4119152046783626e-05,
+      "loss": 1.8043,
+      "step": 6436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2295291423797607,
+      "learning_rate": 1.4118238304093568e-05,
+      "loss": 1.7043,
+      "step": 6437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3972551822662354,
+      "learning_rate": 1.4117324561403509e-05,
+      "loss": 1.4712,
+      "step": 6438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5746779441833496,
+      "learning_rate": 1.4116410818713451e-05,
+      "loss": 1.4634,
+      "step": 6439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4181838035583496,
+      "learning_rate": 1.4115497076023391e-05,
+      "loss": 1.4534,
+      "step": 6440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.777219295501709,
+      "learning_rate": 1.4114583333333334e-05,
+      "loss": 1.4632,
+      "step": 6441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6882894039154053,
+      "learning_rate": 1.4113669590643276e-05,
+      "loss": 1.6602,
+      "step": 6442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9679529666900635,
+      "learning_rate": 1.4112755847953216e-05,
+      "loss": 1.6029,
+      "step": 6443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.521071672439575,
+      "learning_rate": 1.4111842105263159e-05,
+      "loss": 1.4299,
+      "step": 6444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.423550605773926,
+      "learning_rate": 1.4110928362573099e-05,
+      "loss": 1.3626,
+      "step": 6445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6014106273651123,
+      "learning_rate": 1.4110014619883041e-05,
+      "loss": 1.4859,
+      "step": 6446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.510014772415161,
+      "learning_rate": 1.4109100877192982e-05,
+      "loss": 1.5112,
+      "step": 6447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1592226028442383,
+      "learning_rate": 1.4108187134502924e-05,
+      "loss": 1.5916,
+      "step": 6448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0831568241119385,
+      "learning_rate": 1.4107273391812866e-05,
+      "loss": 1.516,
+      "step": 6449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.561302423477173,
+      "learning_rate": 1.4106359649122807e-05,
+      "loss": 1.3923,
+      "step": 6450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.942192554473877,
+      "learning_rate": 1.4105445906432749e-05,
+      "loss": 1.7384,
+      "step": 6451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.603619337081909,
+      "learning_rate": 1.410453216374269e-05,
+      "loss": 2.0547,
+      "step": 6452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.798448324203491,
+      "learning_rate": 1.4103618421052632e-05,
+      "loss": 1.6328,
+      "step": 6453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.055691719055176,
+      "learning_rate": 1.4102704678362572e-05,
+      "loss": 1.0934,
+      "step": 6454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.16817569732666,
+      "learning_rate": 1.4101790935672515e-05,
+      "loss": 1.3723,
+      "step": 6455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3405120372772217,
+      "learning_rate": 1.4100877192982459e-05,
+      "loss": 1.8745,
+      "step": 6456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.825360059738159,
+      "learning_rate": 1.40999634502924e-05,
+      "loss": 1.7189,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3020482063293457,
+      "learning_rate": 1.4099049707602341e-05,
+      "loss": 1.6874,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.931701898574829,
+      "learning_rate": 1.4098135964912283e-05,
+      "loss": 1.6374,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.747696876525879,
+      "learning_rate": 1.4097222222222224e-05,
+      "loss": 1.6325,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.956969976425171,
+      "learning_rate": 1.4096308479532166e-05,
+      "loss": 1.695,
+      "step": 6461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.738619327545166,
+      "learning_rate": 1.4095394736842107e-05,
+      "loss": 1.5617,
+      "step": 6462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3641388416290283,
+      "learning_rate": 1.4094480994152049e-05,
+      "loss": 1.6691,
+      "step": 6463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5093581676483154,
+      "learning_rate": 1.409356725146199e-05,
+      "loss": 1.4203,
+      "step": 6464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.419104814529419,
+      "learning_rate": 1.4092653508771932e-05,
+      "loss": 1.9707,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9986307621002197,
+      "learning_rate": 1.4091739766081874e-05,
+      "loss": 1.5273,
+      "step": 6466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.014436960220337,
+      "learning_rate": 1.4090826023391814e-05,
+      "loss": 1.7921,
+      "step": 6467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5397002696990967,
+      "learning_rate": 1.4089912280701757e-05,
+      "loss": 1.4897,
+      "step": 6468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.912163496017456,
+      "learning_rate": 1.4088998538011697e-05,
+      "loss": 1.5117,
+      "step": 6469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0626871585845947,
+      "learning_rate": 1.408808479532164e-05,
+      "loss": 1.5548,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.345714569091797,
+      "learning_rate": 1.408717105263158e-05,
+      "loss": 1.2712,
+      "step": 6471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.013089656829834,
+      "learning_rate": 1.4086257309941522e-05,
+      "loss": 1.9081,
+      "step": 6472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5411922931671143,
+      "learning_rate": 1.4085343567251464e-05,
+      "loss": 1.3066,
+      "step": 6473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9597887992858887,
+      "learning_rate": 1.4084429824561405e-05,
+      "loss": 1.7803,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.67724609375,
+      "learning_rate": 1.4083516081871347e-05,
+      "loss": 1.4955,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1814873218536377,
+      "learning_rate": 1.4082602339181288e-05,
+      "loss": 1.8272,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5223171710968018,
+      "learning_rate": 1.408168859649123e-05,
+      "loss": 1.2262,
+      "step": 6477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.376291275024414,
+      "learning_rate": 1.408077485380117e-05,
+      "loss": 1.5385,
+      "step": 6478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0251331329345703,
+      "learning_rate": 1.4079861111111113e-05,
+      "loss": 1.6979,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.568866014480591,
+      "learning_rate": 1.4078947368421055e-05,
+      "loss": 1.4316,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9059407711029053,
+      "learning_rate": 1.4078033625730995e-05,
+      "loss": 1.4919,
+      "step": 6481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.504286527633667,
+      "learning_rate": 1.4077119883040938e-05,
+      "loss": 1.6096,
+      "step": 6482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1211764812469482,
+      "learning_rate": 1.4076206140350878e-05,
+      "loss": 1.4,
+      "step": 6483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5568275451660156,
+      "learning_rate": 1.407529239766082e-05,
+      "loss": 1.7878,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1366400718688965,
+      "learning_rate": 1.4074378654970761e-05,
+      "loss": 1.9695,
+      "step": 6485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.427182674407959,
+      "learning_rate": 1.4073464912280703e-05,
+      "loss": 1.2249,
+      "step": 6486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1351828575134277,
+      "learning_rate": 1.4072551169590645e-05,
+      "loss": 1.8239,
+      "step": 6487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6057121753692627,
+      "learning_rate": 1.4071637426900586e-05,
+      "loss": 2.0408,
+      "step": 6488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0845439434051514,
+      "learning_rate": 1.4070723684210528e-05,
+      "loss": 1.9068,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.303119659423828,
+      "learning_rate": 1.4069809941520469e-05,
+      "loss": 1.2115,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3044190406799316,
+      "learning_rate": 1.406889619883041e-05,
+      "loss": 1.9038,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.482377767562866,
+      "learning_rate": 1.4067982456140351e-05,
+      "loss": 1.3708,
+      "step": 6492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8078951835632324,
+      "learning_rate": 1.4067068713450294e-05,
+      "loss": 1.5849,
+      "step": 6493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8620259761810303,
+      "learning_rate": 1.4066154970760236e-05,
+      "loss": 1.7865,
+      "step": 6494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5577194690704346,
+      "learning_rate": 1.4065241228070176e-05,
+      "loss": 1.541,
+      "step": 6495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9296696186065674,
+      "learning_rate": 1.4064327485380119e-05,
+      "loss": 1.3521,
+      "step": 6496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9124536514282227,
+      "learning_rate": 1.4063413742690059e-05,
+      "loss": 1.1095,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5721559524536133,
+      "learning_rate": 1.4062500000000001e-05,
+      "loss": 1.6904,
+      "step": 6498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.069610595703125,
+      "learning_rate": 1.4061586257309942e-05,
+      "loss": 1.9107,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.617709159851074,
+      "learning_rate": 1.4060672514619884e-05,
+      "loss": 1.3235,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.201760768890381,
+      "learning_rate": 1.4059758771929826e-05,
+      "loss": 1.9101,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0023651123046875,
+      "learning_rate": 1.4058845029239767e-05,
+      "loss": 1.8297,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7149839401245117,
+      "learning_rate": 1.4057931286549709e-05,
+      "loss": 1.5747,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.681027889251709,
+      "learning_rate": 1.405701754385965e-05,
+      "loss": 1.6753,
+      "step": 6504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8732407093048096,
+      "learning_rate": 1.4056103801169592e-05,
+      "loss": 1.3765,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.563662528991699,
+      "learning_rate": 1.4055190058479532e-05,
+      "loss": 1.5904,
+      "step": 6506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.040543556213379,
+      "learning_rate": 1.4054276315789475e-05,
+      "loss": 1.6806,
+      "step": 6507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.312272787094116,
+      "learning_rate": 1.4053362573099417e-05,
+      "loss": 2.106,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.601762056350708,
+      "learning_rate": 1.4052448830409357e-05,
+      "loss": 1.408,
+      "step": 6509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.465139150619507,
+      "learning_rate": 1.40515350877193e-05,
+      "loss": 1.6729,
+      "step": 6510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2272791862487793,
+      "learning_rate": 1.405062134502924e-05,
+      "loss": 1.5441,
+      "step": 6511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.483968734741211,
+      "learning_rate": 1.4049707602339182e-05,
+      "loss": 1.2866,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9296071529388428,
+      "learning_rate": 1.4048793859649123e-05,
+      "loss": 1.8863,
+      "step": 6513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4955410957336426,
+      "learning_rate": 1.4047880116959065e-05,
+      "loss": 1.2789,
+      "step": 6514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.121929407119751,
+      "learning_rate": 1.4046966374269007e-05,
+      "loss": 1.8752,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.868783712387085,
+      "learning_rate": 1.4046052631578948e-05,
+      "loss": 1.7001,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4108448028564453,
+      "learning_rate": 1.404513888888889e-05,
+      "loss": 1.8062,
+      "step": 6517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6849493980407715,
+      "learning_rate": 1.404422514619883e-05,
+      "loss": 1.5763,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.23856782913208,
+      "learning_rate": 1.4043311403508773e-05,
+      "loss": 1.3967,
+      "step": 6519
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4070091247558594,
+      "learning_rate": 1.4042397660818713e-05,
+      "loss": 1.1932,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4327731132507324,
+      "learning_rate": 1.4041483918128655e-05,
+      "loss": 2.3303,
+      "step": 6521
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7245705127716064,
+      "learning_rate": 1.4040570175438598e-05,
+      "loss": 1.7174,
+      "step": 6522
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8255252838134766,
+      "learning_rate": 1.4039656432748538e-05,
+      "loss": 1.4901,
+      "step": 6523
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.223803997039795,
+      "learning_rate": 1.403874269005848e-05,
+      "loss": 1.6814,
+      "step": 6524
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9751369953155518,
+      "learning_rate": 1.4037828947368421e-05,
+      "loss": 1.6415,
+      "step": 6525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.097250461578369,
+      "learning_rate": 1.4036915204678363e-05,
+      "loss": 1.5195,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.898127317428589,
+      "learning_rate": 1.4036001461988304e-05,
+      "loss": 1.6576,
+      "step": 6527
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7641496658325195,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 1.4665,
+      "step": 6528
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9249322414398193,
+      "learning_rate": 1.4034173976608188e-05,
+      "loss": 1.5999,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1794047355651855,
+      "learning_rate": 1.4033260233918129e-05,
+      "loss": 1.2684,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8377509117126465,
+      "learning_rate": 1.4032346491228071e-05,
+      "loss": 1.5403,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8402652740478516,
+      "learning_rate": 1.4031432748538011e-05,
+      "loss": 1.4342,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.788893699645996,
+      "learning_rate": 1.4030519005847954e-05,
+      "loss": 1.6439,
+      "step": 6533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5837740898132324,
+      "learning_rate": 1.4029605263157894e-05,
+      "loss": 1.3507,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.327691078186035,
+      "learning_rate": 1.4028691520467836e-05,
+      "loss": 1.9926,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9618849754333496,
+      "learning_rate": 1.4027777777777779e-05,
+      "loss": 1.8086,
+      "step": 6536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9821619987487793,
+      "learning_rate": 1.4026864035087719e-05,
+      "loss": 2.0875,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7873988151550293,
+      "learning_rate": 1.4025950292397661e-05,
+      "loss": 1.5791,
+      "step": 6538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5923213958740234,
+      "learning_rate": 1.4025036549707602e-05,
+      "loss": 2.1756,
+      "step": 6539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.445685386657715,
+      "learning_rate": 1.4024122807017544e-05,
+      "loss": 1.9187,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.851198196411133,
+      "learning_rate": 1.4023209064327488e-05,
+      "loss": 2.0394,
+      "step": 6541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.463862180709839,
+      "learning_rate": 1.4022295321637429e-05,
+      "loss": 1.5937,
+      "step": 6542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.040525436401367,
+      "learning_rate": 1.402138157894737e-05,
+      "loss": 1.7831,
+      "step": 6543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.730196475982666,
+      "learning_rate": 1.4020467836257311e-05,
+      "loss": 1.4146,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9080066680908203,
+      "learning_rate": 1.4019554093567254e-05,
+      "loss": 1.8003,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.560852527618408,
+      "learning_rate": 1.4018640350877196e-05,
+      "loss": 1.5899,
+      "step": 6546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.866041898727417,
+      "learning_rate": 1.4017726608187136e-05,
+      "loss": 1.7973,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4691715240478516,
+      "learning_rate": 1.4016812865497079e-05,
+      "loss": 1.7886,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1459896564483643,
+      "learning_rate": 1.4015899122807019e-05,
+      "loss": 1.7309,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5024967193603516,
+      "learning_rate": 1.4014985380116961e-05,
+      "loss": 1.5025,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.919034957885742,
+      "learning_rate": 1.4014071637426902e-05,
+      "loss": 1.8925,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.358146905899048,
+      "learning_rate": 1.4013157894736844e-05,
+      "loss": 1.7903,
+      "step": 6552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1689562797546387,
+      "learning_rate": 1.4012244152046786e-05,
+      "loss": 1.9933,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.07741379737854,
+      "learning_rate": 1.4011330409356727e-05,
+      "loss": 1.6505,
+      "step": 6554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.93888521194458,
+      "learning_rate": 1.4010416666666669e-05,
+      "loss": 1.6344,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0227391719818115,
+      "learning_rate": 1.400950292397661e-05,
+      "loss": 1.8983,
+      "step": 6556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2290780544281006,
+      "learning_rate": 1.4008589181286552e-05,
+      "loss": 1.3106,
+      "step": 6557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.661959409713745,
+      "learning_rate": 1.4007675438596492e-05,
+      "loss": 1.7108,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9673874378204346,
+      "learning_rate": 1.4006761695906434e-05,
+      "loss": 1.4619,
+      "step": 6559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.711583375930786,
+      "learning_rate": 1.4005847953216377e-05,
+      "loss": 1.257,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9798078536987305,
+      "learning_rate": 1.4004934210526317e-05,
+      "loss": 1.8142,
+      "step": 6561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.832597255706787,
+      "learning_rate": 1.400402046783626e-05,
+      "loss": 1.6804,
+      "step": 6562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.890571117401123,
+      "learning_rate": 1.40031067251462e-05,
+      "loss": 1.6169,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.254058837890625,
+      "learning_rate": 1.4002192982456142e-05,
+      "loss": 1.7557,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9084057807922363,
+      "learning_rate": 1.4001279239766083e-05,
+      "loss": 1.6705,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0419869422912598,
+      "learning_rate": 1.4000365497076025e-05,
+      "loss": 1.7856,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3807008266448975,
+      "learning_rate": 1.3999451754385967e-05,
+      "loss": 1.5474,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3,
+      "eval_train_bleu": 0.22551466647071444,
+      "eval_train_loss": 2.132018804550171,
+      "eval_train_runtime": 25.9882,
+      "eval_train_samples_per_second": 38.518,
+      "eval_train_steps_per_second": 1.231,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8205885887145996,
+      "learning_rate": 1.3998538011695908e-05,
+      "loss": 2.2948,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7889020442962646,
+      "learning_rate": 1.399762426900585e-05,
+      "loss": 1.6458,
+      "step": 6569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.379974842071533,
+      "learning_rate": 1.399671052631579e-05,
+      "loss": 1.4312,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8436660766601562,
+      "learning_rate": 1.3995796783625733e-05,
+      "loss": 1.7255,
+      "step": 6571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.912590503692627,
+      "learning_rate": 1.3994883040935673e-05,
+      "loss": 1.9583,
+      "step": 6572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.225209951400757,
+      "learning_rate": 1.3993969298245615e-05,
+      "loss": 1.6044,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7780191898345947,
+      "learning_rate": 1.3993055555555558e-05,
+      "loss": 1.9408,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8418631553649902,
+      "learning_rate": 1.3992141812865498e-05,
+      "loss": 1.6639,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6670310497283936,
+      "learning_rate": 1.399122807017544e-05,
+      "loss": 1.5744,
+      "step": 6576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.834116220474243,
+      "learning_rate": 1.3990314327485381e-05,
+      "loss": 1.5244,
+      "step": 6577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8625121116638184,
+      "learning_rate": 1.3989400584795323e-05,
+      "loss": 1.6452,
+      "step": 6578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.508974313735962,
+      "learning_rate": 1.3988486842105264e-05,
+      "loss": 1.486,
+      "step": 6579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1183032989501953,
+      "learning_rate": 1.3987573099415206e-05,
+      "loss": 1.3196,
+      "step": 6580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.783736228942871,
+      "learning_rate": 1.3986659356725148e-05,
+      "loss": 1.4887,
+      "step": 6581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.750328540802002,
+      "learning_rate": 1.3985745614035089e-05,
+      "loss": 1.5788,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8137574195861816,
+      "learning_rate": 1.398483187134503e-05,
+      "loss": 1.5516,
+      "step": 6583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0097098350524902,
+      "learning_rate": 1.3983918128654971e-05,
+      "loss": 1.7485,
+      "step": 6584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.789245367050171,
+      "learning_rate": 1.3983004385964914e-05,
+      "loss": 1.2495,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8122811317443848,
+      "learning_rate": 1.3982090643274854e-05,
+      "loss": 1.5989,
+      "step": 6586
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6988203525543213,
+      "learning_rate": 1.3981176900584796e-05,
+      "loss": 1.6931,
+      "step": 6587
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3415255546569824,
+      "learning_rate": 1.3980263157894739e-05,
+      "loss": 1.5447,
+      "step": 6588
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5821356773376465,
+      "learning_rate": 1.3979349415204679e-05,
+      "loss": 1.4317,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5852203369140625,
+      "learning_rate": 1.3978435672514621e-05,
+      "loss": 1.4133,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9407856464385986,
+      "learning_rate": 1.3977521929824562e-05,
+      "loss": 1.6774,
+      "step": 6591
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5229196548461914,
+      "learning_rate": 1.3976608187134504e-05,
+      "loss": 1.4076,
+      "step": 6592
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7118120193481445,
+      "learning_rate": 1.3975694444444445e-05,
+      "loss": 1.6945,
+      "step": 6593
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6112353801727295,
+      "learning_rate": 1.3974780701754387e-05,
+      "loss": 1.3875,
+      "step": 6594
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5801966190338135,
+      "learning_rate": 1.3973866959064329e-05,
+      "loss": 1.654,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0476956367492676,
+      "learning_rate": 1.397295321637427e-05,
+      "loss": 1.7331,
+      "step": 6596
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.371410369873047,
+      "learning_rate": 1.3972039473684212e-05,
+      "loss": 1.3965,
+      "step": 6597
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5534348487854004,
+      "learning_rate": 1.3971125730994152e-05,
+      "loss": 1.6489,
+      "step": 6598
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0425572395324707,
+      "learning_rate": 1.3970211988304095e-05,
+      "loss": 1.9304,
+      "step": 6599
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0657835006713867,
+      "learning_rate": 1.3969298245614035e-05,
+      "loss": 1.6661,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8455708026885986,
+      "learning_rate": 1.3968384502923977e-05,
+      "loss": 1.7593,
+      "step": 6601
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7360172271728516,
+      "learning_rate": 1.396747076023392e-05,
+      "loss": 1.6314,
+      "step": 6602
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3259994983673096,
+      "learning_rate": 1.396655701754386e-05,
+      "loss": 1.7876,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8697052001953125,
+      "learning_rate": 1.3965643274853802e-05,
+      "loss": 1.6658,
+      "step": 6604
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2412641048431396,
+      "learning_rate": 1.3964729532163743e-05,
+      "loss": 2.0418,
+      "step": 6605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.122339963912964,
+      "learning_rate": 1.3963815789473685e-05,
+      "loss": 1.7323,
+      "step": 6606
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.100515604019165,
+      "learning_rate": 1.3962902046783626e-05,
+      "loss": 1.89,
+      "step": 6607
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8553011417388916,
+      "learning_rate": 1.3961988304093568e-05,
+      "loss": 1.8058,
+      "step": 6608
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.706554651260376,
+      "learning_rate": 1.396107456140351e-05,
+      "loss": 1.7491,
+      "step": 6609
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3330085277557373,
+      "learning_rate": 1.396016081871345e-05,
+      "loss": 2.1052,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.131976366043091,
+      "learning_rate": 1.3959247076023393e-05,
+      "loss": 2.0357,
+      "step": 6611
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6270740032196045,
+      "learning_rate": 1.3958333333333333e-05,
+      "loss": 1.6432,
+      "step": 6612
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.644266366958618,
+      "learning_rate": 1.3957419590643275e-05,
+      "loss": 1.456,
+      "step": 6613
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5185883045196533,
+      "learning_rate": 1.3956505847953216e-05,
+      "loss": 1.5341,
+      "step": 6614
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.200831890106201,
+      "learning_rate": 1.3955592105263158e-05,
+      "loss": 1.6961,
+      "step": 6615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.878873109817505,
+      "learning_rate": 1.39546783625731e-05,
+      "loss": 1.6738,
+      "step": 6616
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9173529148101807,
+      "learning_rate": 1.3953764619883041e-05,
+      "loss": 1.6126,
+      "step": 6617
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0428991317749023,
+      "learning_rate": 1.3952850877192983e-05,
+      "loss": 1.9443,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4856362342834473,
+      "learning_rate": 1.3951937134502924e-05,
+      "loss": 1.5524,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.758016586303711,
+      "learning_rate": 1.3951023391812866e-05,
+      "loss": 1.4683,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4270081520080566,
+      "learning_rate": 1.3950109649122806e-05,
+      "loss": 1.5937,
+      "step": 6621
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8033902645111084,
+      "learning_rate": 1.3949195906432749e-05,
+      "loss": 1.7112,
+      "step": 6622
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7682812213897705,
+      "learning_rate": 1.3948282163742691e-05,
+      "loss": 1.7425,
+      "step": 6623
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5834412574768066,
+      "learning_rate": 1.3947368421052631e-05,
+      "loss": 1.4983,
+      "step": 6624
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.343904972076416,
+      "learning_rate": 1.3946454678362574e-05,
+      "loss": 1.3294,
+      "step": 6625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.78882098197937,
+      "learning_rate": 1.3945540935672514e-05,
+      "loss": 1.6444,
+      "step": 6626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3542263507843018,
+      "learning_rate": 1.3944627192982458e-05,
+      "loss": 1.4333,
+      "step": 6627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7537243366241455,
+      "learning_rate": 1.39437134502924e-05,
+      "loss": 1.5809,
+      "step": 6628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.393742561340332,
+      "learning_rate": 1.394279970760234e-05,
+      "loss": 1.726,
+      "step": 6629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7899715900421143,
+      "learning_rate": 1.3941885964912283e-05,
+      "loss": 1.6986,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8352789878845215,
+      "learning_rate": 1.3940972222222224e-05,
+      "loss": 1.5558,
+      "step": 6631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.482501268386841,
+      "learning_rate": 1.3940058479532166e-05,
+      "loss": 1.3066,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1656293869018555,
+      "learning_rate": 1.3939144736842108e-05,
+      "loss": 1.6973,
+      "step": 6633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.831303358078003,
+      "learning_rate": 1.3938230994152049e-05,
+      "loss": 1.6235,
+      "step": 6634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2564809322357178,
+      "learning_rate": 1.393731725146199e-05,
+      "loss": 1.8579,
+      "step": 6635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.905776262283325,
+      "learning_rate": 1.3936403508771931e-05,
+      "loss": 2.0196,
+      "step": 6636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.263751745223999,
+      "learning_rate": 1.3935489766081874e-05,
+      "loss": 2.4627,
+      "step": 6637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7678282260894775,
+      "learning_rate": 1.3934576023391814e-05,
+      "loss": 1.5342,
+      "step": 6638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3895230293273926,
+      "learning_rate": 1.3933662280701756e-05,
+      "loss": 1.4906,
+      "step": 6639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6073107719421387,
+      "learning_rate": 1.3932748538011698e-05,
+      "loss": 1.4312,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.790027618408203,
+      "learning_rate": 1.3931834795321639e-05,
+      "loss": 1.7148,
+      "step": 6641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6608517169952393,
+      "learning_rate": 1.3930921052631581e-05,
+      "loss": 1.3882,
+      "step": 6642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4098775386810303,
+      "learning_rate": 1.3930007309941522e-05,
+      "loss": 1.6847,
+      "step": 6643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.353687286376953,
+      "learning_rate": 1.3929093567251464e-05,
+      "loss": 1.3315,
+      "step": 6644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.577115535736084,
+      "learning_rate": 1.3928179824561405e-05,
+      "loss": 1.8518,
+      "step": 6645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8502073287963867,
+      "learning_rate": 1.3927266081871347e-05,
+      "loss": 1.6852,
+      "step": 6646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8415627479553223,
+      "learning_rate": 1.3926352339181289e-05,
+      "loss": 1.8049,
+      "step": 6647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3035011291503906,
+      "learning_rate": 1.392543859649123e-05,
+      "loss": 1.0844,
+      "step": 6648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.87026047706604,
+      "learning_rate": 1.3924524853801172e-05,
+      "loss": 1.8103,
+      "step": 6649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2861711978912354,
+      "learning_rate": 1.3923611111111112e-05,
+      "loss": 1.8232,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.301989793777466,
+      "learning_rate": 1.3922697368421054e-05,
+      "loss": 1.1706,
+      "step": 6651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4842684268951416,
+      "learning_rate": 1.3921783625730995e-05,
+      "loss": 1.3696,
+      "step": 6652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9612534046173096,
+      "learning_rate": 1.3920869883040937e-05,
+      "loss": 1.7836,
+      "step": 6653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7906248569488525,
+      "learning_rate": 1.391995614035088e-05,
+      "loss": 1.8041,
+      "step": 6654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.020707130432129,
+      "learning_rate": 1.391904239766082e-05,
+      "loss": 1.9183,
+      "step": 6655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.09311842918396,
+      "learning_rate": 1.3918128654970762e-05,
+      "loss": 1.6659,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4578962326049805,
+      "learning_rate": 1.3917214912280703e-05,
+      "loss": 2.0141,
+      "step": 6657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3486392498016357,
+      "learning_rate": 1.3916301169590645e-05,
+      "loss": 1.7643,
+      "step": 6658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9599051475524902,
+      "learning_rate": 1.3915387426900585e-05,
+      "loss": 1.5213,
+      "step": 6659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8103458881378174,
+      "learning_rate": 1.3914473684210528e-05,
+      "loss": 1.7137,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.704557418823242,
+      "learning_rate": 1.391355994152047e-05,
+      "loss": 1.3874,
+      "step": 6661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.68798828125,
+      "learning_rate": 1.391264619883041e-05,
+      "loss": 1.501,
+      "step": 6662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6771488189697266,
+      "learning_rate": 1.3911732456140353e-05,
+      "loss": 1.5408,
+      "step": 6663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.186177968978882,
+      "learning_rate": 1.3910818713450293e-05,
+      "loss": 1.1401,
+      "step": 6664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.724609851837158,
+      "learning_rate": 1.3909904970760235e-05,
+      "loss": 1.6199,
+      "step": 6665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8777503967285156,
+      "learning_rate": 1.3908991228070176e-05,
+      "loss": 1.6118,
+      "step": 6666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.470581293106079,
+      "learning_rate": 1.3908077485380118e-05,
+      "loss": 1.3966,
+      "step": 6667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.094134569168091,
+      "learning_rate": 1.390716374269006e-05,
+      "loss": 1.4671,
+      "step": 6668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.837286949157715,
+      "learning_rate": 1.3906250000000001e-05,
+      "loss": 1.773,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.863492012023926,
+      "learning_rate": 1.3905336257309943e-05,
+      "loss": 1.6514,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8258426189422607,
+      "learning_rate": 1.3904422514619884e-05,
+      "loss": 1.7453,
+      "step": 6671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.542546510696411,
+      "learning_rate": 1.3903508771929826e-05,
+      "loss": 1.8741,
+      "step": 6672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2215113639831543,
+      "learning_rate": 1.3902595029239766e-05,
+      "loss": 1.5731,
+      "step": 6673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.777700185775757,
+      "learning_rate": 1.3901681286549709e-05,
+      "loss": 1.6004,
+      "step": 6674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1828737258911133,
+      "learning_rate": 1.390076754385965e-05,
+      "loss": 2.1112,
+      "step": 6675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3636667728424072,
+      "learning_rate": 1.3899853801169591e-05,
+      "loss": 1.6551,
+      "step": 6676
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.949023962020874,
+      "learning_rate": 1.3898940058479534e-05,
+      "loss": 1.3879,
+      "step": 6677
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.542902946472168,
+      "learning_rate": 1.3898026315789474e-05,
+      "loss": 1.3967,
+      "step": 6678
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.152531862258911,
+      "learning_rate": 1.3897112573099416e-05,
+      "loss": 1.0285,
+      "step": 6679
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8420629501342773,
+      "learning_rate": 1.3896198830409357e-05,
+      "loss": 1.5187,
+      "step": 6680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7342076301574707,
+      "learning_rate": 1.3895285087719299e-05,
+      "loss": 1.442,
+      "step": 6681
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.220863103866577,
+      "learning_rate": 1.3894371345029241e-05,
+      "loss": 1.5641,
+      "step": 6682
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.862306594848633,
+      "learning_rate": 1.3893457602339182e-05,
+      "loss": 1.6761,
+      "step": 6683
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3127639293670654,
+      "learning_rate": 1.3892543859649124e-05,
+      "loss": 1.2753,
+      "step": 6684
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.220876693725586,
+      "learning_rate": 1.3891630116959065e-05,
+      "loss": 1.8317,
+      "step": 6685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7388737201690674,
+      "learning_rate": 1.3890716374269007e-05,
+      "loss": 1.8648,
+      "step": 6686
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6114282608032227,
+      "learning_rate": 1.3889802631578947e-05,
+      "loss": 1.6303,
+      "step": 6687
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.079401731491089,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 1.766,
+      "step": 6688
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.861564874649048,
+      "learning_rate": 1.3887975146198832e-05,
+      "loss": 1.41,
+      "step": 6689
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9316413402557373,
+      "learning_rate": 1.3887061403508772e-05,
+      "loss": 1.9892,
+      "step": 6690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.672434091567993,
+      "learning_rate": 1.3886147660818715e-05,
+      "loss": 1.3747,
+      "step": 6691
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6244001388549805,
+      "learning_rate": 1.3885233918128655e-05,
+      "loss": 1.5092,
+      "step": 6692
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8585116863250732,
+      "learning_rate": 1.3884320175438597e-05,
+      "loss": 1.8609,
+      "step": 6693
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4552018642425537,
+      "learning_rate": 1.3883406432748538e-05,
+      "loss": 1.5778,
+      "step": 6694
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.521862745285034,
+      "learning_rate": 1.388249269005848e-05,
+      "loss": 1.7578,
+      "step": 6695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7968814373016357,
+      "learning_rate": 1.3881578947368422e-05,
+      "loss": 1.629,
+      "step": 6696
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8171489238739014,
+      "learning_rate": 1.3880665204678363e-05,
+      "loss": 1.518,
+      "step": 6697
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1931285858154297,
+      "learning_rate": 1.3879751461988305e-05,
+      "loss": 1.3056,
+      "step": 6698
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8894481658935547,
+      "learning_rate": 1.3878837719298246e-05,
+      "loss": 1.9445,
+      "step": 6699
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1979873180389404,
+      "learning_rate": 1.3877923976608188e-05,
+      "loss": 1.0807,
+      "step": 6700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8235557079315186,
+      "learning_rate": 1.3877010233918128e-05,
+      "loss": 1.7606,
+      "step": 6701
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.435800552368164,
+      "learning_rate": 1.387609649122807e-05,
+      "loss": 1.4438,
+      "step": 6702
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4883248805999756,
+      "learning_rate": 1.3875182748538013e-05,
+      "loss": 1.5122,
+      "step": 6703
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.672346353530884,
+      "learning_rate": 1.3874269005847953e-05,
+      "loss": 1.6434,
+      "step": 6704
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.888458490371704,
+      "learning_rate": 1.3873355263157895e-05,
+      "loss": 1.4405,
+      "step": 6705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.333225727081299,
+      "learning_rate": 1.3872441520467836e-05,
+      "loss": 1.7927,
+      "step": 6706
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.195791721343994,
+      "learning_rate": 1.3871527777777778e-05,
+      "loss": 1.6923,
+      "step": 6707
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7395644187927246,
+      "learning_rate": 1.3870614035087719e-05,
+      "loss": 1.626,
+      "step": 6708
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9430335760116577,
+      "learning_rate": 1.3869700292397661e-05,
+      "loss": 0.8543,
+      "step": 6709
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.717339277267456,
+      "learning_rate": 1.3868786549707603e-05,
+      "loss": 1.4641,
+      "step": 6710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1202821731567383,
+      "learning_rate": 1.3867872807017544e-05,
+      "loss": 1.8808,
+      "step": 6711
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.140056848526001,
+      "learning_rate": 1.3866959064327488e-05,
+      "loss": 1.4691,
+      "step": 6712
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1936395168304443,
+      "learning_rate": 1.386604532163743e-05,
+      "loss": 1.3755,
+      "step": 6713
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.45335054397583,
+      "learning_rate": 1.386513157894737e-05,
+      "loss": 1.4861,
+      "step": 6714
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9241535663604736,
+      "learning_rate": 1.3864217836257313e-05,
+      "loss": 1.688,
+      "step": 6715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7418746948242188,
+      "learning_rate": 1.3863304093567253e-05,
+      "loss": 1.9444,
+      "step": 6716
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7798213958740234,
+      "learning_rate": 1.3862390350877195e-05,
+      "loss": 1.6321,
+      "step": 6717
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5857014656066895,
+      "learning_rate": 1.3861476608187136e-05,
+      "loss": 1.4441,
+      "step": 6718
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1652722358703613,
+      "learning_rate": 1.3860562865497078e-05,
+      "loss": 1.8164,
+      "step": 6719
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6386868953704834,
+      "learning_rate": 1.385964912280702e-05,
+      "loss": 1.42,
+      "step": 6720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6978907585144043,
+      "learning_rate": 1.385873538011696e-05,
+      "loss": 1.6496,
+      "step": 6721
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4740161895751953,
+      "learning_rate": 1.3857821637426903e-05,
+      "loss": 1.4574,
+      "step": 6722
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9834628105163574,
+      "learning_rate": 1.3856907894736844e-05,
+      "loss": 1.7532,
+      "step": 6723
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5874924659729004,
+      "learning_rate": 1.3855994152046786e-05,
+      "loss": 1.2845,
+      "step": 6724
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.48478364944458,
+      "learning_rate": 1.3855080409356726e-05,
+      "loss": 1.3577,
+      "step": 6725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.537146806716919,
+      "learning_rate": 1.3854166666666669e-05,
+      "loss": 2.4532,
+      "step": 6726
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9821856021881104,
+      "learning_rate": 1.385325292397661e-05,
+      "loss": 1.5343,
+      "step": 6727
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.652658700942993,
+      "learning_rate": 1.3852339181286551e-05,
+      "loss": 1.4699,
+      "step": 6728
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8824551105499268,
+      "learning_rate": 1.3851425438596494e-05,
+      "loss": 1.6898,
+      "step": 6729
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9585533142089844,
+      "learning_rate": 1.3850511695906434e-05,
+      "loss": 1.7848,
+      "step": 6730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0884039402008057,
+      "learning_rate": 1.3849597953216376e-05,
+      "loss": 1.6991,
+      "step": 6731
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.427269458770752,
+      "learning_rate": 1.3848684210526317e-05,
+      "loss": 1.6084,
+      "step": 6732
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7282333374023438,
+      "learning_rate": 1.3847770467836259e-05,
+      "loss": 1.8375,
+      "step": 6733
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7356057167053223,
+      "learning_rate": 1.3846856725146201e-05,
+      "loss": 1.6137,
+      "step": 6734
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2540550231933594,
+      "learning_rate": 1.3845942982456142e-05,
+      "loss": 1.6868,
+      "step": 6735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7242588996887207,
+      "learning_rate": 1.3845029239766084e-05,
+      "loss": 1.6077,
+      "step": 6736
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2895243167877197,
+      "learning_rate": 1.3844115497076025e-05,
+      "loss": 1.4368,
+      "step": 6737
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.554555654525757,
+      "learning_rate": 1.3843201754385967e-05,
+      "loss": 1.5725,
+      "step": 6738
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2540535926818848,
+      "learning_rate": 1.3842288011695907e-05,
+      "loss": 1.9946,
+      "step": 6739
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.744932174682617,
+      "learning_rate": 1.384137426900585e-05,
+      "loss": 1.6394,
+      "step": 6740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3971025943756104,
+      "learning_rate": 1.3840460526315792e-05,
+      "loss": 1.5055,
+      "step": 6741
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5589303970336914,
+      "learning_rate": 1.3839546783625732e-05,
+      "loss": 1.4782,
+      "step": 6742
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.213980197906494,
+      "learning_rate": 1.3838633040935674e-05,
+      "loss": 1.218,
+      "step": 6743
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6820027828216553,
+      "learning_rate": 1.3837719298245615e-05,
+      "loss": 1.3955,
+      "step": 6744
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1126720905303955,
+      "learning_rate": 1.3836805555555557e-05,
+      "loss": 1.8298,
+      "step": 6745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.219555616378784,
+      "learning_rate": 1.3835891812865498e-05,
+      "loss": 1.2936,
+      "step": 6746
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2858428955078125,
+      "learning_rate": 1.383497807017544e-05,
+      "loss": 1.2787,
+      "step": 6747
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.462223529815674,
+      "learning_rate": 1.3834064327485382e-05,
+      "loss": 1.8661,
+      "step": 6748
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8489177227020264,
+      "learning_rate": 1.3833150584795323e-05,
+      "loss": 1.5992,
+      "step": 6749
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8385989665985107,
+      "learning_rate": 1.3832236842105265e-05,
+      "loss": 1.4643,
+      "step": 6750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.496729850769043,
+      "learning_rate": 1.3831323099415205e-05,
+      "loss": 1.519,
+      "step": 6751
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.008730173110962,
+      "learning_rate": 1.3830409356725148e-05,
+      "loss": 1.5494,
+      "step": 6752
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9411892890930176,
+      "learning_rate": 1.3829495614035088e-05,
+      "loss": 1.6068,
+      "step": 6753
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.588364362716675,
+      "learning_rate": 1.382858187134503e-05,
+      "loss": 1.4357,
+      "step": 6754
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.561660051345825,
+      "learning_rate": 1.3827668128654973e-05,
+      "loss": 1.5779,
+      "step": 6755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6623730659484863,
+      "learning_rate": 1.3826754385964913e-05,
+      "loss": 1.4197,
+      "step": 6756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.901285171508789,
+      "learning_rate": 1.3825840643274855e-05,
+      "loss": 1.5571,
+      "step": 6757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.650787591934204,
+      "learning_rate": 1.3824926900584796e-05,
+      "loss": 1.5977,
+      "step": 6758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.784062385559082,
+      "learning_rate": 1.3824013157894738e-05,
+      "loss": 1.8881,
+      "step": 6759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.374589204788208,
+      "learning_rate": 1.3823099415204679e-05,
+      "loss": 1.4646,
+      "step": 6760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8574862480163574,
+      "learning_rate": 1.3822185672514621e-05,
+      "loss": 1.5021,
+      "step": 6761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.876465082168579,
+      "learning_rate": 1.3821271929824563e-05,
+      "loss": 1.6109,
+      "step": 6762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.682917594909668,
+      "learning_rate": 1.3820358187134504e-05,
+      "loss": 1.645,
+      "step": 6763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.893181800842285,
+      "learning_rate": 1.3819444444444446e-05,
+      "loss": 1.7945,
+      "step": 6764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1061458587646484,
+      "learning_rate": 1.3818530701754386e-05,
+      "loss": 1.9644,
+      "step": 6765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.813389778137207,
+      "learning_rate": 1.3817616959064329e-05,
+      "loss": 1.4019,
+      "step": 6766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5284085273742676,
+      "learning_rate": 1.3816703216374269e-05,
+      "loss": 1.5973,
+      "step": 6767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.823585033416748,
+      "learning_rate": 1.3815789473684211e-05,
+      "loss": 1.7308,
+      "step": 6768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6164684295654297,
+      "learning_rate": 1.3814875730994154e-05,
+      "loss": 1.5685,
+      "step": 6769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4643657207489014,
+      "learning_rate": 1.3813961988304094e-05,
+      "loss": 1.4222,
+      "step": 6770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8977293968200684,
+      "learning_rate": 1.3813048245614036e-05,
+      "loss": 1.6479,
+      "step": 6771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.792365789413452,
+      "learning_rate": 1.3812134502923977e-05,
+      "loss": 2.0954,
+      "step": 6772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.073976516723633,
+      "learning_rate": 1.3811220760233919e-05,
+      "loss": 1.3448,
+      "step": 6773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1304149627685547,
+      "learning_rate": 1.381030701754386e-05,
+      "loss": 1.7776,
+      "step": 6774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7535173892974854,
+      "learning_rate": 1.3809393274853802e-05,
+      "loss": 1.614,
+      "step": 6775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8621387481689453,
+      "learning_rate": 1.3808479532163744e-05,
+      "loss": 1.607,
+      "step": 6776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5054759979248047,
+      "learning_rate": 1.3807565789473685e-05,
+      "loss": 1.4637,
+      "step": 6777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.443587064743042,
+      "learning_rate": 1.3806652046783627e-05,
+      "loss": 1.9124,
+      "step": 6778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1644272804260254,
+      "learning_rate": 1.3805738304093567e-05,
+      "loss": 1.3834,
+      "step": 6779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2965497970581055,
+      "learning_rate": 1.380482456140351e-05,
+      "loss": 1.3268,
+      "step": 6780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.141469717025757,
+      "learning_rate": 1.380391081871345e-05,
+      "loss": 1.9088,
+      "step": 6781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7538132667541504,
+      "learning_rate": 1.3802997076023392e-05,
+      "loss": 1.6331,
+      "step": 6782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.42398738861084,
+      "learning_rate": 1.3802083333333335e-05,
+      "loss": 1.2761,
+      "step": 6783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6035568714141846,
+      "learning_rate": 1.3801169590643275e-05,
+      "loss": 1.2409,
+      "step": 6784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.35518217086792,
+      "learning_rate": 1.3800255847953217e-05,
+      "loss": 1.257,
+      "step": 6785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5266401767730713,
+      "learning_rate": 1.3799342105263158e-05,
+      "loss": 1.8999,
+      "step": 6786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8089535236358643,
+      "learning_rate": 1.37984283625731e-05,
+      "loss": 1.6228,
+      "step": 6787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5935556888580322,
+      "learning_rate": 1.379751461988304e-05,
+      "loss": 1.4712,
+      "step": 6788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2506067752838135,
+      "learning_rate": 1.3796600877192983e-05,
+      "loss": 1.7005,
+      "step": 6789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.905414342880249,
+      "learning_rate": 1.3795687134502925e-05,
+      "loss": 1.8136,
+      "step": 6790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9010519981384277,
+      "learning_rate": 1.3794773391812866e-05,
+      "loss": 1.7429,
+      "step": 6791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7573885917663574,
+      "learning_rate": 1.3793859649122808e-05,
+      "loss": 1.7995,
+      "step": 6792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.928812265396118,
+      "learning_rate": 1.3792945906432748e-05,
+      "loss": 1.6145,
+      "step": 6793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6746528148651123,
+      "learning_rate": 1.379203216374269e-05,
+      "loss": 1.5936,
+      "step": 6794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.855553150177002,
+      "learning_rate": 1.3791118421052631e-05,
+      "loss": 1.9989,
+      "step": 6795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3293089866638184,
+      "learning_rate": 1.3790204678362573e-05,
+      "loss": 1.3016,
+      "step": 6796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.322483539581299,
+      "learning_rate": 1.3789290935672515e-05,
+      "loss": 1.9777,
+      "step": 6797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.43819522857666,
+      "learning_rate": 1.3788377192982458e-05,
+      "loss": 1.298,
+      "step": 6798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.741579294204712,
+      "learning_rate": 1.37874634502924e-05,
+      "loss": 1.8868,
+      "step": 6799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6391756534576416,
+      "learning_rate": 1.3786549707602342e-05,
+      "loss": 1.5635,
+      "step": 6800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.865779876708984,
+      "learning_rate": 1.3785635964912283e-05,
+      "loss": 1.2401,
+      "step": 6801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8899590969085693,
+      "learning_rate": 1.3784722222222225e-05,
+      "loss": 1.5233,
+      "step": 6802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.020313024520874,
+      "learning_rate": 1.3783808479532165e-05,
+      "loss": 1.7429,
+      "step": 6803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8322765827178955,
+      "learning_rate": 1.3782894736842108e-05,
+      "loss": 1.5569,
+      "step": 6804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4183266162872314,
+      "learning_rate": 1.3781980994152048e-05,
+      "loss": 1.3378,
+      "step": 6805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9566407203674316,
+      "learning_rate": 1.378106725146199e-05,
+      "loss": 1.438,
+      "step": 6806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.820204257965088,
+      "learning_rate": 1.3780153508771933e-05,
+      "loss": 1.8481,
+      "step": 6807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7198047637939453,
+      "learning_rate": 1.3779239766081873e-05,
+      "loss": 1.5923,
+      "step": 6808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.312114953994751,
+      "learning_rate": 1.3778326023391815e-05,
+      "loss": 1.1962,
+      "step": 6809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7412166595458984,
+      "learning_rate": 1.3777412280701756e-05,
+      "loss": 1.577,
+      "step": 6810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4055635929107666,
+      "learning_rate": 1.3776498538011698e-05,
+      "loss": 1.5733,
+      "step": 6811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.335371255874634,
+      "learning_rate": 1.3775584795321639e-05,
+      "loss": 1.5176,
+      "step": 6812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.476555109024048,
+      "learning_rate": 1.377467105263158e-05,
+      "loss": 1.5995,
+      "step": 6813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2852771282196045,
+      "learning_rate": 1.3773757309941523e-05,
+      "loss": 1.7136,
+      "step": 6814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.298440933227539,
+      "learning_rate": 1.3772843567251464e-05,
+      "loss": 1.8823,
+      "step": 6815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0247209072113037,
+      "learning_rate": 1.3771929824561406e-05,
+      "loss": 1.8023,
+      "step": 6816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1023809909820557,
+      "learning_rate": 1.3771016081871346e-05,
+      "loss": 2.1741,
+      "step": 6817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.288228988647461,
+      "learning_rate": 1.3770102339181289e-05,
+      "loss": 1.5032,
+      "step": 6818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7364161014556885,
+      "learning_rate": 1.3769188596491229e-05,
+      "loss": 1.5688,
+      "step": 6819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.409148693084717,
+      "learning_rate": 1.3768274853801171e-05,
+      "loss": 2.0175,
+      "step": 6820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8065028190612793,
+      "learning_rate": 1.3767361111111114e-05,
+      "loss": 1.6602,
+      "step": 6821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.847278594970703,
+      "learning_rate": 1.3766447368421054e-05,
+      "loss": 1.5133,
+      "step": 6822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8224966526031494,
+      "learning_rate": 1.3765533625730996e-05,
+      "loss": 1.4831,
+      "step": 6823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.781510829925537,
+      "learning_rate": 1.3764619883040937e-05,
+      "loss": 1.4302,
+      "step": 6824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6217098236083984,
+      "learning_rate": 1.3763706140350879e-05,
+      "loss": 1.5845,
+      "step": 6825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.277649164199829,
+      "learning_rate": 1.376279239766082e-05,
+      "loss": 1.9575,
+      "step": 6826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.48419189453125,
+      "learning_rate": 1.3761878654970762e-05,
+      "loss": 1.4132,
+      "step": 6827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.828747510910034,
+      "learning_rate": 1.3760964912280704e-05,
+      "loss": 1.6635,
+      "step": 6828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.070056438446045,
+      "learning_rate": 1.3760051169590645e-05,
+      "loss": 1.3384,
+      "step": 6829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.120586395263672,
+      "learning_rate": 1.3759137426900587e-05,
+      "loss": 1.7798,
+      "step": 6830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7212843894958496,
+      "learning_rate": 1.3758223684210527e-05,
+      "loss": 1.5527,
+      "step": 6831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9869585037231445,
+      "learning_rate": 1.375730994152047e-05,
+      "loss": 1.683,
+      "step": 6832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3921425342559814,
+      "learning_rate": 1.375639619883041e-05,
+      "loss": 1.5753,
+      "step": 6833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2232048511505127,
+      "learning_rate": 1.3755482456140352e-05,
+      "loss": 1.3403,
+      "step": 6834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.951045274734497,
+      "learning_rate": 1.3754568713450294e-05,
+      "loss": 1.7546,
+      "step": 6835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6766936779022217,
+      "learning_rate": 1.3753654970760235e-05,
+      "loss": 1.5903,
+      "step": 6836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7826833724975586,
+      "learning_rate": 1.3752741228070177e-05,
+      "loss": 1.693,
+      "step": 6837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3125789165496826,
+      "learning_rate": 1.3751827485380118e-05,
+      "loss": 1.3642,
+      "step": 6838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.843789577484131,
+      "learning_rate": 1.375091374269006e-05,
+      "loss": 1.4053,
+      "step": 6839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9857707023620605,
+      "learning_rate": 1.375e-05,
+      "loss": 1.6255,
+      "step": 6840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.177401304244995,
+      "learning_rate": 1.3749086257309943e-05,
+      "loss": 1.8285,
+      "step": 6841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.598741054534912,
+      "learning_rate": 1.3748172514619885e-05,
+      "loss": 1.432,
+      "step": 6842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6077005863189697,
+      "learning_rate": 1.3747258771929825e-05,
+      "loss": 1.5488,
+      "step": 6843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.845355987548828,
+      "learning_rate": 1.3746345029239768e-05,
+      "loss": 1.4719,
+      "step": 6844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5320067405700684,
+      "learning_rate": 1.3745431286549708e-05,
+      "loss": 1.2946,
+      "step": 6845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.532179832458496,
+      "learning_rate": 1.374451754385965e-05,
+      "loss": 1.4664,
+      "step": 6846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6254098415374756,
+      "learning_rate": 1.3743603801169591e-05,
+      "loss": 1.525,
+      "step": 6847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6689224243164062,
+      "learning_rate": 1.3742690058479533e-05,
+      "loss": 1.4597,
+      "step": 6848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.26853346824646,
+      "learning_rate": 1.3741776315789475e-05,
+      "loss": 1.4111,
+      "step": 6849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9943795204162598,
+      "learning_rate": 1.3740862573099416e-05,
+      "loss": 1.4901,
+      "step": 6850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1507887840270996,
+      "learning_rate": 1.3739948830409358e-05,
+      "loss": 1.7344,
+      "step": 6851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6281771659851074,
+      "learning_rate": 1.3739035087719299e-05,
+      "loss": 1.4214,
+      "step": 6852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6193861961364746,
+      "learning_rate": 1.3738121345029241e-05,
+      "loss": 1.4975,
+      "step": 6853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8620223999023438,
+      "learning_rate": 1.3737207602339181e-05,
+      "loss": 1.8154,
+      "step": 6854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3181004524230957,
+      "learning_rate": 1.3736293859649124e-05,
+      "loss": 1.6112,
+      "step": 6855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7044262886047363,
+      "learning_rate": 1.3735380116959066e-05,
+      "loss": 1.8003,
+      "step": 6856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.799069404602051,
+      "learning_rate": 1.3734466374269006e-05,
+      "loss": 1.6293,
+      "step": 6857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6533377170562744,
+      "learning_rate": 1.3733552631578949e-05,
+      "loss": 1.6106,
+      "step": 6858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4016623497009277,
+      "learning_rate": 1.3732638888888889e-05,
+      "loss": 2.3038,
+      "step": 6859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.382294178009033,
+      "learning_rate": 1.3731725146198831e-05,
+      "loss": 1.4141,
+      "step": 6860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.70401668548584,
+      "learning_rate": 1.3730811403508772e-05,
+      "loss": 1.3144,
+      "step": 6861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3324873447418213,
+      "learning_rate": 1.3729897660818714e-05,
+      "loss": 1.1371,
+      "step": 6862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.416271448135376,
+      "learning_rate": 1.3728983918128656e-05,
+      "loss": 1.4449,
+      "step": 6863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.336777925491333,
+      "learning_rate": 1.3728070175438597e-05,
+      "loss": 2.16,
+      "step": 6864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0915095806121826,
+      "learning_rate": 1.3727156432748539e-05,
+      "loss": 1.3925,
+      "step": 6865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5183794498443604,
+      "learning_rate": 1.372624269005848e-05,
+      "loss": 1.6241,
+      "step": 6866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6201629638671875,
+      "learning_rate": 1.3725328947368422e-05,
+      "loss": 1.5694,
+      "step": 6867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9611291885375977,
+      "learning_rate": 1.3724415204678362e-05,
+      "loss": 1.5227,
+      "step": 6868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.179899215698242,
+      "learning_rate": 1.3723501461988305e-05,
+      "loss": 1.8655,
+      "step": 6869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6127207279205322,
+      "learning_rate": 1.3722587719298247e-05,
+      "loss": 1.3952,
+      "step": 6870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.747121572494507,
+      "learning_rate": 1.3721673976608187e-05,
+      "loss": 1.515,
+      "step": 6871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.197197437286377,
+      "learning_rate": 1.372076023391813e-05,
+      "loss": 1.7019,
+      "step": 6872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2074966430664062,
+      "learning_rate": 1.371984649122807e-05,
+      "loss": 1.2888,
+      "step": 6873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5747504234313965,
+      "learning_rate": 1.3718932748538012e-05,
+      "loss": 1.435,
+      "step": 6874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.145078420639038,
+      "learning_rate": 1.3718019005847953e-05,
+      "loss": 1.739,
+      "step": 6875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5711567401885986,
+      "learning_rate": 1.3717105263157895e-05,
+      "loss": 1.5763,
+      "step": 6876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9655838012695312,
+      "learning_rate": 1.3716191520467836e-05,
+      "loss": 1.5032,
+      "step": 6877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.02750563621521,
+      "learning_rate": 1.3715277777777778e-05,
+      "loss": 2.0279,
+      "step": 6878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.482017755508423,
+      "learning_rate": 1.371436403508772e-05,
+      "loss": 1.6114,
+      "step": 6879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5683741569519043,
+      "learning_rate": 1.371345029239766e-05,
+      "loss": 1.2497,
+      "step": 6880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.824859142303467,
+      "learning_rate": 1.3712536549707603e-05,
+      "loss": 1.8388,
+      "step": 6881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8598506450653076,
+      "learning_rate": 1.3711622807017543e-05,
+      "loss": 1.7647,
+      "step": 6882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.665745973587036,
+      "learning_rate": 1.3710709064327487e-05,
+      "loss": 1.6662,
+      "step": 6883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3611128330230713,
+      "learning_rate": 1.370979532163743e-05,
+      "loss": 1.171,
+      "step": 6884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9686148166656494,
+      "learning_rate": 1.370888157894737e-05,
+      "loss": 1.4907,
+      "step": 6885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7915053367614746,
+      "learning_rate": 1.3707967836257312e-05,
+      "loss": 1.6569,
+      "step": 6886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.790268898010254,
+      "learning_rate": 1.3707054093567254e-05,
+      "loss": 1.54,
+      "step": 6887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.594082832336426,
+      "learning_rate": 1.3706140350877195e-05,
+      "loss": 1.5004,
+      "step": 6888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3849406242370605,
+      "learning_rate": 1.3705226608187137e-05,
+      "loss": 1.39,
+      "step": 6889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.512270927429199,
+      "learning_rate": 1.3704312865497078e-05,
+      "loss": 1.3486,
+      "step": 6890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.880706310272217,
+      "learning_rate": 1.370339912280702e-05,
+      "loss": 1.6027,
+      "step": 6891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.193636655807495,
+      "learning_rate": 1.370248538011696e-05,
+      "loss": 1.0771,
+      "step": 6892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8800690174102783,
+      "learning_rate": 1.3701571637426903e-05,
+      "loss": 1.7955,
+      "step": 6893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.15055251121521,
+      "learning_rate": 1.3700657894736845e-05,
+      "loss": 1.6563,
+      "step": 6894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3572330474853516,
+      "learning_rate": 1.3699744152046785e-05,
+      "loss": 1.4279,
+      "step": 6895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.82804274559021,
+      "learning_rate": 1.3698830409356728e-05,
+      "loss": 1.7979,
+      "step": 6896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.553452253341675,
+      "learning_rate": 1.3697916666666668e-05,
+      "loss": 1.6493,
+      "step": 6897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0874617099761963,
+      "learning_rate": 1.369700292397661e-05,
+      "loss": 2.1785,
+      "step": 6898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7265143394470215,
+      "learning_rate": 1.3696089181286551e-05,
+      "loss": 1.4021,
+      "step": 6899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2980902194976807,
+      "learning_rate": 1.3695175438596493e-05,
+      "loss": 1.3208,
+      "step": 6900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3559138774871826,
+      "learning_rate": 1.3694261695906435e-05,
+      "loss": 1.256,
+      "step": 6901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.848970651626587,
+      "learning_rate": 1.3693347953216376e-05,
+      "loss": 1.821,
+      "step": 6902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1844913959503174,
+      "learning_rate": 1.3692434210526318e-05,
+      "loss": 1.8453,
+      "step": 6903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4080259799957275,
+      "learning_rate": 1.3691520467836259e-05,
+      "loss": 1.3447,
+      "step": 6904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.09549617767334,
+      "learning_rate": 1.36906067251462e-05,
+      "loss": 1.3883,
+      "step": 6905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7748210430145264,
+      "learning_rate": 1.3689692982456141e-05,
+      "loss": 1.4237,
+      "step": 6906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.796407461166382,
+      "learning_rate": 1.3688779239766084e-05,
+      "loss": 1.7054,
+      "step": 6907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.251763343811035,
+      "learning_rate": 1.3687865497076026e-05,
+      "loss": 1.4454,
+      "step": 6908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.590575933456421,
+      "learning_rate": 1.3686951754385966e-05,
+      "loss": 1.6556,
+      "step": 6909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.883186101913452,
+      "learning_rate": 1.3686038011695909e-05,
+      "loss": 1.7666,
+      "step": 6910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.555344343185425,
+      "learning_rate": 1.3685124269005849e-05,
+      "loss": 1.332,
+      "step": 6911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.596463918685913,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 1.4494,
+      "step": 6912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8013625144958496,
+      "learning_rate": 1.3683296783625732e-05,
+      "loss": 1.7553,
+      "step": 6913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7450766563415527,
+      "learning_rate": 1.3682383040935674e-05,
+      "loss": 1.6887,
+      "step": 6914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.963374614715576,
+      "learning_rate": 1.3681469298245615e-05,
+      "loss": 1.6506,
+      "step": 6915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.786616086959839,
+      "learning_rate": 1.3680555555555557e-05,
+      "loss": 1.6131,
+      "step": 6916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1608455181121826,
+      "learning_rate": 1.3679641812865499e-05,
+      "loss": 1.8771,
+      "step": 6917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6996333599090576,
+      "learning_rate": 1.367872807017544e-05,
+      "loss": 1.4066,
+      "step": 6918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.286597490310669,
+      "learning_rate": 1.3677814327485382e-05,
+      "loss": 1.86,
+      "step": 6919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.106703042984009,
+      "learning_rate": 1.3676900584795322e-05,
+      "loss": 1.2948,
+      "step": 6920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8148531913757324,
+      "learning_rate": 1.3675986842105265e-05,
+      "loss": 1.523,
+      "step": 6921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3784542083740234,
+      "learning_rate": 1.3675073099415205e-05,
+      "loss": 1.3121,
+      "step": 6922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4827349185943604,
+      "learning_rate": 1.3674159356725147e-05,
+      "loss": 1.4994,
+      "step": 6923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5420806407928467,
+      "learning_rate": 1.367324561403509e-05,
+      "loss": 1.7891,
+      "step": 6924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3642513751983643,
+      "learning_rate": 1.367233187134503e-05,
+      "loss": 1.2219,
+      "step": 6925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3842403888702393,
+      "learning_rate": 1.3671418128654972e-05,
+      "loss": 1.2165,
+      "step": 6926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.634286880493164,
+      "learning_rate": 1.3670504385964913e-05,
+      "loss": 1.3023,
+      "step": 6927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.625933885574341,
+      "learning_rate": 1.3669590643274855e-05,
+      "loss": 1.4743,
+      "step": 6928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.279099702835083,
+      "learning_rate": 1.3668676900584796e-05,
+      "loss": 1.1474,
+      "step": 6929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2430553436279297,
+      "learning_rate": 1.3667763157894738e-05,
+      "loss": 1.7663,
+      "step": 6930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.820307970046997,
+      "learning_rate": 1.366684941520468e-05,
+      "loss": 1.5973,
+      "step": 6931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8250555992126465,
+      "learning_rate": 1.366593567251462e-05,
+      "loss": 1.8796,
+      "step": 6932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.385274648666382,
+      "learning_rate": 1.3665021929824563e-05,
+      "loss": 1.2971,
+      "step": 6933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.901336193084717,
+      "learning_rate": 1.3664108187134503e-05,
+      "loss": 1.6785,
+      "step": 6934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.857414484024048,
+      "learning_rate": 1.3663194444444445e-05,
+      "loss": 1.5616,
+      "step": 6935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3681046962738037,
+      "learning_rate": 1.3662280701754386e-05,
+      "loss": 1.4249,
+      "step": 6936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1732962131500244,
+      "learning_rate": 1.3661366959064328e-05,
+      "loss": 1.7559,
+      "step": 6937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.934558391571045,
+      "learning_rate": 1.366045321637427e-05,
+      "loss": 1.7365,
+      "step": 6938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.791466474533081,
+      "learning_rate": 1.3659539473684211e-05,
+      "loss": 1.4496,
+      "step": 6939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.701282501220703,
+      "learning_rate": 1.3658625730994153e-05,
+      "loss": 1.3467,
+      "step": 6940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.951768398284912,
+      "learning_rate": 1.3657711988304094e-05,
+      "loss": 1.2527,
+      "step": 6941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.753519296646118,
+      "learning_rate": 1.3656798245614036e-05,
+      "loss": 1.5613,
+      "step": 6942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.629753828048706,
+      "learning_rate": 1.3655884502923976e-05,
+      "loss": 1.5578,
+      "step": 6943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1057324409484863,
+      "learning_rate": 1.3654970760233919e-05,
+      "loss": 1.7927,
+      "step": 6944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.541609048843384,
+      "learning_rate": 1.3654057017543861e-05,
+      "loss": 1.9044,
+      "step": 6945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7221686840057373,
+      "learning_rate": 1.3653143274853801e-05,
+      "loss": 1.6391,
+      "step": 6946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.795020580291748,
+      "learning_rate": 1.3652229532163744e-05,
+      "loss": 1.5322,
+      "step": 6947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1852633953094482,
+      "learning_rate": 1.3651315789473684e-05,
+      "loss": 1.5459,
+      "step": 6948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.61502742767334,
+      "learning_rate": 1.3650402046783626e-05,
+      "loss": 1.3842,
+      "step": 6949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7228546142578125,
+      "learning_rate": 1.3649488304093567e-05,
+      "loss": 1.6503,
+      "step": 6950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.585345983505249,
+      "learning_rate": 1.3648574561403509e-05,
+      "loss": 1.6401,
+      "step": 6951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6237730979919434,
+      "learning_rate": 1.3647660818713451e-05,
+      "loss": 1.5332,
+      "step": 6952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2040419578552246,
+      "learning_rate": 1.3646747076023392e-05,
+      "loss": 1.1115,
+      "step": 6953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.750194787979126,
+      "learning_rate": 1.3645833333333334e-05,
+      "loss": 1.6846,
+      "step": 6954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7223901748657227,
+      "learning_rate": 1.3644919590643275e-05,
+      "loss": 1.4641,
+      "step": 6955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8100852966308594,
+      "learning_rate": 1.3644005847953217e-05,
+      "loss": 2.1023,
+      "step": 6956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.205705404281616,
+      "learning_rate": 1.3643092105263157e-05,
+      "loss": 1.3888,
+      "step": 6957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6006529331207275,
+      "learning_rate": 1.36421783625731e-05,
+      "loss": 1.4522,
+      "step": 6958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.481837272644043,
+      "learning_rate": 1.3641264619883042e-05,
+      "loss": 1.3889,
+      "step": 6959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.214006185531616,
+      "learning_rate": 1.3640350877192982e-05,
+      "loss": 1.7813,
+      "step": 6960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6398725509643555,
+      "learning_rate": 1.3639437134502925e-05,
+      "loss": 1.6096,
+      "step": 6961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6537649631500244,
+      "learning_rate": 1.3638523391812865e-05,
+      "loss": 2.1427,
+      "step": 6962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.808300733566284,
+      "learning_rate": 1.3637609649122807e-05,
+      "loss": 1.5728,
+      "step": 6963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7178359031677246,
+      "learning_rate": 1.3636695906432748e-05,
+      "loss": 1.8732,
+      "step": 6964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9507675170898438,
+      "learning_rate": 1.363578216374269e-05,
+      "loss": 1.8757,
+      "step": 6965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0485453605651855,
+      "learning_rate": 1.3634868421052632e-05,
+      "loss": 1.9062,
+      "step": 6966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5736186504364014,
+      "learning_rate": 1.3633954678362573e-05,
+      "loss": 1.7685,
+      "step": 6967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.295440196990967,
+      "learning_rate": 1.3633040935672515e-05,
+      "loss": 1.7704,
+      "step": 6968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4629931449890137,
+      "learning_rate": 1.3632127192982459e-05,
+      "loss": 1.1677,
+      "step": 6969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3960819244384766,
+      "learning_rate": 1.36312134502924e-05,
+      "loss": 1.4372,
+      "step": 6970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.894286870956421,
+      "learning_rate": 1.3630299707602342e-05,
+      "loss": 1.6996,
+      "step": 6971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5014407634735107,
+      "learning_rate": 1.3629385964912282e-05,
+      "loss": 1.4323,
+      "step": 6972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.256707191467285,
+      "learning_rate": 1.3628472222222224e-05,
+      "loss": 1.3965,
+      "step": 6973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5738775730133057,
+      "learning_rate": 1.3627558479532165e-05,
+      "loss": 1.1003,
+      "step": 6974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.007394313812256,
+      "learning_rate": 1.3626644736842107e-05,
+      "loss": 1.7734,
+      "step": 6975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.144076347351074,
+      "learning_rate": 1.362573099415205e-05,
+      "loss": 1.7845,
+      "step": 6976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.711521863937378,
+      "learning_rate": 1.362481725146199e-05,
+      "loss": 1.5817,
+      "step": 6977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7720961570739746,
+      "learning_rate": 1.3623903508771932e-05,
+      "loss": 1.794,
+      "step": 6978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6726269721984863,
+      "learning_rate": 1.3622989766081873e-05,
+      "loss": 1.3919,
+      "step": 6979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7141425609588623,
+      "learning_rate": 1.3622076023391815e-05,
+      "loss": 1.4387,
+      "step": 6980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.704746723175049,
+      "learning_rate": 1.3621162280701755e-05,
+      "loss": 1.4521,
+      "step": 6981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.662642240524292,
+      "learning_rate": 1.3620248538011698e-05,
+      "loss": 1.464,
+      "step": 6982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.693985939025879,
+      "learning_rate": 1.361933479532164e-05,
+      "loss": 1.5773,
+      "step": 6983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.855538845062256,
+      "learning_rate": 1.361842105263158e-05,
+      "loss": 1.4718,
+      "step": 6984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.752016544342041,
+      "learning_rate": 1.3617507309941523e-05,
+      "loss": 1.4948,
+      "step": 6985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9444832801818848,
+      "learning_rate": 1.3616593567251463e-05,
+      "loss": 1.3943,
+      "step": 6986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8482041358947754,
+      "learning_rate": 1.3615679824561405e-05,
+      "loss": 1.9234,
+      "step": 6987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4258170127868652,
+      "learning_rate": 1.3614766081871346e-05,
+      "loss": 1.3061,
+      "step": 6988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.730182647705078,
+      "learning_rate": 1.3613852339181288e-05,
+      "loss": 1.4355,
+      "step": 6989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.816906690597534,
+      "learning_rate": 1.361293859649123e-05,
+      "loss": 1.8063,
+      "step": 6990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.727513313293457,
+      "learning_rate": 1.3612024853801171e-05,
+      "loss": 1.6212,
+      "step": 6991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3107922077178955,
+      "learning_rate": 1.3611111111111113e-05,
+      "loss": 1.2711,
+      "step": 6992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.168626546859741,
+      "learning_rate": 1.3610197368421054e-05,
+      "loss": 1.3726,
+      "step": 6993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1157350540161133,
+      "learning_rate": 1.3609283625730996e-05,
+      "loss": 1.8146,
+      "step": 6994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.837942600250244,
+      "learning_rate": 1.3608369883040936e-05,
+      "loss": 1.3958,
+      "step": 6995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0406086444854736,
+      "learning_rate": 1.3607456140350879e-05,
+      "loss": 1.8671,
+      "step": 6996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.523745059967041,
+      "learning_rate": 1.360654239766082e-05,
+      "loss": 1.4062,
+      "step": 6997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.339362382888794,
+      "learning_rate": 1.3605628654970761e-05,
+      "loss": 1.382,
+      "step": 6998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3637216091156006,
+      "learning_rate": 1.3604714912280704e-05,
+      "loss": 1.5891,
+      "step": 6999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9001810550689697,
+      "learning_rate": 1.3603801169590644e-05,
+      "loss": 1.525,
+      "step": 7000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2228188514709473,
+      "learning_rate": 1.3602887426900586e-05,
+      "loss": 1.2122,
+      "step": 7001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.819352149963379,
+      "learning_rate": 1.3601973684210527e-05,
+      "loss": 1.7501,
+      "step": 7002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.290510654449463,
+      "learning_rate": 1.3601059941520469e-05,
+      "loss": 1.4923,
+      "step": 7003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.562480926513672,
+      "learning_rate": 1.3600146198830411e-05,
+      "loss": 1.6992,
+      "step": 7004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.036576509475708,
+      "learning_rate": 1.3599232456140352e-05,
+      "loss": 1.7818,
+      "step": 7005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.38167142868042,
+      "learning_rate": 1.3598318713450294e-05,
+      "loss": 1.249,
+      "step": 7006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9718613624572754,
+      "learning_rate": 1.3597404970760235e-05,
+      "loss": 1.7321,
+      "step": 7007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9777636528015137,
+      "learning_rate": 1.3596491228070177e-05,
+      "loss": 1.9084,
+      "step": 7008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7711520195007324,
+      "learning_rate": 1.3595577485380117e-05,
+      "loss": 1.8002,
+      "step": 7009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.718377113342285,
+      "learning_rate": 1.359466374269006e-05,
+      "loss": 1.6075,
+      "step": 7010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.661559581756592,
+      "learning_rate": 1.3593750000000002e-05,
+      "loss": 1.9384,
+      "step": 7011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0522847175598145,
+      "learning_rate": 1.3592836257309942e-05,
+      "loss": 1.824,
+      "step": 7012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3332784175872803,
+      "learning_rate": 1.3591922514619885e-05,
+      "loss": 1.2433,
+      "step": 7013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.127793312072754,
+      "learning_rate": 1.3591008771929825e-05,
+      "loss": 1.5418,
+      "step": 7014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.0405192375183105,
+      "learning_rate": 1.3590095029239767e-05,
+      "loss": 1.7035,
+      "step": 7015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2254042625427246,
+      "learning_rate": 1.3589181286549708e-05,
+      "loss": 1.5833,
+      "step": 7016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.318960666656494,
+      "learning_rate": 1.358826754385965e-05,
+      "loss": 1.4044,
+      "step": 7017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.000833511352539,
+      "learning_rate": 1.3587353801169592e-05,
+      "loss": 1.7465,
+      "step": 7018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6419310569763184,
+      "learning_rate": 1.3586440058479533e-05,
+      "loss": 1.3547,
+      "step": 7019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4682934284210205,
+      "learning_rate": 1.3585526315789475e-05,
+      "loss": 1.2754,
+      "step": 7020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9674408435821533,
+      "learning_rate": 1.3584612573099416e-05,
+      "loss": 1.8745,
+      "step": 7021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.69246768951416,
+      "learning_rate": 1.3583698830409358e-05,
+      "loss": 1.2658,
+      "step": 7022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2515478134155273,
+      "learning_rate": 1.3582785087719298e-05,
+      "loss": 2.0382,
+      "step": 7023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.988879919052124,
+      "learning_rate": 1.358187134502924e-05,
+      "loss": 1.5914,
+      "step": 7024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.126823902130127,
+      "learning_rate": 1.3580957602339183e-05,
+      "loss": 1.9863,
+      "step": 7025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.437666893005371,
+      "learning_rate": 1.3580043859649123e-05,
+      "loss": 2.0964,
+      "step": 7026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8934743404388428,
+      "learning_rate": 1.3579130116959065e-05,
+      "loss": 1.4684,
+      "step": 7027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.664886713027954,
+      "learning_rate": 1.3578216374269006e-05,
+      "loss": 1.5579,
+      "step": 7028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.723557233810425,
+      "learning_rate": 1.3577302631578948e-05,
+      "loss": 1.4056,
+      "step": 7029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9328787326812744,
+      "learning_rate": 1.3576388888888889e-05,
+      "loss": 1.98,
+      "step": 7030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5786588191986084,
+      "learning_rate": 1.3575475146198831e-05,
+      "loss": 1.6178,
+      "step": 7031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6897268295288086,
+      "learning_rate": 1.3574561403508773e-05,
+      "loss": 1.7394,
+      "step": 7032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.715622663497925,
+      "learning_rate": 1.3573647660818714e-05,
+      "loss": 1.6722,
+      "step": 7033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5404582023620605,
+      "learning_rate": 1.3572733918128656e-05,
+      "loss": 1.7992,
+      "step": 7034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5433876514434814,
+      "learning_rate": 1.3571820175438596e-05,
+      "loss": 1.3945,
+      "step": 7035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.446230173110962,
+      "learning_rate": 1.3570906432748539e-05,
+      "loss": 1.5683,
+      "step": 7036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3371775150299072,
+      "learning_rate": 1.356999269005848e-05,
+      "loss": 1.5149,
+      "step": 7037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4583370685577393,
+      "learning_rate": 1.3569078947368421e-05,
+      "loss": 1.3148,
+      "step": 7038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7625725269317627,
+      "learning_rate": 1.3568165204678364e-05,
+      "loss": 1.2946,
+      "step": 7039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.924320697784424,
+      "learning_rate": 1.3567251461988304e-05,
+      "loss": 1.628,
+      "step": 7040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.444265604019165,
+      "learning_rate": 1.3566337719298246e-05,
+      "loss": 1.3825,
+      "step": 7041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7673282623291016,
+      "learning_rate": 1.3565423976608187e-05,
+      "loss": 1.7782,
+      "step": 7042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.933469772338867,
+      "learning_rate": 1.3564510233918129e-05,
+      "loss": 1.4429,
+      "step": 7043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1370301246643066,
+      "learning_rate": 1.356359649122807e-05,
+      "loss": 1.9142,
+      "step": 7044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2937443256378174,
+      "learning_rate": 1.3562682748538012e-05,
+      "loss": 1.8233,
+      "step": 7045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.913942337036133,
+      "learning_rate": 1.3561769005847954e-05,
+      "loss": 1.6333,
+      "step": 7046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8897340297698975,
+      "learning_rate": 1.3560855263157895e-05,
+      "loss": 1.2754,
+      "step": 7047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7832508087158203,
+      "learning_rate": 1.3559941520467837e-05,
+      "loss": 1.6681,
+      "step": 7048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8224782943725586,
+      "learning_rate": 1.3559027777777777e-05,
+      "loss": 1.4877,
+      "step": 7049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8462255001068115,
+      "learning_rate": 1.355811403508772e-05,
+      "loss": 1.3923,
+      "step": 7050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5629000663757324,
+      "learning_rate": 1.355720029239766e-05,
+      "loss": 1.4519,
+      "step": 7051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.959212303161621,
+      "learning_rate": 1.3556286549707602e-05,
+      "loss": 1.4572,
+      "step": 7052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.538008689880371,
+      "learning_rate": 1.3555372807017545e-05,
+      "loss": 1.3577,
+      "step": 7053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.888551950454712,
+      "learning_rate": 1.3554459064327487e-05,
+      "loss": 1.3697,
+      "step": 7054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.286038637161255,
+      "learning_rate": 1.3553545321637429e-05,
+      "loss": 1.7519,
+      "step": 7055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.06600022315979,
+      "learning_rate": 1.3552631578947371e-05,
+      "loss": 1.897,
+      "step": 7056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6005516052246094,
+      "learning_rate": 1.3551717836257312e-05,
+      "loss": 1.5678,
+      "step": 7057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7499120235443115,
+      "learning_rate": 1.3550804093567254e-05,
+      "loss": 1.6274,
+      "step": 7058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8427417278289795,
+      "learning_rate": 1.3549890350877195e-05,
+      "loss": 1.437,
+      "step": 7059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.635312557220459,
+      "learning_rate": 1.3548976608187137e-05,
+      "loss": 1.5994,
+      "step": 7060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.889521837234497,
+      "learning_rate": 1.3548062865497077e-05,
+      "loss": 1.5847,
+      "step": 7061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.645719051361084,
+      "learning_rate": 1.354714912280702e-05,
+      "loss": 1.4797,
+      "step": 7062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1309030055999756,
+      "learning_rate": 1.3546235380116962e-05,
+      "loss": 1.7909,
+      "step": 7063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9080920219421387,
+      "learning_rate": 1.3545321637426902e-05,
+      "loss": 1.6453,
+      "step": 7064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4647176265716553,
+      "learning_rate": 1.3544407894736844e-05,
+      "loss": 1.2993,
+      "step": 7065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9146981239318848,
+      "learning_rate": 1.3543494152046785e-05,
+      "loss": 1.7034,
+      "step": 7066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.039647102355957,
+      "learning_rate": 1.3542580409356727e-05,
+      "loss": 1.761,
+      "step": 7067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0467991828918457,
+      "learning_rate": 1.3541666666666668e-05,
+      "loss": 1.811,
+      "step": 7068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6522915363311768,
+      "learning_rate": 1.354075292397661e-05,
+      "loss": 1.5929,
+      "step": 7069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.657583475112915,
+      "learning_rate": 1.3539839181286552e-05,
+      "loss": 1.7141,
+      "step": 7070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.665259838104248,
+      "learning_rate": 1.3538925438596493e-05,
+      "loss": 1.7554,
+      "step": 7071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9458353519439697,
+      "learning_rate": 1.3538011695906435e-05,
+      "loss": 2.1142,
+      "step": 7072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.349562406539917,
+      "learning_rate": 1.3537097953216375e-05,
+      "loss": 1.7909,
+      "step": 7073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0245747566223145,
+      "learning_rate": 1.3536184210526318e-05,
+      "loss": 1.5404,
+      "step": 7074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.192756175994873,
+      "learning_rate": 1.3535270467836258e-05,
+      "loss": 1.4034,
+      "step": 7075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0179405212402344,
+      "learning_rate": 1.35343567251462e-05,
+      "loss": 1.4409,
+      "step": 7076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5299787521362305,
+      "learning_rate": 1.3533442982456143e-05,
+      "loss": 1.6244,
+      "step": 7077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1314189434051514,
+      "learning_rate": 1.3532529239766083e-05,
+      "loss": 1.8856,
+      "step": 7078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.081772327423096,
+      "learning_rate": 1.3531615497076025e-05,
+      "loss": 2.0896,
+      "step": 7079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5372331142425537,
+      "learning_rate": 1.3530701754385966e-05,
+      "loss": 1.5154,
+      "step": 7080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.667343854904175,
+      "learning_rate": 1.3529788011695908e-05,
+      "loss": 1.4387,
+      "step": 7081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5914230346679688,
+      "learning_rate": 1.3528874269005849e-05,
+      "loss": 1.5759,
+      "step": 7082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.112645149230957,
+      "learning_rate": 1.3527960526315791e-05,
+      "loss": 1.9158,
+      "step": 7083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.756967067718506,
+      "learning_rate": 1.3527046783625733e-05,
+      "loss": 1.0165,
+      "step": 7084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.283897638320923,
+      "learning_rate": 1.3526133040935674e-05,
+      "loss": 1.8303,
+      "step": 7085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.813640594482422,
+      "learning_rate": 1.3525219298245616e-05,
+      "loss": 1.7549,
+      "step": 7086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.528203248977661,
+      "learning_rate": 1.3524305555555556e-05,
+      "loss": 1.7392,
+      "step": 7087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5961225032806396,
+      "learning_rate": 1.3523391812865499e-05,
+      "loss": 1.3713,
+      "step": 7088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2778093814849854,
+      "learning_rate": 1.3522478070175439e-05,
+      "loss": 1.4117,
+      "step": 7089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8434417247772217,
+      "learning_rate": 1.3521564327485381e-05,
+      "loss": 1.3268,
+      "step": 7090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8075473308563232,
+      "learning_rate": 1.3520650584795324e-05,
+      "loss": 1.6047,
+      "step": 7091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.714573621749878,
+      "learning_rate": 1.3519736842105264e-05,
+      "loss": 1.2702,
+      "step": 7092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.669494390487671,
+      "learning_rate": 1.3518823099415206e-05,
+      "loss": 1.7245,
+      "step": 7093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.845693588256836,
+      "learning_rate": 1.3517909356725147e-05,
+      "loss": 1.9949,
+      "step": 7094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4726181030273438,
+      "learning_rate": 1.3516995614035089e-05,
+      "loss": 1.8112,
+      "step": 7095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1273133754730225,
+      "learning_rate": 1.351608187134503e-05,
+      "loss": 1.9454,
+      "step": 7096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3312196731567383,
+      "learning_rate": 1.3515168128654972e-05,
+      "loss": 1.3121,
+      "step": 7097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.773812770843506,
+      "learning_rate": 1.3514254385964914e-05,
+      "loss": 1.5978,
+      "step": 7098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0039901733398438,
+      "learning_rate": 1.3513340643274855e-05,
+      "loss": 1.4818,
+      "step": 7099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.208252191543579,
+      "learning_rate": 1.3512426900584797e-05,
+      "loss": 1.3431,
+      "step": 7100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2811362743377686,
+      "learning_rate": 1.3511513157894737e-05,
+      "loss": 1.2393,
+      "step": 7101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4138479232788086,
+      "learning_rate": 1.351059941520468e-05,
+      "loss": 1.2644,
+      "step": 7102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0925681591033936,
+      "learning_rate": 1.350968567251462e-05,
+      "loss": 1.767,
+      "step": 7103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.514991283416748,
+      "learning_rate": 1.3508771929824562e-05,
+      "loss": 1.4621,
+      "step": 7104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7352819442749023,
+      "learning_rate": 1.3507858187134505e-05,
+      "loss": 1.4531,
+      "step": 7105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5538153648376465,
+      "learning_rate": 1.3506944444444445e-05,
+      "loss": 1.7553,
+      "step": 7106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.521428108215332,
+      "learning_rate": 1.3506030701754387e-05,
+      "loss": 1.6806,
+      "step": 7107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7588047981262207,
+      "learning_rate": 1.3505116959064328e-05,
+      "loss": 1.7483,
+      "step": 7108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2255911827087402,
+      "learning_rate": 1.350420321637427e-05,
+      "loss": 2.0079,
+      "step": 7109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4398679733276367,
+      "learning_rate": 1.350328947368421e-05,
+      "loss": 1.4944,
+      "step": 7110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.445737361907959,
+      "learning_rate": 1.3502375730994153e-05,
+      "loss": 1.4055,
+      "step": 7111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.953087329864502,
+      "learning_rate": 1.3501461988304095e-05,
+      "loss": 1.5757,
+      "step": 7112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4989607334136963,
+      "learning_rate": 1.3500548245614036e-05,
+      "loss": 1.4447,
+      "step": 7113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.884321928024292,
+      "learning_rate": 1.3499634502923978e-05,
+      "loss": 1.7766,
+      "step": 7114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.916013717651367,
+      "learning_rate": 1.3498720760233918e-05,
+      "loss": 1.3687,
+      "step": 7115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5870957374572754,
+      "learning_rate": 1.349780701754386e-05,
+      "loss": 1.7683,
+      "step": 7116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.969017267227173,
+      "learning_rate": 1.3496893274853801e-05,
+      "loss": 1.404,
+      "step": 7117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.442168951034546,
+      "learning_rate": 1.3495979532163743e-05,
+      "loss": 1.1191,
+      "step": 7118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.495973825454712,
+      "learning_rate": 1.3495065789473685e-05,
+      "loss": 1.9313,
+      "step": 7119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.676032781600952,
+      "learning_rate": 1.3494152046783626e-05,
+      "loss": 1.447,
+      "step": 7120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.158863067626953,
+      "learning_rate": 1.3493238304093568e-05,
+      "loss": 1.5847,
+      "step": 7121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.621788501739502,
+      "learning_rate": 1.3492324561403509e-05,
+      "loss": 1.5273,
+      "step": 7122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9655847549438477,
+      "learning_rate": 1.3491410818713451e-05,
+      "loss": 1.6841,
+      "step": 7123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.060056447982788,
+      "learning_rate": 1.3490497076023391e-05,
+      "loss": 1.8388,
+      "step": 7124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.907499313354492,
+      "learning_rate": 1.3489583333333334e-05,
+      "loss": 1.419,
+      "step": 7125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3678791522979736,
+      "learning_rate": 1.3488669590643276e-05,
+      "loss": 1.4701,
+      "step": 7126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.499298095703125,
+      "learning_rate": 1.3487755847953216e-05,
+      "loss": 1.6111,
+      "step": 7127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4970197677612305,
+      "learning_rate": 1.3486842105263159e-05,
+      "loss": 1.3911,
+      "step": 7128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.076756000518799,
+      "learning_rate": 1.34859283625731e-05,
+      "loss": 1.1758,
+      "step": 7129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.993680953979492,
+      "learning_rate": 1.3485014619883041e-05,
+      "loss": 1.7131,
+      "step": 7130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3916585445404053,
+      "learning_rate": 1.3484100877192982e-05,
+      "loss": 1.3901,
+      "step": 7131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9004247188568115,
+      "learning_rate": 1.3483187134502924e-05,
+      "loss": 1.7178,
+      "step": 7132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.496751308441162,
+      "learning_rate": 1.3482273391812866e-05,
+      "loss": 1.8567,
+      "step": 7133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3815927505493164,
+      "learning_rate": 1.3481359649122807e-05,
+      "loss": 1.203,
+      "step": 7134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.620612859725952,
+      "learning_rate": 1.3480445906432749e-05,
+      "loss": 1.4623,
+      "step": 7135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7818448543548584,
+      "learning_rate": 1.347953216374269e-05,
+      "loss": 1.5292,
+      "step": 7136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.529689073562622,
+      "learning_rate": 1.3478618421052632e-05,
+      "loss": 1.5415,
+      "step": 7137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6317367553710938,
+      "learning_rate": 1.3477704678362572e-05,
+      "loss": 1.6651,
+      "step": 7138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4975013732910156,
+      "learning_rate": 1.3476790935672515e-05,
+      "loss": 1.2643,
+      "step": 7139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7835941314697266,
+      "learning_rate": 1.3475877192982459e-05,
+      "loss": 1.6549,
+      "step": 7140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.743851661682129,
+      "learning_rate": 1.3474963450292399e-05,
+      "loss": 1.6987,
+      "step": 7141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8043742179870605,
+      "learning_rate": 1.3474049707602341e-05,
+      "loss": 1.5467,
+      "step": 7142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3111488819122314,
+      "learning_rate": 1.3473135964912284e-05,
+      "loss": 2.0574,
+      "step": 7143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7701611518859863,
+      "learning_rate": 1.3472222222222224e-05,
+      "loss": 1.4913,
+      "step": 7144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.125088930130005,
+      "learning_rate": 1.3471308479532166e-05,
+      "loss": 1.6501,
+      "step": 7145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9757730960845947,
+      "learning_rate": 1.3470394736842107e-05,
+      "loss": 1.6226,
+      "step": 7146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1139168739318848,
+      "learning_rate": 1.3469480994152049e-05,
+      "loss": 1.7948,
+      "step": 7147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.540360927581787,
+      "learning_rate": 1.346856725146199e-05,
+      "loss": 1.2921,
+      "step": 7148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.850022077560425,
+      "learning_rate": 1.3467653508771932e-05,
+      "loss": 1.8131,
+      "step": 7149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4316248893737793,
+      "learning_rate": 1.3466739766081874e-05,
+      "loss": 1.1852,
+      "step": 7150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1509792804718018,
+      "learning_rate": 1.3465826023391815e-05,
+      "loss": 2.0776,
+      "step": 7151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6335151195526123,
+      "learning_rate": 1.3464912280701757e-05,
+      "loss": 1.6868,
+      "step": 7152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3081631660461426,
+      "learning_rate": 1.3463998538011697e-05,
+      "loss": 1.6748,
+      "step": 7153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.944478988647461,
+      "learning_rate": 1.346308479532164e-05,
+      "loss": 1.7963,
+      "step": 7154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7040421962738037,
+      "learning_rate": 1.346217105263158e-05,
+      "loss": 1.4244,
+      "step": 7155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.824533224105835,
+      "learning_rate": 1.3461257309941522e-05,
+      "loss": 1.5806,
+      "step": 7156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.370218276977539,
+      "learning_rate": 1.3460343567251464e-05,
+      "loss": 1.3909,
+      "step": 7157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5175070762634277,
+      "learning_rate": 1.3459429824561405e-05,
+      "loss": 1.4007,
+      "step": 7158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.522095203399658,
+      "learning_rate": 1.3458516081871347e-05,
+      "loss": 1.6307,
+      "step": 7159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2889983654022217,
+      "learning_rate": 1.3457602339181288e-05,
+      "loss": 1.5369,
+      "step": 7160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.713676691055298,
+      "learning_rate": 1.345668859649123e-05,
+      "loss": 1.3502,
+      "step": 7161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5351827144622803,
+      "learning_rate": 1.345577485380117e-05,
+      "loss": 1.5533,
+      "step": 7162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.896463632583618,
+      "learning_rate": 1.3454861111111113e-05,
+      "loss": 1.7166,
+      "step": 7163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.245468854904175,
+      "learning_rate": 1.3453947368421055e-05,
+      "loss": 1.6545,
+      "step": 7164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.925501585006714,
+      "learning_rate": 1.3453033625730995e-05,
+      "loss": 1.7281,
+      "step": 7165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4529147148132324,
+      "learning_rate": 1.3452119883040938e-05,
+      "loss": 1.7316,
+      "step": 7166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4384779930114746,
+      "learning_rate": 1.3451206140350878e-05,
+      "loss": 1.6626,
+      "step": 7167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7437326908111572,
+      "learning_rate": 1.345029239766082e-05,
+      "loss": 1.7166,
+      "step": 7168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.433137893676758,
+      "learning_rate": 1.3449378654970761e-05,
+      "loss": 1.889,
+      "step": 7169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.06622576713562,
+      "learning_rate": 1.3448464912280703e-05,
+      "loss": 1.6439,
+      "step": 7170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3210272789001465,
+      "learning_rate": 1.3447551169590645e-05,
+      "loss": 1.6741,
+      "step": 7171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3865184783935547,
+      "learning_rate": 1.3446637426900586e-05,
+      "loss": 1.9552,
+      "step": 7172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.977743625640869,
+      "learning_rate": 1.3445723684210528e-05,
+      "loss": 1.7776,
+      "step": 7173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9540884494781494,
+      "learning_rate": 1.3444809941520469e-05,
+      "loss": 1.7228,
+      "step": 7174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4476823806762695,
+      "learning_rate": 1.3443896198830411e-05,
+      "loss": 1.5345,
+      "step": 7175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.655212640762329,
+      "learning_rate": 1.3442982456140351e-05,
+      "loss": 1.5644,
+      "step": 7176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6077380180358887,
+      "learning_rate": 1.3442068713450294e-05,
+      "loss": 1.2876,
+      "step": 7177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.10982608795166,
+      "learning_rate": 1.3441154970760236e-05,
+      "loss": 1.7316,
+      "step": 7178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9468605518341064,
+      "learning_rate": 1.3440241228070176e-05,
+      "loss": 1.5145,
+      "step": 7179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2442259788513184,
+      "learning_rate": 1.3439327485380119e-05,
+      "loss": 1.7359,
+      "step": 7180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.594085454940796,
+      "learning_rate": 1.3438413742690059e-05,
+      "loss": 1.5254,
+      "step": 7181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.668487787246704,
+      "learning_rate": 1.3437500000000001e-05,
+      "loss": 1.327,
+      "step": 7182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.641662836074829,
+      "learning_rate": 1.3436586257309942e-05,
+      "loss": 1.6554,
+      "step": 7183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4329702854156494,
+      "learning_rate": 1.3435672514619884e-05,
+      "loss": 1.372,
+      "step": 7184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8671059608459473,
+      "learning_rate": 1.3434758771929826e-05,
+      "loss": 1.7304,
+      "step": 7185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7059998512268066,
+      "learning_rate": 1.3433845029239767e-05,
+      "loss": 1.2951,
+      "step": 7186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.049729824066162,
+      "learning_rate": 1.3432931286549709e-05,
+      "loss": 1.6663,
+      "step": 7187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.941978693008423,
+      "learning_rate": 1.343201754385965e-05,
+      "loss": 1.7775,
+      "step": 7188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.221613645553589,
+      "learning_rate": 1.3431103801169592e-05,
+      "loss": 1.7265,
+      "step": 7189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.90566349029541,
+      "learning_rate": 1.3430190058479532e-05,
+      "loss": 1.6305,
+      "step": 7190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1525869369506836,
+      "learning_rate": 1.3429276315789475e-05,
+      "loss": 1.6004,
+      "step": 7191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2678494453430176,
+      "learning_rate": 1.3428362573099417e-05,
+      "loss": 1.86,
+      "step": 7192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7161855697631836,
+      "learning_rate": 1.3427448830409357e-05,
+      "loss": 2.2604,
+      "step": 7193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7951066493988037,
+      "learning_rate": 1.34265350877193e-05,
+      "loss": 1.5182,
+      "step": 7194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6375391483306885,
+      "learning_rate": 1.342562134502924e-05,
+      "loss": 1.7475,
+      "step": 7195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.347843885421753,
+      "learning_rate": 1.3424707602339182e-05,
+      "loss": 1.5926,
+      "step": 7196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0317745208740234,
+      "learning_rate": 1.3423793859649123e-05,
+      "loss": 1.7858,
+      "step": 7197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.738111734390259,
+      "learning_rate": 1.3422880116959065e-05,
+      "loss": 1.6498,
+      "step": 7198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8600401878356934,
+      "learning_rate": 1.3421966374269007e-05,
+      "loss": 1.6246,
+      "step": 7199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.574077844619751,
+      "learning_rate": 1.3421052631578948e-05,
+      "loss": 1.4895,
+      "step": 7200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0809874534606934,
+      "learning_rate": 1.342013888888889e-05,
+      "loss": 1.8246,
+      "step": 7201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3626811504364014,
+      "learning_rate": 1.341922514619883e-05,
+      "loss": 1.2762,
+      "step": 7202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8543484210968018,
+      "learning_rate": 1.3418311403508773e-05,
+      "loss": 1.8646,
+      "step": 7203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0404038429260254,
+      "learning_rate": 1.3417397660818713e-05,
+      "loss": 1.529,
+      "step": 7204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9856228828430176,
+      "learning_rate": 1.3416483918128656e-05,
+      "loss": 1.8069,
+      "step": 7205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3866302967071533,
+      "learning_rate": 1.3415570175438598e-05,
+      "loss": 1.4462,
+      "step": 7206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0631868839263916,
+      "learning_rate": 1.3414656432748538e-05,
+      "loss": 1.5546,
+      "step": 7207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6966066360473633,
+      "learning_rate": 1.341374269005848e-05,
+      "loss": 1.7286,
+      "step": 7208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9791433811187744,
+      "learning_rate": 1.3412828947368421e-05,
+      "loss": 1.8785,
+      "step": 7209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.878279685974121,
+      "learning_rate": 1.3411915204678363e-05,
+      "loss": 1.6297,
+      "step": 7210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.687354326248169,
+      "learning_rate": 1.3411001461988304e-05,
+      "loss": 1.5364,
+      "step": 7211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.610659599304199,
+      "learning_rate": 1.3410087719298246e-05,
+      "loss": 1.5629,
+      "step": 7212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.965411901473999,
+      "learning_rate": 1.3409173976608188e-05,
+      "loss": 1.1437,
+      "step": 7213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7369580268859863,
+      "learning_rate": 1.3408260233918129e-05,
+      "loss": 1.7053,
+      "step": 7214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5963664054870605,
+      "learning_rate": 1.3407346491228071e-05,
+      "loss": 1.5875,
+      "step": 7215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.770402193069458,
+      "learning_rate": 1.3406432748538011e-05,
+      "loss": 1.6583,
+      "step": 7216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.405651569366455,
+      "learning_rate": 1.3405519005847954e-05,
+      "loss": 1.3222,
+      "step": 7217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.818246364593506,
+      "learning_rate": 1.3404605263157894e-05,
+      "loss": 1.7014,
+      "step": 7218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2262377738952637,
+      "learning_rate": 1.3403691520467836e-05,
+      "loss": 1.8287,
+      "step": 7219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.727980136871338,
+      "learning_rate": 1.3402777777777779e-05,
+      "loss": 1.7962,
+      "step": 7220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.536222457885742,
+      "learning_rate": 1.340186403508772e-05,
+      "loss": 1.2855,
+      "step": 7221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1254022121429443,
+      "learning_rate": 1.3400950292397661e-05,
+      "loss": 1.8196,
+      "step": 7222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5021474361419678,
+      "learning_rate": 1.3400036549707602e-05,
+      "loss": 1.4065,
+      "step": 7223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8311374187469482,
+      "learning_rate": 1.3399122807017544e-05,
+      "loss": 1.6998,
+      "step": 7224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.970578193664551,
+      "learning_rate": 1.3398209064327488e-05,
+      "loss": 1.4678,
+      "step": 7225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5539679527282715,
+      "learning_rate": 1.3397295321637429e-05,
+      "loss": 1.3624,
+      "step": 7226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.25606369972229,
+      "learning_rate": 1.339638157894737e-05,
+      "loss": 1.2423,
+      "step": 7227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.986990451812744,
+      "learning_rate": 1.3395467836257311e-05,
+      "loss": 1.7915,
+      "step": 7228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.975602149963379,
+      "learning_rate": 1.3394554093567254e-05,
+      "loss": 1.3778,
+      "step": 7229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.57975172996521,
+      "learning_rate": 1.3393640350877196e-05,
+      "loss": 1.6804,
+      "step": 7230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7638602256774902,
+      "learning_rate": 1.3392726608187136e-05,
+      "loss": 1.607,
+      "step": 7231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1362314224243164,
+      "learning_rate": 1.3391812865497079e-05,
+      "loss": 1.8805,
+      "step": 7232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.127788782119751,
+      "learning_rate": 1.3390899122807019e-05,
+      "loss": 1.8337,
+      "step": 7233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6998589038848877,
+      "learning_rate": 1.3389985380116961e-05,
+      "loss": 1.5597,
+      "step": 7234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6918818950653076,
+      "learning_rate": 1.3389071637426902e-05,
+      "loss": 1.4784,
+      "step": 7235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.79529070854187,
+      "learning_rate": 1.3388157894736844e-05,
+      "loss": 1.6486,
+      "step": 7236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9101269245147705,
+      "learning_rate": 1.3387244152046786e-05,
+      "loss": 1.418,
+      "step": 7237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.562577724456787,
+      "learning_rate": 1.3386330409356727e-05,
+      "loss": 1.4685,
+      "step": 7238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0120840072631836,
+      "learning_rate": 1.3385416666666669e-05,
+      "loss": 1.5637,
+      "step": 7239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.144063711166382,
+      "learning_rate": 1.338450292397661e-05,
+      "loss": 1.7098,
+      "step": 7240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.556934118270874,
+      "learning_rate": 1.3383589181286552e-05,
+      "loss": 1.3508,
+      "step": 7241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5157318115234375,
+      "learning_rate": 1.3382675438596492e-05,
+      "loss": 1.4844,
+      "step": 7242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9360246658325195,
+      "learning_rate": 1.3381761695906435e-05,
+      "loss": 1.5911,
+      "step": 7243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8858871459960938,
+      "learning_rate": 1.3380847953216377e-05,
+      "loss": 1.5231,
+      "step": 7244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7288150787353516,
+      "learning_rate": 1.3379934210526317e-05,
+      "loss": 1.825,
+      "step": 7245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7770280838012695,
+      "learning_rate": 1.337902046783626e-05,
+      "loss": 1.7446,
+      "step": 7246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6335866451263428,
+      "learning_rate": 1.33781067251462e-05,
+      "loss": 1.7232,
+      "step": 7247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.794302225112915,
+      "learning_rate": 1.3377192982456142e-05,
+      "loss": 1.9964,
+      "step": 7248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9838707447052,
+      "learning_rate": 1.3376279239766083e-05,
+      "loss": 1.5368,
+      "step": 7249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.94075870513916,
+      "learning_rate": 1.3375365497076025e-05,
+      "loss": 1.8576,
+      "step": 7250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5901846885681152,
+      "learning_rate": 1.3374451754385967e-05,
+      "loss": 1.5749,
+      "step": 7251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5518078804016113,
+      "learning_rate": 1.3373538011695908e-05,
+      "loss": 1.6547,
+      "step": 7252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.804410219192505,
+      "learning_rate": 1.337262426900585e-05,
+      "loss": 1.5363,
+      "step": 7253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6823220252990723,
+      "learning_rate": 1.337171052631579e-05,
+      "loss": 1.6383,
+      "step": 7254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.012861967086792,
+      "learning_rate": 1.3370796783625733e-05,
+      "loss": 1.7242,
+      "step": 7255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4388396739959717,
+      "learning_rate": 1.3369883040935673e-05,
+      "loss": 2.0591,
+      "step": 7256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.482856035232544,
+      "learning_rate": 1.3368969298245615e-05,
+      "loss": 1.3072,
+      "step": 7257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.002906322479248,
+      "learning_rate": 1.3368055555555558e-05,
+      "loss": 1.7182,
+      "step": 7258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7685325145721436,
+      "learning_rate": 1.3367141812865498e-05,
+      "loss": 1.5452,
+      "step": 7259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.417029619216919,
+      "learning_rate": 1.336622807017544e-05,
+      "loss": 2.0224,
+      "step": 7260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7297093868255615,
+      "learning_rate": 1.3365314327485381e-05,
+      "loss": 1.523,
+      "step": 7261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.404510974884033,
+      "learning_rate": 1.3364400584795323e-05,
+      "loss": 1.4292,
+      "step": 7262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3224472999572754,
+      "learning_rate": 1.3363486842105264e-05,
+      "loss": 1.6341,
+      "step": 7263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.605706214904785,
+      "learning_rate": 1.3362573099415206e-05,
+      "loss": 1.449,
+      "step": 7264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8465938568115234,
+      "learning_rate": 1.3361659356725148e-05,
+      "loss": 1.495,
+      "step": 7265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.899970054626465,
+      "learning_rate": 1.3360745614035089e-05,
+      "loss": 1.6591,
+      "step": 7266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.430647850036621,
+      "learning_rate": 1.3359831871345031e-05,
+      "loss": 2.1496,
+      "step": 7267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7293925285339355,
+      "learning_rate": 1.3358918128654971e-05,
+      "loss": 1.5862,
+      "step": 7268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.640639305114746,
+      "learning_rate": 1.3358004385964914e-05,
+      "loss": 1.5943,
+      "step": 7269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.040719985961914,
+      "learning_rate": 1.3357090643274854e-05,
+      "loss": 1.756,
+      "step": 7270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6824686527252197,
+      "learning_rate": 1.3356176900584796e-05,
+      "loss": 1.2832,
+      "step": 7271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.612055540084839,
+      "learning_rate": 1.3355263157894739e-05,
+      "loss": 1.3771,
+      "step": 7272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2112865447998047,
+      "learning_rate": 1.3354349415204679e-05,
+      "loss": 1.3052,
+      "step": 7273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.002694845199585,
+      "learning_rate": 1.3353435672514621e-05,
+      "loss": 2.0457,
+      "step": 7274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.578232526779175,
+      "learning_rate": 1.3352521929824562e-05,
+      "loss": 1.7101,
+      "step": 7275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.562703847885132,
+      "learning_rate": 1.3351608187134504e-05,
+      "loss": 1.5134,
+      "step": 7276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.636953353881836,
+      "learning_rate": 1.3350694444444445e-05,
+      "loss": 1.7476,
+      "step": 7277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5336058139801025,
+      "learning_rate": 1.3349780701754387e-05,
+      "loss": 1.3169,
+      "step": 7278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6161508560180664,
+      "learning_rate": 1.3348866959064329e-05,
+      "loss": 1.5287,
+      "step": 7279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1207826137542725,
+      "learning_rate": 1.334795321637427e-05,
+      "loss": 1.4719,
+      "step": 7280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8759729862213135,
+      "learning_rate": 1.3347039473684212e-05,
+      "loss": 2.0602,
+      "step": 7281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.861452102661133,
+      "learning_rate": 1.3346125730994152e-05,
+      "loss": 1.5467,
+      "step": 7282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.629696846008301,
+      "learning_rate": 1.3345211988304095e-05,
+      "loss": 1.8103,
+      "step": 7283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.780815362930298,
+      "learning_rate": 1.3344298245614035e-05,
+      "loss": 1.2539,
+      "step": 7284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8468899726867676,
+      "learning_rate": 1.3343384502923977e-05,
+      "loss": 1.8575,
+      "step": 7285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.365894079208374,
+      "learning_rate": 1.334247076023392e-05,
+      "loss": 1.3227,
+      "step": 7286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.161914587020874,
+      "learning_rate": 1.334155701754386e-05,
+      "loss": 1.5215,
+      "step": 7287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.350565195083618,
+      "learning_rate": 1.3340643274853802e-05,
+      "loss": 1.9193,
+      "step": 7288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2004928588867188,
+      "learning_rate": 1.3339729532163743e-05,
+      "loss": 1.6126,
+      "step": 7289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5705628395080566,
+      "learning_rate": 1.3338815789473685e-05,
+      "loss": 1.2469,
+      "step": 7290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0843560695648193,
+      "learning_rate": 1.3337902046783626e-05,
+      "loss": 1.5896,
+      "step": 7291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.566866397857666,
+      "learning_rate": 1.3336988304093568e-05,
+      "loss": 1.5477,
+      "step": 7292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9803595542907715,
+      "learning_rate": 1.333607456140351e-05,
+      "loss": 1.3063,
+      "step": 7293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6116485595703125,
+      "learning_rate": 1.333516081871345e-05,
+      "loss": 1.5644,
+      "step": 7294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1097702980041504,
+      "learning_rate": 1.3334247076023393e-05,
+      "loss": 2.0756,
+      "step": 7295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3222591876983643,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.1191,
+      "step": 7296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.523000478744507,
+      "learning_rate": 1.3332419590643276e-05,
+      "loss": 1.8234,
+      "step": 7297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.070355176925659,
+      "learning_rate": 1.3331505847953216e-05,
+      "loss": 1.2064,
+      "step": 7298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4949238300323486,
+      "learning_rate": 1.3330592105263158e-05,
+      "loss": 1.5849,
+      "step": 7299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4147465229034424,
+      "learning_rate": 1.33296783625731e-05,
+      "loss": 1.3858,
+      "step": 7300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1394379138946533,
+      "learning_rate": 1.3328764619883041e-05,
+      "loss": 1.6414,
+      "step": 7301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0031471252441406,
+      "learning_rate": 1.3327850877192983e-05,
+      "loss": 1.4823,
+      "step": 7302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1974005699157715,
+      "learning_rate": 1.3326937134502924e-05,
+      "loss": 1.1678,
+      "step": 7303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7727627754211426,
+      "learning_rate": 1.3326023391812866e-05,
+      "loss": 1.7387,
+      "step": 7304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9587831497192383,
+      "learning_rate": 1.3325109649122807e-05,
+      "loss": 1.8112,
+      "step": 7305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6487655639648438,
+      "learning_rate": 1.3324195906432749e-05,
+      "loss": 1.5438,
+      "step": 7306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9976284503936768,
+      "learning_rate": 1.3323282163742691e-05,
+      "loss": 1.7034,
+      "step": 7307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8145558834075928,
+      "learning_rate": 1.3322368421052631e-05,
+      "loss": 1.7598,
+      "step": 7308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3352370262145996,
+      "learning_rate": 1.3321454678362574e-05,
+      "loss": 1.2778,
+      "step": 7309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7214694023132324,
+      "learning_rate": 1.3320540935672514e-05,
+      "loss": 1.2432,
+      "step": 7310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.878653049468994,
+      "learning_rate": 1.3319627192982458e-05,
+      "loss": 1.6179,
+      "step": 7311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.702284097671509,
+      "learning_rate": 1.33187134502924e-05,
+      "loss": 1.553,
+      "step": 7312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5788631439208984,
+      "learning_rate": 1.3317799707602341e-05,
+      "loss": 1.5521,
+      "step": 7313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2717230319976807,
+      "learning_rate": 1.3316885964912283e-05,
+      "loss": 1.6112,
+      "step": 7314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2569968700408936,
+      "learning_rate": 1.3315972222222224e-05,
+      "loss": 1.1759,
+      "step": 7315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3373467922210693,
+      "learning_rate": 1.3315058479532166e-05,
+      "loss": 1.7174,
+      "step": 7316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4285690784454346,
+      "learning_rate": 1.3314144736842108e-05,
+      "loss": 1.3954,
+      "step": 7317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6672379970550537,
+      "learning_rate": 1.3313230994152049e-05,
+      "loss": 1.4912,
+      "step": 7318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8547420501708984,
+      "learning_rate": 1.331231725146199e-05,
+      "loss": 1.429,
+      "step": 7319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2573139667510986,
+      "learning_rate": 1.3311403508771931e-05,
+      "loss": 1.2347,
+      "step": 7320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.251683473587036,
+      "learning_rate": 1.3310489766081874e-05,
+      "loss": 1.2328,
+      "step": 7321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.21793794631958,
+      "learning_rate": 1.3309576023391814e-05,
+      "loss": 1.2147,
+      "step": 7322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4739201068878174,
+      "learning_rate": 1.3308662280701756e-05,
+      "loss": 1.3972,
+      "step": 7323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.856907606124878,
+      "learning_rate": 1.3307748538011699e-05,
+      "loss": 1.7598,
+      "step": 7324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2482731342315674,
+      "learning_rate": 1.3306834795321639e-05,
+      "loss": 1.3303,
+      "step": 7325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.548443555831909,
+      "learning_rate": 1.3305921052631581e-05,
+      "loss": 1.3568,
+      "step": 7326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5384576320648193,
+      "learning_rate": 1.3305007309941522e-05,
+      "loss": 1.3777,
+      "step": 7327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5632283687591553,
+      "learning_rate": 1.3304093567251464e-05,
+      "loss": 1.5959,
+      "step": 7328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6393721103668213,
+      "learning_rate": 1.3303179824561405e-05,
+      "loss": 1.3615,
+      "step": 7329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2380404472351074,
+      "learning_rate": 1.3302266081871347e-05,
+      "loss": 1.6534,
+      "step": 7330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1410675048828125,
+      "learning_rate": 1.3301352339181289e-05,
+      "loss": 1.598,
+      "step": 7331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7173924446105957,
+      "learning_rate": 1.330043859649123e-05,
+      "loss": 1.3074,
+      "step": 7332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.060896158218384,
+      "learning_rate": 1.3299524853801172e-05,
+      "loss": 1.6813,
+      "step": 7333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5629589557647705,
+      "learning_rate": 1.3298611111111112e-05,
+      "loss": 1.6611,
+      "step": 7334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.433089256286621,
+      "learning_rate": 1.3297697368421055e-05,
+      "loss": 1.387,
+      "step": 7335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8024253845214844,
+      "learning_rate": 1.3296783625730995e-05,
+      "loss": 1.5622,
+      "step": 7336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5486862659454346,
+      "learning_rate": 1.3295869883040937e-05,
+      "loss": 1.2312,
+      "step": 7337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.694453001022339,
+      "learning_rate": 1.329495614035088e-05,
+      "loss": 1.6139,
+      "step": 7338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.330380439758301,
+      "learning_rate": 1.329404239766082e-05,
+      "loss": 1.3621,
+      "step": 7339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.255774974822998,
+      "learning_rate": 1.3293128654970762e-05,
+      "loss": 1.3649,
+      "step": 7340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5412096977233887,
+      "learning_rate": 1.3292214912280703e-05,
+      "loss": 1.4619,
+      "step": 7341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0810439586639404,
+      "learning_rate": 1.3291301169590645e-05,
+      "loss": 1.5567,
+      "step": 7342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.902184009552002,
+      "learning_rate": 1.3290387426900586e-05,
+      "loss": 1.9387,
+      "step": 7343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4091808795928955,
+      "learning_rate": 1.3289473684210528e-05,
+      "loss": 1.5429,
+      "step": 7344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.005871534347534,
+      "learning_rate": 1.328855994152047e-05,
+      "loss": 1.3691,
+      "step": 7345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.655944347381592,
+      "learning_rate": 1.328764619883041e-05,
+      "loss": 1.2529,
+      "step": 7346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8942413330078125,
+      "learning_rate": 1.3286732456140353e-05,
+      "loss": 1.4154,
+      "step": 7347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.451422929763794,
+      "learning_rate": 1.3285818713450293e-05,
+      "loss": 1.5086,
+      "step": 7348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5570507049560547,
+      "learning_rate": 1.3284904970760235e-05,
+      "loss": 1.5201,
+      "step": 7349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0173420906066895,
+      "learning_rate": 1.3283991228070176e-05,
+      "loss": 1.602,
+      "step": 7350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.98885440826416,
+      "learning_rate": 1.3283077485380118e-05,
+      "loss": 1.7072,
+      "step": 7351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.031817436218262,
+      "learning_rate": 1.328216374269006e-05,
+      "loss": 1.9511,
+      "step": 7352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3569374084472656,
+      "learning_rate": 1.3281250000000001e-05,
+      "loss": 1.893,
+      "step": 7353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9661877155303955,
+      "learning_rate": 1.3280336257309943e-05,
+      "loss": 1.7361,
+      "step": 7354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.61279034614563,
+      "learning_rate": 1.3279422514619884e-05,
+      "loss": 1.3844,
+      "step": 7355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4339654445648193,
+      "learning_rate": 1.3278508771929826e-05,
+      "loss": 1.3445,
+      "step": 7356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9629132747650146,
+      "learning_rate": 1.3277595029239766e-05,
+      "loss": 1.7914,
+      "step": 7357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2869808673858643,
+      "learning_rate": 1.3276681286549709e-05,
+      "loss": 1.6387,
+      "step": 7358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.672407388687134,
+      "learning_rate": 1.3275767543859651e-05,
+      "loss": 1.5523,
+      "step": 7359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1171700954437256,
+      "learning_rate": 1.3274853801169591e-05,
+      "loss": 1.8603,
+      "step": 7360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8346781730651855,
+      "learning_rate": 1.3273940058479534e-05,
+      "loss": 1.4642,
+      "step": 7361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7815234661102295,
+      "learning_rate": 1.3273026315789474e-05,
+      "loss": 1.5261,
+      "step": 7362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.724586248397827,
+      "learning_rate": 1.3272112573099416e-05,
+      "loss": 1.2355,
+      "step": 7363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6491141319274902,
+      "learning_rate": 1.3271198830409357e-05,
+      "loss": 1.5973,
+      "step": 7364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8178319931030273,
+      "learning_rate": 1.3270285087719299e-05,
+      "loss": 1.6171,
+      "step": 7365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.685699939727783,
+      "learning_rate": 1.326937134502924e-05,
+      "loss": 1.6039,
+      "step": 7366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.581364393234253,
+      "learning_rate": 1.3268457602339182e-05,
+      "loss": 1.8614,
+      "step": 7367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.819481372833252,
+      "learning_rate": 1.3267543859649124e-05,
+      "loss": 1.5796,
+      "step": 7368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.168811321258545,
+      "learning_rate": 1.3266630116959065e-05,
+      "loss": 2.0505,
+      "step": 7369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4263079166412354,
+      "learning_rate": 1.3265716374269007e-05,
+      "loss": 1.3086,
+      "step": 7370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5497212409973145,
+      "learning_rate": 1.3264802631578947e-05,
+      "loss": 1.5675,
+      "step": 7371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7188186645507812,
+      "learning_rate": 1.326388888888889e-05,
+      "loss": 1.5324,
+      "step": 7372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6839401721954346,
+      "learning_rate": 1.326297514619883e-05,
+      "loss": 1.7443,
+      "step": 7373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7122786045074463,
+      "learning_rate": 1.3262061403508772e-05,
+      "loss": 1.4901,
+      "step": 7374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.032392978668213,
+      "learning_rate": 1.3261147660818715e-05,
+      "loss": 1.8584,
+      "step": 7375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4068055152893066,
+      "learning_rate": 1.3260233918128655e-05,
+      "loss": 1.0307,
+      "step": 7376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.180506706237793,
+      "learning_rate": 1.3259320175438597e-05,
+      "loss": 1.7067,
+      "step": 7377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4691760540008545,
+      "learning_rate": 1.3258406432748538e-05,
+      "loss": 1.425,
+      "step": 7378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1919772624969482,
+      "learning_rate": 1.325749269005848e-05,
+      "loss": 1.8108,
+      "step": 7379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7047030925750732,
+      "learning_rate": 1.325657894736842e-05,
+      "loss": 1.5278,
+      "step": 7380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3283920288085938,
+      "learning_rate": 1.3255665204678363e-05,
+      "loss": 1.3655,
+      "step": 7381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.82409930229187,
+      "learning_rate": 1.3254751461988305e-05,
+      "loss": 1.7524,
+      "step": 7382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0099263191223145,
+      "learning_rate": 1.3253837719298246e-05,
+      "loss": 1.8075,
+      "step": 7383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8313770294189453,
+      "learning_rate": 1.3252923976608188e-05,
+      "loss": 1.6762,
+      "step": 7384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.958136558532715,
+      "learning_rate": 1.3252010233918128e-05,
+      "loss": 1.6929,
+      "step": 7385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9644107818603516,
+      "learning_rate": 1.325109649122807e-05,
+      "loss": 1.7109,
+      "step": 7386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6905734539031982,
+      "learning_rate": 1.3250182748538011e-05,
+      "loss": 1.5718,
+      "step": 7387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.647716760635376,
+      "learning_rate": 1.3249269005847953e-05,
+      "loss": 1.3884,
+      "step": 7388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.521646022796631,
+      "learning_rate": 1.3248355263157896e-05,
+      "loss": 1.4088,
+      "step": 7389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1324586868286133,
+      "learning_rate": 1.3247441520467836e-05,
+      "loss": 1.2829,
+      "step": 7390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.290647506713867,
+      "learning_rate": 1.3246527777777778e-05,
+      "loss": 1.2882,
+      "step": 7391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.52830171585083,
+      "learning_rate": 1.3245614035087719e-05,
+      "loss": 1.5353,
+      "step": 7392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9218811988830566,
+      "learning_rate": 1.3244700292397661e-05,
+      "loss": 2.0064,
+      "step": 7393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.342543125152588,
+      "learning_rate": 1.3243786549707602e-05,
+      "loss": 1.2628,
+      "step": 7394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0310914516448975,
+      "learning_rate": 1.3242872807017544e-05,
+      "loss": 1.6666,
+      "step": 7395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8118035793304443,
+      "learning_rate": 1.3241959064327488e-05,
+      "loss": 1.4721,
+      "step": 7396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9417383670806885,
+      "learning_rate": 1.324104532163743e-05,
+      "loss": 1.6724,
+      "step": 7397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6946754455566406,
+      "learning_rate": 1.324013157894737e-05,
+      "loss": 1.4713,
+      "step": 7398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0855305194854736,
+      "learning_rate": 1.3239217836257313e-05,
+      "loss": 1.3707,
+      "step": 7399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8120012283325195,
+      "learning_rate": 1.3238304093567253e-05,
+      "loss": 1.4584,
+      "step": 7400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3857884407043457,
+      "learning_rate": 1.3237390350877195e-05,
+      "loss": 1.6987,
+      "step": 7401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5128772258758545,
+      "learning_rate": 1.3236476608187136e-05,
+      "loss": 1.514,
+      "step": 7402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9959025382995605,
+      "learning_rate": 1.3235562865497078e-05,
+      "loss": 1.338,
+      "step": 7403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6969404220581055,
+      "learning_rate": 1.323464912280702e-05,
+      "loss": 1.3683,
+      "step": 7404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8732101917266846,
+      "learning_rate": 1.3233735380116961e-05,
+      "loss": 1.6793,
+      "step": 7405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.787325382232666,
+      "learning_rate": 1.3232821637426903e-05,
+      "loss": 1.4058,
+      "step": 7406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7728023529052734,
+      "learning_rate": 1.3231907894736844e-05,
+      "loss": 1.5817,
+      "step": 7407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4601876735687256,
+      "learning_rate": 1.3230994152046786e-05,
+      "loss": 1.347,
+      "step": 7408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2770395278930664,
+      "learning_rate": 1.3230080409356726e-05,
+      "loss": 1.4619,
+      "step": 7409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.507550001144409,
+      "learning_rate": 1.3229166666666669e-05,
+      "loss": 1.4726,
+      "step": 7410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9143364429473877,
+      "learning_rate": 1.3228252923976609e-05,
+      "loss": 1.6363,
+      "step": 7411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9306752681732178,
+      "learning_rate": 1.3227339181286551e-05,
+      "loss": 1.6121,
+      "step": 7412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.221947193145752,
+      "learning_rate": 1.3226425438596494e-05,
+      "loss": 1.4222,
+      "step": 7413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6698453426361084,
+      "learning_rate": 1.3225511695906434e-05,
+      "loss": 1.69,
+      "step": 7414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1267271041870117,
+      "learning_rate": 1.3224597953216376e-05,
+      "loss": 1.1522,
+      "step": 7415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6300947666168213,
+      "learning_rate": 1.3223684210526317e-05,
+      "loss": 1.2858,
+      "step": 7416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7821617126464844,
+      "learning_rate": 1.3222770467836259e-05,
+      "loss": 1.6106,
+      "step": 7417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7877752780914307,
+      "learning_rate": 1.32218567251462e-05,
+      "loss": 1.7449,
+      "step": 7418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6628577709198,
+      "learning_rate": 1.3220942982456142e-05,
+      "loss": 1.6292,
+      "step": 7419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3470911979675293,
+      "learning_rate": 1.3220029239766084e-05,
+      "loss": 2.0182,
+      "step": 7420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.794325828552246,
+      "learning_rate": 1.3219115497076025e-05,
+      "loss": 1.491,
+      "step": 7421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.802269220352173,
+      "learning_rate": 1.3218201754385967e-05,
+      "loss": 1.7237,
+      "step": 7422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8884811401367188,
+      "learning_rate": 1.3217288011695907e-05,
+      "loss": 1.7356,
+      "step": 7423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7049875259399414,
+      "learning_rate": 1.321637426900585e-05,
+      "loss": 1.5174,
+      "step": 7424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.042642593383789,
+      "learning_rate": 1.321546052631579e-05,
+      "loss": 1.7853,
+      "step": 7425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.998415231704712,
+      "learning_rate": 1.3214546783625732e-05,
+      "loss": 1.6711,
+      "step": 7426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.918598175048828,
+      "learning_rate": 1.3213633040935675e-05,
+      "loss": 1.7158,
+      "step": 7427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9858145713806152,
+      "learning_rate": 1.3212719298245615e-05,
+      "loss": 1.8005,
+      "step": 7428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.463015079498291,
+      "learning_rate": 1.3211805555555557e-05,
+      "loss": 1.4366,
+      "step": 7429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.07146954536438,
+      "learning_rate": 1.3210891812865498e-05,
+      "loss": 1.7121,
+      "step": 7430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6749982833862305,
+      "learning_rate": 1.320997807017544e-05,
+      "loss": 1.5105,
+      "step": 7431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.82189679145813,
+      "learning_rate": 1.320906432748538e-05,
+      "loss": 1.55,
+      "step": 7432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4429993629455566,
+      "learning_rate": 1.3208150584795323e-05,
+      "loss": 1.3338,
+      "step": 7433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2899608612060547,
+      "learning_rate": 1.3207236842105265e-05,
+      "loss": 1.8551,
+      "step": 7434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.622770309448242,
+      "learning_rate": 1.3206323099415206e-05,
+      "loss": 1.476,
+      "step": 7435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4037907123565674,
+      "learning_rate": 1.3205409356725148e-05,
+      "loss": 1.2754,
+      "step": 7436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.507402181625366,
+      "learning_rate": 1.3204495614035088e-05,
+      "loss": 1.7043,
+      "step": 7437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.115473985671997,
+      "learning_rate": 1.320358187134503e-05,
+      "loss": 1.7512,
+      "step": 7438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4264042377471924,
+      "learning_rate": 1.3202668128654971e-05,
+      "loss": 2.2155,
+      "step": 7439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4902472496032715,
+      "learning_rate": 1.3201754385964913e-05,
+      "loss": 1.5402,
+      "step": 7440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6359686851501465,
+      "learning_rate": 1.3200840643274855e-05,
+      "loss": 1.185,
+      "step": 7441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.283764600753784,
+      "learning_rate": 1.3199926900584796e-05,
+      "loss": 1.543,
+      "step": 7442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6757826805114746,
+      "learning_rate": 1.3199013157894738e-05,
+      "loss": 1.5386,
+      "step": 7443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.320169687271118,
+      "learning_rate": 1.3198099415204679e-05,
+      "loss": 1.3439,
+      "step": 7444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.878943920135498,
+      "learning_rate": 1.3197185672514621e-05,
+      "loss": 1.8032,
+      "step": 7445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.973975419998169,
+      "learning_rate": 1.3196271929824561e-05,
+      "loss": 1.6498,
+      "step": 7446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.775691270828247,
+      "learning_rate": 1.3195358187134504e-05,
+      "loss": 1.4685,
+      "step": 7447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4879250526428223,
+      "learning_rate": 1.3194444444444446e-05,
+      "loss": 1.5193,
+      "step": 7448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.480525016784668,
+      "learning_rate": 1.3193530701754386e-05,
+      "loss": 1.3942,
+      "step": 7449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7249419689178467,
+      "learning_rate": 1.3192616959064329e-05,
+      "loss": 1.696,
+      "step": 7450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.915666103363037,
+      "learning_rate": 1.319170321637427e-05,
+      "loss": 2.0439,
+      "step": 7451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5125348567962646,
+      "learning_rate": 1.3190789473684211e-05,
+      "loss": 1.5176,
+      "step": 7452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.73323655128479,
+      "learning_rate": 1.3189875730994152e-05,
+      "loss": 1.6516,
+      "step": 7453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9381368160247803,
+      "learning_rate": 1.3188961988304094e-05,
+      "loss": 1.744,
+      "step": 7454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0982987880706787,
+      "learning_rate": 1.3188048245614036e-05,
+      "loss": 1.8575,
+      "step": 7455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.469006299972534,
+      "learning_rate": 1.3187134502923977e-05,
+      "loss": 1.5878,
+      "step": 7456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.078296661376953,
+      "learning_rate": 1.3186220760233919e-05,
+      "loss": 1.4051,
+      "step": 7457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.575103759765625,
+      "learning_rate": 1.318530701754386e-05,
+      "loss": 1.5815,
+      "step": 7458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9280755519866943,
+      "learning_rate": 1.3184393274853802e-05,
+      "loss": 1.5919,
+      "step": 7459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2290103435516357,
+      "learning_rate": 1.3183479532163742e-05,
+      "loss": 1.3427,
+      "step": 7460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.254648447036743,
+      "learning_rate": 1.3182565789473685e-05,
+      "loss": 1.52,
+      "step": 7461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.627131223678589,
+      "learning_rate": 1.3181652046783627e-05,
+      "loss": 1.4583,
+      "step": 7462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3927931785583496,
+      "learning_rate": 1.3180738304093567e-05,
+      "loss": 1.3616,
+      "step": 7463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3435277938842773,
+      "learning_rate": 1.317982456140351e-05,
+      "loss": 1.2836,
+      "step": 7464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.810774326324463,
+      "learning_rate": 1.317891081871345e-05,
+      "loss": 1.8535,
+      "step": 7465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2936007976531982,
+      "learning_rate": 1.3177997076023392e-05,
+      "loss": 1.6894,
+      "step": 7466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.515669107437134,
+      "learning_rate": 1.3177083333333333e-05,
+      "loss": 1.6026,
+      "step": 7467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6036713123321533,
+      "learning_rate": 1.3176169590643275e-05,
+      "loss": 1.4596,
+      "step": 7468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6362762451171875,
+      "learning_rate": 1.3175255847953217e-05,
+      "loss": 1.7383,
+      "step": 7469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2806437015533447,
+      "learning_rate": 1.3174342105263158e-05,
+      "loss": 1.2623,
+      "step": 7470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.857222318649292,
+      "learning_rate": 1.31734283625731e-05,
+      "loss": 1.5588,
+      "step": 7471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.209399461746216,
+      "learning_rate": 1.317251461988304e-05,
+      "loss": 1.1973,
+      "step": 7472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2435810565948486,
+      "learning_rate": 1.3171600877192983e-05,
+      "loss": 2.2232,
+      "step": 7473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.655555009841919,
+      "learning_rate": 1.3170687134502923e-05,
+      "loss": 1.5014,
+      "step": 7474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4341163635253906,
+      "learning_rate": 1.3169773391812866e-05,
+      "loss": 1.4717,
+      "step": 7475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.358368396759033,
+      "learning_rate": 1.3168859649122808e-05,
+      "loss": 1.2893,
+      "step": 7476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.222130060195923,
+      "learning_rate": 1.3167945906432748e-05,
+      "loss": 1.2484,
+      "step": 7477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7359304428100586,
+      "learning_rate": 1.316703216374269e-05,
+      "loss": 1.4592,
+      "step": 7478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9808802604675293,
+      "learning_rate": 1.3166118421052631e-05,
+      "loss": 1.4133,
+      "step": 7479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4296936988830566,
+      "learning_rate": 1.3165204678362573e-05,
+      "loss": 1.3679,
+      "step": 7480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9773683547973633,
+      "learning_rate": 1.3164290935672514e-05,
+      "loss": 1.3634,
+      "step": 7481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7845067977905273,
+      "learning_rate": 1.3163377192982458e-05,
+      "loss": 1.6555,
+      "step": 7482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0260682106018066,
+      "learning_rate": 1.31624634502924e-05,
+      "loss": 1.5021,
+      "step": 7483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.699028968811035,
+      "learning_rate": 1.316154970760234e-05,
+      "loss": 1.4957,
+      "step": 7484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5353217124938965,
+      "learning_rate": 1.3160635964912283e-05,
+      "loss": 1.5723,
+      "step": 7485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.803952932357788,
+      "learning_rate": 1.3159722222222225e-05,
+      "loss": 1.469,
+      "step": 7486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.523667573928833,
+      "learning_rate": 1.3158808479532165e-05,
+      "loss": 1.4436,
+      "step": 7487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6007039546966553,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 1.2275,
+      "step": 7488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6830222606658936,
+      "learning_rate": 1.3156980994152048e-05,
+      "loss": 1.606,
+      "step": 7489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.696786403656006,
+      "learning_rate": 1.315606725146199e-05,
+      "loss": 1.5438,
+      "step": 7490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.983767509460449,
+      "learning_rate": 1.3155153508771931e-05,
+      "loss": 1.7937,
+      "step": 7491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.662984848022461,
+      "learning_rate": 1.3154239766081873e-05,
+      "loss": 1.308,
+      "step": 7492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1246745586395264,
+      "learning_rate": 1.3153326023391815e-05,
+      "loss": 1.7914,
+      "step": 7493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.456368923187256,
+      "learning_rate": 1.3152412280701756e-05,
+      "loss": 1.4625,
+      "step": 7494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2862985134124756,
+      "learning_rate": 1.3151498538011698e-05,
+      "loss": 1.1168,
+      "step": 7495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6601009368896484,
+      "learning_rate": 1.3150584795321639e-05,
+      "loss": 1.364,
+      "step": 7496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.482325553894043,
+      "learning_rate": 1.3149671052631581e-05,
+      "loss": 1.5663,
+      "step": 7497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9976296424865723,
+      "learning_rate": 1.3148757309941521e-05,
+      "loss": 1.4631,
+      "step": 7498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.384181499481201,
+      "learning_rate": 1.3147843567251464e-05,
+      "loss": 1.2776,
+      "step": 7499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9757752418518066,
+      "learning_rate": 1.3146929824561406e-05,
+      "loss": 1.7827,
+      "step": 7500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3009378910064697,
+      "learning_rate": 1.3146016081871346e-05,
+      "loss": 1.2129,
+      "step": 7501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2874839305877686,
+      "learning_rate": 1.3145102339181289e-05,
+      "loss": 1.3994,
+      "step": 7502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.613372325897217,
+      "learning_rate": 1.3144188596491229e-05,
+      "loss": 1.4139,
+      "step": 7503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6794087886810303,
+      "learning_rate": 1.3143274853801171e-05,
+      "loss": 1.7133,
+      "step": 7504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5697343349456787,
+      "learning_rate": 1.3142361111111112e-05,
+      "loss": 2.1317,
+      "step": 7505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9298925399780273,
+      "learning_rate": 1.3141447368421054e-05,
+      "loss": 1.727,
+      "step": 7506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.443211555480957,
+      "learning_rate": 1.3140533625730996e-05,
+      "loss": 1.6791,
+      "step": 7507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.674600601196289,
+      "learning_rate": 1.3139619883040937e-05,
+      "loss": 1.2433,
+      "step": 7508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.775961399078369,
+      "learning_rate": 1.3138706140350879e-05,
+      "loss": 1.4685,
+      "step": 7509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2689218521118164,
+      "learning_rate": 1.313779239766082e-05,
+      "loss": 1.8904,
+      "step": 7510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.749438762664795,
+      "learning_rate": 1.3136878654970762e-05,
+      "loss": 1.422,
+      "step": 7511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.613713026046753,
+      "learning_rate": 1.3135964912280702e-05,
+      "loss": 1.6387,
+      "step": 7512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6441614627838135,
+      "learning_rate": 1.3135051169590645e-05,
+      "loss": 1.5895,
+      "step": 7513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5480639934539795,
+      "learning_rate": 1.3134137426900587e-05,
+      "loss": 1.4415,
+      "step": 7514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9317939281463623,
+      "learning_rate": 1.3133223684210527e-05,
+      "loss": 1.5519,
+      "step": 7515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.954190492630005,
+      "learning_rate": 1.313230994152047e-05,
+      "loss": 1.7495,
+      "step": 7516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6930949687957764,
+      "learning_rate": 1.313139619883041e-05,
+      "loss": 1.3516,
+      "step": 7517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.521967649459839,
+      "learning_rate": 1.3130482456140352e-05,
+      "loss": 1.8913,
+      "step": 7518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4103362560272217,
+      "learning_rate": 1.3129568713450293e-05,
+      "loss": 1.3126,
+      "step": 7519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.789335250854492,
+      "learning_rate": 1.3128654970760235e-05,
+      "loss": 1.3783,
+      "step": 7520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6402463912963867,
+      "learning_rate": 1.3127741228070177e-05,
+      "loss": 1.3472,
+      "step": 7521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.025484561920166,
+      "learning_rate": 1.3126827485380118e-05,
+      "loss": 1.6695,
+      "step": 7522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.801391839981079,
+      "learning_rate": 1.312591374269006e-05,
+      "loss": 1.6882,
+      "step": 7523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2688584327697754,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.6856,
+      "step": 7524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.541736602783203,
+      "learning_rate": 1.3124086257309943e-05,
+      "loss": 1.6453,
+      "step": 7525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.715869188308716,
+      "learning_rate": 1.3123172514619883e-05,
+      "loss": 1.8615,
+      "step": 7526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.594226598739624,
+      "learning_rate": 1.3122258771929826e-05,
+      "loss": 1.4895,
+      "step": 7527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8726227283477783,
+      "learning_rate": 1.3121345029239768e-05,
+      "loss": 1.6131,
+      "step": 7528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.103917121887207,
+      "learning_rate": 1.3120431286549708e-05,
+      "loss": 1.7008,
+      "step": 7529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7714974880218506,
+      "learning_rate": 1.311951754385965e-05,
+      "loss": 1.6424,
+      "step": 7530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6079299449920654,
+      "learning_rate": 1.3118603801169591e-05,
+      "loss": 1.4432,
+      "step": 7531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6598544120788574,
+      "learning_rate": 1.3117690058479533e-05,
+      "loss": 1.3893,
+      "step": 7532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.280085802078247,
+      "learning_rate": 1.3116776315789474e-05,
+      "loss": 1.9201,
+      "step": 7533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.43029522895813,
+      "learning_rate": 1.3115862573099416e-05,
+      "loss": 1.9571,
+      "step": 7534
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.819730281829834,
+      "learning_rate": 1.3114948830409358e-05,
+      "loss": 1.3128,
+      "step": 7535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7355308532714844,
+      "learning_rate": 1.3114035087719299e-05,
+      "loss": 1.7821,
+      "step": 7536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.264524221420288,
+      "learning_rate": 1.3113121345029241e-05,
+      "loss": 1.4059,
+      "step": 7537
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.747714042663574,
+      "learning_rate": 1.3112207602339181e-05,
+      "loss": 1.427,
+      "step": 7538
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.403470277786255,
+      "learning_rate": 1.3111293859649124e-05,
+      "loss": 1.4245,
+      "step": 7539
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8437349796295166,
+      "learning_rate": 1.3110380116959064e-05,
+      "loss": 1.3395,
+      "step": 7540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.584524631500244,
+      "learning_rate": 1.3109466374269006e-05,
+      "loss": 1.3752,
+      "step": 7541
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.343517303466797,
+      "learning_rate": 1.3108552631578949e-05,
+      "loss": 1.2478,
+      "step": 7542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3542120456695557,
+      "learning_rate": 1.310763888888889e-05,
+      "loss": 1.3278,
+      "step": 7543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2885069847106934,
+      "learning_rate": 1.3106725146198831e-05,
+      "loss": 2.1874,
+      "step": 7544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.359647035598755,
+      "learning_rate": 1.3105811403508772e-05,
+      "loss": 1.6865,
+      "step": 7545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2118844985961914,
+      "learning_rate": 1.3104897660818714e-05,
+      "loss": 1.3394,
+      "step": 7546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.921527147293091,
+      "learning_rate": 1.3103983918128655e-05,
+      "loss": 1.5015,
+      "step": 7547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.593461513519287,
+      "learning_rate": 1.3103070175438597e-05,
+      "loss": 1.3328,
+      "step": 7548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3510282039642334,
+      "learning_rate": 1.3102156432748539e-05,
+      "loss": 1.5916,
+      "step": 7549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.998934268951416,
+      "learning_rate": 1.310124269005848e-05,
+      "loss": 1.5284,
+      "step": 7550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.588298797607422,
+      "learning_rate": 1.3100328947368422e-05,
+      "loss": 1.3125,
+      "step": 7551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.585730791091919,
+      "learning_rate": 1.3099415204678362e-05,
+      "loss": 1.5355,
+      "step": 7552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.635124444961548,
+      "learning_rate": 1.3098501461988305e-05,
+      "loss": 1.5876,
+      "step": 7553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9313249588012695,
+      "learning_rate": 1.3097587719298245e-05,
+      "loss": 1.8774,
+      "step": 7554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4984500408172607,
+      "learning_rate": 1.3096673976608187e-05,
+      "loss": 1.7024,
+      "step": 7555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.129915714263916,
+      "learning_rate": 1.309576023391813e-05,
+      "loss": 1.1062,
+      "step": 7556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.276872396469116,
+      "learning_rate": 1.309484649122807e-05,
+      "loss": 1.9321,
+      "step": 7557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8152027130126953,
+      "learning_rate": 1.3093932748538012e-05,
+      "loss": 1.552,
+      "step": 7558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0491483211517334,
+      "learning_rate": 1.3093019005847953e-05,
+      "loss": 1.7101,
+      "step": 7559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9436588287353516,
+      "learning_rate": 1.3092105263157895e-05,
+      "loss": 1.5625,
+      "step": 7560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2359354496002197,
+      "learning_rate": 1.3091191520467836e-05,
+      "loss": 1.2612,
+      "step": 7561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.50085711479187,
+      "learning_rate": 1.3090277777777778e-05,
+      "loss": 1.4814,
+      "step": 7562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3079915046691895,
+      "learning_rate": 1.308936403508772e-05,
+      "loss": 1.2489,
+      "step": 7563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.446021318435669,
+      "learning_rate": 1.308845029239766e-05,
+      "loss": 1.3072,
+      "step": 7564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.762510299682617,
+      "learning_rate": 1.3087536549707603e-05,
+      "loss": 1.5682,
+      "step": 7565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4104926586151123,
+      "learning_rate": 1.3086622807017543e-05,
+      "loss": 1.6395,
+      "step": 7566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.622901439666748,
+      "learning_rate": 1.3085709064327487e-05,
+      "loss": 1.6401,
+      "step": 7567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6951169967651367,
+      "learning_rate": 1.308479532163743e-05,
+      "loss": 1.9935,
+      "step": 7568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.553347587585449,
+      "learning_rate": 1.308388157894737e-05,
+      "loss": 1.5406,
+      "step": 7569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.393467664718628,
+      "learning_rate": 1.3082967836257312e-05,
+      "loss": 1.44,
+      "step": 7570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7523200511932373,
+      "learning_rate": 1.3082054093567253e-05,
+      "loss": 1.5695,
+      "step": 7571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5028460025787354,
+      "learning_rate": 1.3081140350877195e-05,
+      "loss": 1.6405,
+      "step": 7572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9068799018859863,
+      "learning_rate": 1.3080226608187137e-05,
+      "loss": 1.8196,
+      "step": 7573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.445441484451294,
+      "learning_rate": 1.3079312865497078e-05,
+      "loss": 1.1714,
+      "step": 7574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.96091365814209,
+      "learning_rate": 1.307839912280702e-05,
+      "loss": 1.9808,
+      "step": 7575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8526605367660522,
+      "learning_rate": 1.307748538011696e-05,
+      "loss": 1.2394,
+      "step": 7576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3508598804473877,
+      "learning_rate": 1.3076571637426903e-05,
+      "loss": 1.3135,
+      "step": 7577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.267124652862549,
+      "learning_rate": 1.3075657894736843e-05,
+      "loss": 1.1745,
+      "step": 7578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4664361476898193,
+      "learning_rate": 1.3074744152046785e-05,
+      "loss": 1.3376,
+      "step": 7579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3890697956085205,
+      "learning_rate": 1.3073830409356728e-05,
+      "loss": 1.3644,
+      "step": 7580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.794367790222168,
+      "learning_rate": 1.3072916666666668e-05,
+      "loss": 1.5858,
+      "step": 7581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.496839761734009,
+      "learning_rate": 1.307200292397661e-05,
+      "loss": 1.8356,
+      "step": 7582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.118877410888672,
+      "learning_rate": 1.3071089181286551e-05,
+      "loss": 1.19,
+      "step": 7583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.569328546524048,
+      "learning_rate": 1.3070175438596493e-05,
+      "loss": 2.1215,
+      "step": 7584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.059717893600464,
+      "learning_rate": 1.3069261695906434e-05,
+      "loss": 1.7985,
+      "step": 7585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3498072624206543,
+      "learning_rate": 1.3068347953216376e-05,
+      "loss": 1.7004,
+      "step": 7586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6418344974517822,
+      "learning_rate": 1.3067434210526318e-05,
+      "loss": 1.7474,
+      "step": 7587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.552260160446167,
+      "learning_rate": 1.3066520467836259e-05,
+      "loss": 1.5513,
+      "step": 7588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3867197036743164,
+      "learning_rate": 1.3065606725146201e-05,
+      "loss": 1.2199,
+      "step": 7589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.532175302505493,
+      "learning_rate": 1.3064692982456141e-05,
+      "loss": 1.2941,
+      "step": 7590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.286618947982788,
+      "learning_rate": 1.3063779239766084e-05,
+      "loss": 1.7436,
+      "step": 7591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.386000633239746,
+      "learning_rate": 1.3062865497076024e-05,
+      "loss": 1.2695,
+      "step": 7592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0102484226226807,
+      "learning_rate": 1.3061951754385966e-05,
+      "loss": 1.5924,
+      "step": 7593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.366584300994873,
+      "learning_rate": 1.3061038011695909e-05,
+      "loss": 1.4682,
+      "step": 7594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9150102138519287,
+      "learning_rate": 1.3060124269005849e-05,
+      "loss": 1.6269,
+      "step": 7595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4307971000671387,
+      "learning_rate": 1.3059210526315791e-05,
+      "loss": 1.7096,
+      "step": 7596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7923924922943115,
+      "learning_rate": 1.3058296783625732e-05,
+      "loss": 1.7022,
+      "step": 7597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9618723392486572,
+      "learning_rate": 1.3057383040935674e-05,
+      "loss": 1.6198,
+      "step": 7598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.646665096282959,
+      "learning_rate": 1.3056469298245615e-05,
+      "loss": 1.4403,
+      "step": 7599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7312166690826416,
+      "learning_rate": 1.3055555555555557e-05,
+      "loss": 1.7382,
+      "step": 7600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9903066158294678,
+      "learning_rate": 1.3054641812865499e-05,
+      "loss": 1.8317,
+      "step": 7601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.025710344314575,
+      "learning_rate": 1.305372807017544e-05,
+      "loss": 1.2976,
+      "step": 7602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3255646228790283,
+      "learning_rate": 1.3052814327485382e-05,
+      "loss": 1.3321,
+      "step": 7603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6302545070648193,
+      "learning_rate": 1.3051900584795322e-05,
+      "loss": 1.2982,
+      "step": 7604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1142091751098633,
+      "learning_rate": 1.3050986842105265e-05,
+      "loss": 1.2731,
+      "step": 7605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.14876389503479,
+      "learning_rate": 1.3050073099415205e-05,
+      "loss": 1.6192,
+      "step": 7606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.096499443054199,
+      "learning_rate": 1.3049159356725147e-05,
+      "loss": 1.3054,
+      "step": 7607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4724371433258057,
+      "learning_rate": 1.304824561403509e-05,
+      "loss": 1.4294,
+      "step": 7608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.633986711502075,
+      "learning_rate": 1.304733187134503e-05,
+      "loss": 1.3505,
+      "step": 7609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.476583242416382,
+      "learning_rate": 1.3046418128654972e-05,
+      "loss": 1.2556,
+      "step": 7610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9069745540618896,
+      "learning_rate": 1.3045504385964913e-05,
+      "loss": 1.5245,
+      "step": 7611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8430123329162598,
+      "learning_rate": 1.3044590643274855e-05,
+      "loss": 1.6809,
+      "step": 7612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6381399631500244,
+      "learning_rate": 1.3043676900584796e-05,
+      "loss": 1.373,
+      "step": 7613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3231465816497803,
+      "learning_rate": 1.3042763157894738e-05,
+      "loss": 1.2813,
+      "step": 7614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5924928188323975,
+      "learning_rate": 1.304184941520468e-05,
+      "loss": 1.5817,
+      "step": 7615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.991715431213379,
+      "learning_rate": 1.304093567251462e-05,
+      "loss": 1.8293,
+      "step": 7616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.781002998352051,
+      "learning_rate": 1.3040021929824563e-05,
+      "loss": 1.5257,
+      "step": 7617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0320537090301514,
+      "learning_rate": 1.3039108187134503e-05,
+      "loss": 1.5837,
+      "step": 7618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.190152645111084,
+      "learning_rate": 1.3038194444444446e-05,
+      "loss": 1.5958,
+      "step": 7619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0921597480773926,
+      "learning_rate": 1.3037280701754386e-05,
+      "loss": 1.6756,
+      "step": 7620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.578763484954834,
+      "learning_rate": 1.3036366959064328e-05,
+      "loss": 1.5551,
+      "step": 7621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1255714893341064,
+      "learning_rate": 1.303545321637427e-05,
+      "loss": 1.5667,
+      "step": 7622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6548473834991455,
+      "learning_rate": 1.3034539473684211e-05,
+      "loss": 1.4598,
+      "step": 7623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7109217643737793,
+      "learning_rate": 1.3033625730994153e-05,
+      "loss": 1.412,
+      "step": 7624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.021446943283081,
+      "learning_rate": 1.3032711988304094e-05,
+      "loss": 1.3966,
+      "step": 7625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4192917346954346,
+      "learning_rate": 1.3031798245614036e-05,
+      "loss": 1.3735,
+      "step": 7626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.488851547241211,
+      "learning_rate": 1.3030884502923977e-05,
+      "loss": 1.4382,
+      "step": 7627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8319313526153564,
+      "learning_rate": 1.3029970760233919e-05,
+      "loss": 1.6091,
+      "step": 7628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.360970973968506,
+      "learning_rate": 1.3029057017543861e-05,
+      "loss": 1.7635,
+      "step": 7629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1964786052703857,
+      "learning_rate": 1.3028143274853801e-05,
+      "loss": 1.8716,
+      "step": 7630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9663896560668945,
+      "learning_rate": 1.3027229532163744e-05,
+      "loss": 1.4922,
+      "step": 7631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.673290491104126,
+      "learning_rate": 1.3026315789473684e-05,
+      "loss": 1.5772,
+      "step": 7632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2870802879333496,
+      "learning_rate": 1.3025402046783626e-05,
+      "loss": 1.4369,
+      "step": 7633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3955652713775635,
+      "learning_rate": 1.3024488304093567e-05,
+      "loss": 1.4327,
+      "step": 7634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.531557321548462,
+      "learning_rate": 1.302357456140351e-05,
+      "loss": 1.9778,
+      "step": 7635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0050339698791504,
+      "learning_rate": 1.3022660818713451e-05,
+      "loss": 1.669,
+      "step": 7636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.32407283782959,
+      "learning_rate": 1.3021747076023392e-05,
+      "loss": 1.1837,
+      "step": 7637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3339385986328125,
+      "learning_rate": 1.3020833333333334e-05,
+      "loss": 1.4116,
+      "step": 7638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7897183895111084,
+      "learning_rate": 1.3019919590643275e-05,
+      "loss": 1.3807,
+      "step": 7639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6842963695526123,
+      "learning_rate": 1.3019005847953217e-05,
+      "loss": 1.4965,
+      "step": 7640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2272703647613525,
+      "learning_rate": 1.3018092105263157e-05,
+      "loss": 1.7015,
+      "step": 7641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.158797264099121,
+      "learning_rate": 1.30171783625731e-05,
+      "loss": 1.431,
+      "step": 7642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.918997287750244,
+      "learning_rate": 1.3016264619883042e-05,
+      "loss": 1.6968,
+      "step": 7643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6367173194885254,
+      "learning_rate": 1.3015350877192982e-05,
+      "loss": 1.6011,
+      "step": 7644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4528391361236572,
+      "learning_rate": 1.3014437134502925e-05,
+      "loss": 1.5295,
+      "step": 7645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.78804349899292,
+      "learning_rate": 1.3013523391812865e-05,
+      "loss": 1.4567,
+      "step": 7646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.713848114013672,
+      "learning_rate": 1.3012609649122807e-05,
+      "loss": 1.5972,
+      "step": 7647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.428323745727539,
+      "learning_rate": 1.3011695906432748e-05,
+      "loss": 1.301,
+      "step": 7648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.159023284912109,
+      "learning_rate": 1.301078216374269e-05,
+      "loss": 1.637,
+      "step": 7649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.885981321334839,
+      "learning_rate": 1.3009868421052632e-05,
+      "loss": 1.8323,
+      "step": 7650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.179409503936768,
+      "learning_rate": 1.3008954678362573e-05,
+      "loss": 1.7018,
+      "step": 7651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.051621437072754,
+      "learning_rate": 1.3008040935672515e-05,
+      "loss": 1.8997,
+      "step": 7652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2715981006622314,
+      "learning_rate": 1.3007127192982459e-05,
+      "loss": 1.3289,
+      "step": 7653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.524278163909912,
+      "learning_rate": 1.30062134502924e-05,
+      "loss": 1.5205,
+      "step": 7654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0145232677459717,
+      "learning_rate": 1.3005299707602342e-05,
+      "loss": 1.514,
+      "step": 7655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.30082368850708,
+      "learning_rate": 1.3004385964912282e-05,
+      "loss": 1.6636,
+      "step": 7656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6784725189208984,
+      "learning_rate": 1.3003472222222225e-05,
+      "loss": 1.4355,
+      "step": 7657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3025126457214355,
+      "learning_rate": 1.3002558479532165e-05,
+      "loss": 1.3806,
+      "step": 7658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6025829315185547,
+      "learning_rate": 1.3001644736842107e-05,
+      "loss": 1.5281,
+      "step": 7659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9911653995513916,
+      "learning_rate": 1.300073099415205e-05,
+      "loss": 1.4568,
+      "step": 7660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0437960624694824,
+      "learning_rate": 1.299981725146199e-05,
+      "loss": 1.8094,
+      "step": 7661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.004815101623535,
+      "learning_rate": 1.2998903508771932e-05,
+      "loss": 1.7191,
+      "step": 7662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7614965438842773,
+      "learning_rate": 1.2997989766081873e-05,
+      "loss": 1.4741,
+      "step": 7663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8327720165252686,
+      "learning_rate": 1.2997076023391815e-05,
+      "loss": 1.7833,
+      "step": 7664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.954817771911621,
+      "learning_rate": 1.2996162280701756e-05,
+      "loss": 1.717,
+      "step": 7665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1399800777435303,
+      "learning_rate": 1.2995248538011698e-05,
+      "loss": 1.2148,
+      "step": 7666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2218551635742188,
+      "learning_rate": 1.299433479532164e-05,
+      "loss": 1.5807,
+      "step": 7667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.308690309524536,
+      "learning_rate": 1.299342105263158e-05,
+      "loss": 1.4055,
+      "step": 7668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.045820474624634,
+      "learning_rate": 1.2992507309941523e-05,
+      "loss": 1.7711,
+      "step": 7669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6161251068115234,
+      "learning_rate": 1.2991593567251463e-05,
+      "loss": 1.4966,
+      "step": 7670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0183441638946533,
+      "learning_rate": 1.2990679824561405e-05,
+      "loss": 1.8705,
+      "step": 7671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.646249294281006,
+      "learning_rate": 1.2989766081871346e-05,
+      "loss": 1.688,
+      "step": 7672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5656652450561523,
+      "learning_rate": 1.2988852339181288e-05,
+      "loss": 1.5044,
+      "step": 7673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.542783260345459,
+      "learning_rate": 1.298793859649123e-05,
+      "loss": 1.1984,
+      "step": 7674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1016159057617188,
+      "learning_rate": 1.2987024853801171e-05,
+      "loss": 1.5495,
+      "step": 7675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4614477157592773,
+      "learning_rate": 1.2986111111111113e-05,
+      "loss": 1.4841,
+      "step": 7676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2863309383392334,
+      "learning_rate": 1.2985197368421054e-05,
+      "loss": 1.0553,
+      "step": 7677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.181607723236084,
+      "learning_rate": 1.2984283625730996e-05,
+      "loss": 1.6938,
+      "step": 7678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3847906589508057,
+      "learning_rate": 1.2983369883040936e-05,
+      "loss": 1.5642,
+      "step": 7679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3369884490966797,
+      "learning_rate": 1.2982456140350879e-05,
+      "loss": 1.4863,
+      "step": 7680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7074062824249268,
+      "learning_rate": 1.2981542397660821e-05,
+      "loss": 1.4918,
+      "step": 7681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2207815647125244,
+      "learning_rate": 1.2980628654970761e-05,
+      "loss": 2.1437,
+      "step": 7682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.437697172164917,
+      "learning_rate": 1.2979714912280704e-05,
+      "loss": 1.6913,
+      "step": 7683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.852585554122925,
+      "learning_rate": 1.2978801169590644e-05,
+      "loss": 1.2153,
+      "step": 7684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.753481149673462,
+      "learning_rate": 1.2977887426900586e-05,
+      "loss": 1.5768,
+      "step": 7685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8732268810272217,
+      "learning_rate": 1.2976973684210527e-05,
+      "loss": 1.729,
+      "step": 7686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2230072021484375,
+      "learning_rate": 1.2976059941520469e-05,
+      "loss": 1.6184,
+      "step": 7687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.312185525894165,
+      "learning_rate": 1.2975146198830411e-05,
+      "loss": 1.3615,
+      "step": 7688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3182404041290283,
+      "learning_rate": 1.2974232456140352e-05,
+      "loss": 1.5431,
+      "step": 7689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.710153818130493,
+      "learning_rate": 1.2973318713450294e-05,
+      "loss": 1.7961,
+      "step": 7690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.446765184402466,
+      "learning_rate": 1.2972404970760235e-05,
+      "loss": 1.719,
+      "step": 7691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.497523784637451,
+      "learning_rate": 1.2971491228070177e-05,
+      "loss": 1.4446,
+      "step": 7692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.712714195251465,
+      "learning_rate": 1.2970577485380117e-05,
+      "loss": 1.7223,
+      "step": 7693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.525545120239258,
+      "learning_rate": 1.296966374269006e-05,
+      "loss": 1.3819,
+      "step": 7694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0876026153564453,
+      "learning_rate": 1.2968750000000002e-05,
+      "loss": 1.753,
+      "step": 7695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.895507574081421,
+      "learning_rate": 1.2967836257309942e-05,
+      "loss": 1.5691,
+      "step": 7696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.876636266708374,
+      "learning_rate": 1.2966922514619885e-05,
+      "loss": 1.5231,
+      "step": 7697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5438106060028076,
+      "learning_rate": 1.2966008771929825e-05,
+      "loss": 1.5892,
+      "step": 7698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6307833194732666,
+      "learning_rate": 1.2965095029239767e-05,
+      "loss": 1.8021,
+      "step": 7699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2400612831115723,
+      "learning_rate": 1.2964181286549708e-05,
+      "loss": 1.6896,
+      "step": 7700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.78497576713562,
+      "learning_rate": 1.296326754385965e-05,
+      "loss": 1.5678,
+      "step": 7701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9493491649627686,
+      "learning_rate": 1.2962353801169592e-05,
+      "loss": 1.0531,
+      "step": 7702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7631852626800537,
+      "learning_rate": 1.2961440058479533e-05,
+      "loss": 1.2196,
+      "step": 7703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.700488567352295,
+      "learning_rate": 1.2960526315789475e-05,
+      "loss": 2.2593,
+      "step": 7704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9914467334747314,
+      "learning_rate": 1.2959612573099416e-05,
+      "loss": 2.1095,
+      "step": 7705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.20833420753479,
+      "learning_rate": 1.2958698830409358e-05,
+      "loss": 1.2822,
+      "step": 7706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.591688871383667,
+      "learning_rate": 1.2957785087719298e-05,
+      "loss": 1.6547,
+      "step": 7707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.411773204803467,
+      "learning_rate": 1.295687134502924e-05,
+      "loss": 1.5907,
+      "step": 7708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.93913197517395,
+      "learning_rate": 1.2955957602339183e-05,
+      "loss": 1.8442,
+      "step": 7709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.790070056915283,
+      "learning_rate": 1.2955043859649123e-05,
+      "loss": 1.4923,
+      "step": 7710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8402862548828125,
+      "learning_rate": 1.2954130116959066e-05,
+      "loss": 1.5399,
+      "step": 7711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7838480472564697,
+      "learning_rate": 1.2953216374269006e-05,
+      "loss": 1.5942,
+      "step": 7712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.937056541442871,
+      "learning_rate": 1.2952302631578948e-05,
+      "loss": 1.8734,
+      "step": 7713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.857221841812134,
+      "learning_rate": 1.2951388888888889e-05,
+      "loss": 1.8259,
+      "step": 7714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.72666335105896,
+      "learning_rate": 1.2950475146198831e-05,
+      "loss": 1.5052,
+      "step": 7715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.335400342941284,
+      "learning_rate": 1.2949561403508773e-05,
+      "loss": 1.4292,
+      "step": 7716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1012814044952393,
+      "learning_rate": 1.2948647660818714e-05,
+      "loss": 1.723,
+      "step": 7717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.709864377975464,
+      "learning_rate": 1.2947733918128656e-05,
+      "loss": 1.3326,
+      "step": 7718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0276238918304443,
+      "learning_rate": 1.2946820175438597e-05,
+      "loss": 1.8503,
+      "step": 7719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.07392954826355,
+      "learning_rate": 1.2945906432748539e-05,
+      "loss": 1.6106,
+      "step": 7720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2292122840881348,
+      "learning_rate": 1.294499269005848e-05,
+      "loss": 1.5346,
+      "step": 7721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.840869426727295,
+      "learning_rate": 1.2944078947368421e-05,
+      "loss": 1.7713,
+      "step": 7722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.821861505508423,
+      "learning_rate": 1.2943165204678364e-05,
+      "loss": 1.4443,
+      "step": 7723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9449336528778076,
+      "learning_rate": 1.2942251461988304e-05,
+      "loss": 1.7031,
+      "step": 7724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4788947105407715,
+      "learning_rate": 1.2941337719298246e-05,
+      "loss": 2.0578,
+      "step": 7725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8950133323669434,
+      "learning_rate": 1.2940423976608187e-05,
+      "loss": 1.701,
+      "step": 7726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8961501121520996,
+      "learning_rate": 1.293951023391813e-05,
+      "loss": 1.8407,
+      "step": 7727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.462369680404663,
+      "learning_rate": 1.293859649122807e-05,
+      "loss": 1.5174,
+      "step": 7728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.662972927093506,
+      "learning_rate": 1.2937682748538012e-05,
+      "loss": 1.643,
+      "step": 7729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9564671516418457,
+      "learning_rate": 1.2936769005847954e-05,
+      "loss": 1.6631,
+      "step": 7730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.756410837173462,
+      "learning_rate": 1.2935855263157895e-05,
+      "loss": 1.6596,
+      "step": 7731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6060402393341064,
+      "learning_rate": 1.2934941520467837e-05,
+      "loss": 1.6603,
+      "step": 7732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4694905281066895,
+      "learning_rate": 1.2934027777777777e-05,
+      "loss": 1.4059,
+      "step": 7733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.014190673828125,
+      "learning_rate": 1.293311403508772e-05,
+      "loss": 1.9479,
+      "step": 7734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1365766525268555,
+      "learning_rate": 1.293220029239766e-05,
+      "loss": 1.9102,
+      "step": 7735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.758409261703491,
+      "learning_rate": 1.2931286549707602e-05,
+      "loss": 1.8307,
+      "step": 7736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.620863437652588,
+      "learning_rate": 1.2930372807017545e-05,
+      "loss": 1.519,
+      "step": 7737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2581493854522705,
+      "learning_rate": 1.2929459064327487e-05,
+      "loss": 2.0171,
+      "step": 7738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4912455081939697,
+      "learning_rate": 1.2928545321637429e-05,
+      "loss": 1.2464,
+      "step": 7739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6773877143859863,
+      "learning_rate": 1.2927631578947371e-05,
+      "loss": 1.568,
+      "step": 7740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9517884254455566,
+      "learning_rate": 1.2926717836257312e-05,
+      "loss": 1.8344,
+      "step": 7741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4982051849365234,
+      "learning_rate": 1.2925804093567254e-05,
+      "loss": 1.5425,
+      "step": 7742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.544952630996704,
+      "learning_rate": 1.2924890350877195e-05,
+      "loss": 1.6096,
+      "step": 7743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.713510513305664,
+      "learning_rate": 1.2923976608187137e-05,
+      "loss": 1.6838,
+      "step": 7744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.621405601501465,
+      "learning_rate": 1.2923062865497077e-05,
+      "loss": 2.202,
+      "step": 7745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3539812564849854,
+      "learning_rate": 1.292214912280702e-05,
+      "loss": 1.3375,
+      "step": 7746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7825047969818115,
+      "learning_rate": 1.2921235380116962e-05,
+      "loss": 1.8474,
+      "step": 7747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.663121223449707,
+      "learning_rate": 1.2920321637426902e-05,
+      "loss": 1.8045,
+      "step": 7748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8649630546569824,
+      "learning_rate": 1.2919407894736845e-05,
+      "loss": 2.0102,
+      "step": 7749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9504175186157227,
+      "learning_rate": 1.2918494152046785e-05,
+      "loss": 1.7934,
+      "step": 7750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6739048957824707,
+      "learning_rate": 1.2917580409356727e-05,
+      "loss": 1.5817,
+      "step": 7751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.226635694503784,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 1.6219,
+      "step": 7752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.852090358734131,
+      "learning_rate": 1.291575292397661e-05,
+      "loss": 1.9625,
+      "step": 7753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9404513835906982,
+      "learning_rate": 1.2914839181286552e-05,
+      "loss": 1.4824,
+      "step": 7754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.486804962158203,
+      "learning_rate": 1.2913925438596493e-05,
+      "loss": 1.2828,
+      "step": 7755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.396188735961914,
+      "learning_rate": 1.2913011695906435e-05,
+      "loss": 2.0856,
+      "step": 7756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6666831970214844,
+      "learning_rate": 1.2912097953216376e-05,
+      "loss": 1.7803,
+      "step": 7757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.426179885864258,
+      "learning_rate": 1.2911184210526318e-05,
+      "loss": 1.3337,
+      "step": 7758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.417659044265747,
+      "learning_rate": 1.2910270467836258e-05,
+      "loss": 1.7273,
+      "step": 7759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.278146505355835,
+      "learning_rate": 1.29093567251462e-05,
+      "loss": 1.2689,
+      "step": 7760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8385446071624756,
+      "learning_rate": 1.2908442982456143e-05,
+      "loss": 1.6696,
+      "step": 7761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6555514335632324,
+      "learning_rate": 1.2907529239766083e-05,
+      "loss": 1.4496,
+      "step": 7762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.151371479034424,
+      "learning_rate": 1.2906615497076025e-05,
+      "loss": 1.2792,
+      "step": 7763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0117290019989014,
+      "learning_rate": 1.2905701754385966e-05,
+      "loss": 1.7113,
+      "step": 7764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9356815814971924,
+      "learning_rate": 1.2904788011695908e-05,
+      "loss": 1.7412,
+      "step": 7765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.930814743041992,
+      "learning_rate": 1.2903874269005849e-05,
+      "loss": 1.5451,
+      "step": 7766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.359916925430298,
+      "learning_rate": 1.2902960526315791e-05,
+      "loss": 1.5159,
+      "step": 7767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.392301082611084,
+      "learning_rate": 1.2902046783625733e-05,
+      "loss": 1.3554,
+      "step": 7768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6952953338623047,
+      "learning_rate": 1.2901133040935674e-05,
+      "loss": 1.5067,
+      "step": 7769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.126692771911621,
+      "learning_rate": 1.2900219298245616e-05,
+      "loss": 1.3797,
+      "step": 7770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5190212726593018,
+      "learning_rate": 1.2899305555555556e-05,
+      "loss": 1.4305,
+      "step": 7771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.5044734477996826,
+      "learning_rate": 1.2898391812865499e-05,
+      "loss": 1.9378,
+      "step": 7772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.757378101348877,
+      "learning_rate": 1.289747807017544e-05,
+      "loss": 1.4862,
+      "step": 7773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6405856609344482,
+      "learning_rate": 1.2896564327485381e-05,
+      "loss": 1.5106,
+      "step": 7774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.999145269393921,
+      "learning_rate": 1.2895650584795324e-05,
+      "loss": 1.5466,
+      "step": 7775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.503081798553467,
+      "learning_rate": 1.2894736842105264e-05,
+      "loss": 1.4268,
+      "step": 7776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9625494480133057,
+      "learning_rate": 1.2893823099415206e-05,
+      "loss": 1.5464,
+      "step": 7777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.638820171356201,
+      "learning_rate": 1.2892909356725147e-05,
+      "loss": 1.6274,
+      "step": 7778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7140798568725586,
+      "learning_rate": 1.2891995614035089e-05,
+      "loss": 1.5711,
+      "step": 7779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6383309364318848,
+      "learning_rate": 1.289108187134503e-05,
+      "loss": 1.4892,
+      "step": 7780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9941591024398804,
+      "learning_rate": 1.2890168128654972e-05,
+      "loss": 1.1749,
+      "step": 7781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6435673236846924,
+      "learning_rate": 1.2889254385964914e-05,
+      "loss": 1.81,
+      "step": 7782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.855181932449341,
+      "learning_rate": 1.2888340643274855e-05,
+      "loss": 1.4585,
+      "step": 7783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.456141710281372,
+      "learning_rate": 1.2887426900584797e-05,
+      "loss": 1.4052,
+      "step": 7784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2357492446899414,
+      "learning_rate": 1.2886513157894737e-05,
+      "loss": 1.1575,
+      "step": 7785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.719575881958008,
+      "learning_rate": 1.288559941520468e-05,
+      "loss": 1.2612,
+      "step": 7786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8698558807373047,
+      "learning_rate": 1.288468567251462e-05,
+      "loss": 1.5909,
+      "step": 7787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7109549045562744,
+      "learning_rate": 1.2883771929824562e-05,
+      "loss": 1.6796,
+      "step": 7788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9564270973205566,
+      "learning_rate": 1.2882858187134505e-05,
+      "loss": 1.4431,
+      "step": 7789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8119468688964844,
+      "learning_rate": 1.2881944444444445e-05,
+      "loss": 1.6576,
+      "step": 7790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5334973335266113,
+      "learning_rate": 1.2881030701754387e-05,
+      "loss": 1.3726,
+      "step": 7791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.445345640182495,
+      "learning_rate": 1.2880116959064328e-05,
+      "loss": 1.5026,
+      "step": 7792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8191070556640625,
+      "learning_rate": 1.287920321637427e-05,
+      "loss": 1.776,
+      "step": 7793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2860281467437744,
+      "learning_rate": 1.287828947368421e-05,
+      "loss": 1.2534,
+      "step": 7794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.063065528869629,
+      "learning_rate": 1.2877375730994153e-05,
+      "loss": 2.0368,
+      "step": 7795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.792527914047241,
+      "learning_rate": 1.2876461988304095e-05,
+      "loss": 1.5807,
+      "step": 7796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.152884006500244,
+      "learning_rate": 1.2875548245614036e-05,
+      "loss": 1.3231,
+      "step": 7797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5772995948791504,
+      "learning_rate": 1.2874634502923978e-05,
+      "loss": 1.622,
+      "step": 7798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.418821096420288,
+      "learning_rate": 1.2873720760233918e-05,
+      "loss": 1.1911,
+      "step": 7799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3413503170013428,
+      "learning_rate": 1.287280701754386e-05,
+      "loss": 1.2217,
+      "step": 7800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.649393320083618,
+      "learning_rate": 1.2871893274853801e-05,
+      "loss": 1.4739,
+      "step": 7801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.979614734649658,
+      "learning_rate": 1.2870979532163743e-05,
+      "loss": 1.687,
+      "step": 7802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.248433828353882,
+      "learning_rate": 1.2870065789473686e-05,
+      "loss": 1.4153,
+      "step": 7803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.432821750640869,
+      "learning_rate": 1.2869152046783626e-05,
+      "loss": 1.497,
+      "step": 7804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0642714500427246,
+      "learning_rate": 1.2868238304093568e-05,
+      "loss": 0.9753,
+      "step": 7805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.507110834121704,
+      "learning_rate": 1.2867324561403509e-05,
+      "loss": 1.5855,
+      "step": 7806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6095101833343506,
+      "learning_rate": 1.2866410818713451e-05,
+      "loss": 1.632,
+      "step": 7807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.95267391204834,
+      "learning_rate": 1.2865497076023392e-05,
+      "loss": 1.4938,
+      "step": 7808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.337686538696289,
+      "learning_rate": 1.2864583333333334e-05,
+      "loss": 1.5748,
+      "step": 7809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6918272972106934,
+      "learning_rate": 1.2863669590643276e-05,
+      "loss": 1.4383,
+      "step": 7810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0536110401153564,
+      "learning_rate": 1.2862755847953217e-05,
+      "loss": 1.6136,
+      "step": 7811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6876518726348877,
+      "learning_rate": 1.2861842105263159e-05,
+      "loss": 1.1962,
+      "step": 7812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5976064205169678,
+      "learning_rate": 1.28609283625731e-05,
+      "loss": 1.5275,
+      "step": 7813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.48844313621521,
+      "learning_rate": 1.2860014619883041e-05,
+      "loss": 1.3227,
+      "step": 7814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2803189754486084,
+      "learning_rate": 1.2859100877192982e-05,
+      "loss": 2.0281,
+      "step": 7815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.569676399230957,
+      "learning_rate": 1.2858187134502924e-05,
+      "loss": 1.3357,
+      "step": 7816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.075697660446167,
+      "learning_rate": 1.2857273391812865e-05,
+      "loss": 1.8298,
+      "step": 7817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5705809593200684,
+      "learning_rate": 1.2856359649122807e-05,
+      "loss": 1.4904,
+      "step": 7818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9496054649353027,
+      "learning_rate": 1.285544590643275e-05,
+      "loss": 1.5334,
+      "step": 7819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4818129539489746,
+      "learning_rate": 1.285453216374269e-05,
+      "loss": 1.4276,
+      "step": 7820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4338905811309814,
+      "learning_rate": 1.2853618421052632e-05,
+      "loss": 1.5293,
+      "step": 7821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9041078090667725,
+      "learning_rate": 1.2852704678362572e-05,
+      "loss": 1.8961,
+      "step": 7822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7378063201904297,
+      "learning_rate": 1.2851790935672515e-05,
+      "loss": 1.3736,
+      "step": 7823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9046101570129395,
+      "learning_rate": 1.2850877192982459e-05,
+      "loss": 1.4999,
+      "step": 7824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6227688789367676,
+      "learning_rate": 1.2849963450292399e-05,
+      "loss": 1.5215,
+      "step": 7825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.187577486038208,
+      "learning_rate": 1.2849049707602341e-05,
+      "loss": 1.1522,
+      "step": 7826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6402478218078613,
+      "learning_rate": 1.2848135964912284e-05,
+      "loss": 1.5118,
+      "step": 7827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0620288848876953,
+      "learning_rate": 1.2847222222222224e-05,
+      "loss": 1.734,
+      "step": 7828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.856441020965576,
+      "learning_rate": 1.2846308479532166e-05,
+      "loss": 1.7829,
+      "step": 7829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6026480197906494,
+      "learning_rate": 1.2845394736842107e-05,
+      "loss": 2.0511,
+      "step": 7830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2191977500915527,
+      "learning_rate": 1.2844480994152049e-05,
+      "loss": 1.1566,
+      "step": 7831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3874449729919434,
+      "learning_rate": 1.284356725146199e-05,
+      "loss": 2.0209,
+      "step": 7832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7936720848083496,
+      "learning_rate": 1.2842653508771932e-05,
+      "loss": 1.5738,
+      "step": 7833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4006035327911377,
+      "learning_rate": 1.2841739766081874e-05,
+      "loss": 1.4533,
+      "step": 7834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.251798629760742,
+      "learning_rate": 1.2840826023391815e-05,
+      "loss": 1.72,
+      "step": 7835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5949318408966064,
+      "learning_rate": 1.2839912280701757e-05,
+      "loss": 1.4996,
+      "step": 7836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3298840522766113,
+      "learning_rate": 1.2838998538011697e-05,
+      "loss": 1.1214,
+      "step": 7837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1294589042663574,
+      "learning_rate": 1.283808479532164e-05,
+      "loss": 1.8372,
+      "step": 7838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4234235286712646,
+      "learning_rate": 1.283717105263158e-05,
+      "loss": 1.404,
+      "step": 7839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.640564203262329,
+      "learning_rate": 1.2836257309941522e-05,
+      "loss": 1.8252,
+      "step": 7840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9734244346618652,
+      "learning_rate": 1.2835343567251465e-05,
+      "loss": 1.7567,
+      "step": 7841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5750107765197754,
+      "learning_rate": 1.2834429824561405e-05,
+      "loss": 1.6711,
+      "step": 7842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.332205295562744,
+      "learning_rate": 1.2833516081871347e-05,
+      "loss": 2.0454,
+      "step": 7843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.841599941253662,
+      "learning_rate": 1.2832602339181288e-05,
+      "loss": 1.6316,
+      "step": 7844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0226128101348877,
+      "learning_rate": 1.283168859649123e-05,
+      "loss": 1.8833,
+      "step": 7845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6376616954803467,
+      "learning_rate": 1.283077485380117e-05,
+      "loss": 1.5946,
+      "step": 7846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4773261547088623,
+      "learning_rate": 1.2829861111111113e-05,
+      "loss": 1.5121,
+      "step": 7847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8150296211242676,
+      "learning_rate": 1.2828947368421055e-05,
+      "loss": 1.5544,
+      "step": 7848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5229320526123047,
+      "learning_rate": 1.2828033625730996e-05,
+      "loss": 1.2968,
+      "step": 7849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.731741428375244,
+      "learning_rate": 1.2827119883040938e-05,
+      "loss": 1.5655,
+      "step": 7850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.884183645248413,
+      "learning_rate": 1.2826206140350878e-05,
+      "loss": 1.7804,
+      "step": 7851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6512765884399414,
+      "learning_rate": 1.282529239766082e-05,
+      "loss": 1.3701,
+      "step": 7852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8303802013397217,
+      "learning_rate": 1.2824378654970761e-05,
+      "loss": 1.4717,
+      "step": 7853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.864285707473755,
+      "learning_rate": 1.2823464912280703e-05,
+      "loss": 1.573,
+      "step": 7854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0813164710998535,
+      "learning_rate": 1.2822551169590645e-05,
+      "loss": 1.9208,
+      "step": 7855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.685028076171875,
+      "learning_rate": 1.2821637426900586e-05,
+      "loss": 1.5041,
+      "step": 7856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0060484409332275,
+      "learning_rate": 1.2820723684210528e-05,
+      "loss": 1.0137,
+      "step": 7857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.922553539276123,
+      "learning_rate": 1.2819809941520469e-05,
+      "loss": 1.5432,
+      "step": 7858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.78312349319458,
+      "learning_rate": 1.2818896198830411e-05,
+      "loss": 1.6476,
+      "step": 7859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3014256954193115,
+      "learning_rate": 1.2817982456140351e-05,
+      "loss": 1.1062,
+      "step": 7860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.435598373413086,
+      "learning_rate": 1.2817068713450294e-05,
+      "loss": 1.5063,
+      "step": 7861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6364786624908447,
+      "learning_rate": 1.2816154970760234e-05,
+      "loss": 1.3707,
+      "step": 7862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.126497507095337,
+      "learning_rate": 1.2815241228070176e-05,
+      "loss": 1.7118,
+      "step": 7863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8483691215515137,
+      "learning_rate": 1.2814327485380119e-05,
+      "loss": 1.6263,
+      "step": 7864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7278494834899902,
+      "learning_rate": 1.281341374269006e-05,
+      "loss": 1.6468,
+      "step": 7865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.871074914932251,
+      "learning_rate": 1.2812500000000001e-05,
+      "loss": 1.6631,
+      "step": 7866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.549651861190796,
+      "learning_rate": 1.2811586257309942e-05,
+      "loss": 1.2223,
+      "step": 7867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2979087829589844,
+      "learning_rate": 1.2810672514619884e-05,
+      "loss": 1.8043,
+      "step": 7868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.642763614654541,
+      "learning_rate": 1.2809758771929825e-05,
+      "loss": 1.5563,
+      "step": 7869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3590891361236572,
+      "learning_rate": 1.2808845029239767e-05,
+      "loss": 1.3885,
+      "step": 7870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7954812049865723,
+      "learning_rate": 1.2807931286549709e-05,
+      "loss": 1.7907,
+      "step": 7871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3910293579101562,
+      "learning_rate": 1.280701754385965e-05,
+      "loss": 1.551,
+      "step": 7872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8561911582946777,
+      "learning_rate": 1.2806103801169592e-05,
+      "loss": 1.7817,
+      "step": 7873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6865179538726807,
+      "learning_rate": 1.2805190058479532e-05,
+      "loss": 1.7904,
+      "step": 7874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.370478868484497,
+      "learning_rate": 1.2804276315789475e-05,
+      "loss": 1.4291,
+      "step": 7875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9174680709838867,
+      "learning_rate": 1.2803362573099415e-05,
+      "loss": 1.5938,
+      "step": 7876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3079612255096436,
+      "learning_rate": 1.2802448830409357e-05,
+      "loss": 1.1986,
+      "step": 7877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.491466760635376,
+      "learning_rate": 1.28015350877193e-05,
+      "loss": 1.4859,
+      "step": 7878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.497004747390747,
+      "learning_rate": 1.280062134502924e-05,
+      "loss": 1.3795,
+      "step": 7879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7646827697753906,
+      "learning_rate": 1.2799707602339182e-05,
+      "loss": 1.676,
+      "step": 7880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.587336778640747,
+      "learning_rate": 1.2798793859649123e-05,
+      "loss": 1.4339,
+      "step": 7881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.688960075378418,
+      "learning_rate": 1.2797880116959065e-05,
+      "loss": 1.7613,
+      "step": 7882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7157154083251953,
+      "learning_rate": 1.2796966374269006e-05,
+      "loss": 1.4891,
+      "step": 7883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8046557903289795,
+      "learning_rate": 1.2796052631578948e-05,
+      "loss": 1.5613,
+      "step": 7884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.387561559677124,
+      "learning_rate": 1.279513888888889e-05,
+      "loss": 1.5378,
+      "step": 7885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.472764492034912,
+      "learning_rate": 1.279422514619883e-05,
+      "loss": 1.4629,
+      "step": 7886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9628195762634277,
+      "learning_rate": 1.2793311403508773e-05,
+      "loss": 1.5105,
+      "step": 7887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8098814487457275,
+      "learning_rate": 1.2792397660818713e-05,
+      "loss": 1.3137,
+      "step": 7888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.338752269744873,
+      "learning_rate": 1.2791483918128656e-05,
+      "loss": 1.2127,
+      "step": 7889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8861677646636963,
+      "learning_rate": 1.2790570175438596e-05,
+      "loss": 1.6019,
+      "step": 7890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.752060651779175,
+      "learning_rate": 1.2789656432748538e-05,
+      "loss": 1.4027,
+      "step": 7891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4805266857147217,
+      "learning_rate": 1.278874269005848e-05,
+      "loss": 1.4581,
+      "step": 7892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2383029460906982,
+      "learning_rate": 1.2787828947368421e-05,
+      "loss": 1.7384,
+      "step": 7893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2005062103271484,
+      "learning_rate": 1.2786915204678363e-05,
+      "loss": 1.7482,
+      "step": 7894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.69576358795166,
+      "learning_rate": 1.2786001461988304e-05,
+      "loss": 1.6453,
+      "step": 7895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7574169635772705,
+      "learning_rate": 1.2785087719298246e-05,
+      "loss": 1.9401,
+      "step": 7896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.088395833969116,
+      "learning_rate": 1.2784173976608187e-05,
+      "loss": 1.9214,
+      "step": 7897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.582761287689209,
+      "learning_rate": 1.2783260233918129e-05,
+      "loss": 1.6236,
+      "step": 7898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.377876043319702,
+      "learning_rate": 1.2782346491228071e-05,
+      "loss": 1.4955,
+      "step": 7899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.642194986343384,
+      "learning_rate": 1.2781432748538012e-05,
+      "loss": 1.5604,
+      "step": 7900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.648076057434082,
+      "learning_rate": 1.2780519005847954e-05,
+      "loss": 1.5691,
+      "step": 7901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5353236198425293,
+      "learning_rate": 1.2779605263157894e-05,
+      "loss": 1.6537,
+      "step": 7902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.270700693130493,
+      "learning_rate": 1.2778691520467837e-05,
+      "loss": 1.3908,
+      "step": 7903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7080209255218506,
+      "learning_rate": 1.2777777777777777e-05,
+      "loss": 1.9188,
+      "step": 7904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0465004444122314,
+      "learning_rate": 1.277686403508772e-05,
+      "loss": 1.5706,
+      "step": 7905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.338702440261841,
+      "learning_rate": 1.2775950292397661e-05,
+      "loss": 1.3028,
+      "step": 7906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.374671459197998,
+      "learning_rate": 1.2775036549707602e-05,
+      "loss": 1.9792,
+      "step": 7907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5504884719848633,
+      "learning_rate": 1.2774122807017544e-05,
+      "loss": 1.4108,
+      "step": 7908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.825834035873413,
+      "learning_rate": 1.2773209064327488e-05,
+      "loss": 1.4316,
+      "step": 7909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8588459491729736,
+      "learning_rate": 1.2772295321637429e-05,
+      "loss": 1.6291,
+      "step": 7910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8014516830444336,
+      "learning_rate": 1.2771381578947371e-05,
+      "loss": 1.467,
+      "step": 7911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.885730743408203,
+      "learning_rate": 1.2770467836257311e-05,
+      "loss": 1.4649,
+      "step": 7912
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.444396495819092,
+      "learning_rate": 1.2769554093567254e-05,
+      "loss": 1.2774,
+      "step": 7913
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.523958683013916,
+      "learning_rate": 1.2768640350877194e-05,
+      "loss": 1.2864,
+      "step": 7914
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4847280979156494,
+      "learning_rate": 1.2767726608187136e-05,
+      "loss": 1.5377,
+      "step": 7915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.816372871398926,
+      "learning_rate": 1.2766812865497079e-05,
+      "loss": 1.835,
+      "step": 7916
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.548891067504883,
+      "learning_rate": 1.2765899122807019e-05,
+      "loss": 1.6112,
+      "step": 7917
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.827434539794922,
+      "learning_rate": 1.2764985380116961e-05,
+      "loss": 1.8002,
+      "step": 7918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.225632667541504,
+      "learning_rate": 1.2764071637426902e-05,
+      "loss": 1.167,
+      "step": 7919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.901219367980957,
+      "learning_rate": 1.2763157894736844e-05,
+      "loss": 1.7682,
+      "step": 7920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.847550392150879,
+      "learning_rate": 1.2762244152046785e-05,
+      "loss": 1.7687,
+      "step": 7921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2610270977020264,
+      "learning_rate": 1.2761330409356727e-05,
+      "loss": 1.2874,
+      "step": 7922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6327221393585205,
+      "learning_rate": 1.2760416666666669e-05,
+      "loss": 1.3879,
+      "step": 7923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.097769021987915,
+      "learning_rate": 1.275950292397661e-05,
+      "loss": 2.0102,
+      "step": 7924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3063855171203613,
+      "learning_rate": 1.2758589181286552e-05,
+      "loss": 1.2599,
+      "step": 7925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.058701515197754,
+      "learning_rate": 1.2757675438596492e-05,
+      "loss": 1.3746,
+      "step": 7926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7349648475646973,
+      "learning_rate": 1.2756761695906435e-05,
+      "loss": 1.5133,
+      "step": 7927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.546931266784668,
+      "learning_rate": 1.2755847953216375e-05,
+      "loss": 1.3918,
+      "step": 7928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0493719577789307,
+      "learning_rate": 1.2754934210526317e-05,
+      "loss": 1.6557,
+      "step": 7929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2115650177001953,
+      "learning_rate": 1.275402046783626e-05,
+      "loss": 1.7872,
+      "step": 7930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.046967029571533,
+      "learning_rate": 1.27531067251462e-05,
+      "loss": 1.9234,
+      "step": 7931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6548349857330322,
+      "learning_rate": 1.2752192982456142e-05,
+      "loss": 1.5291,
+      "step": 7932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8649322986602783,
+      "learning_rate": 1.2751279239766083e-05,
+      "loss": 1.6663,
+      "step": 7933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2106895446777344,
+      "learning_rate": 1.2750365497076025e-05,
+      "loss": 1.8432,
+      "step": 7934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3908133506774902,
+      "learning_rate": 1.2749451754385966e-05,
+      "loss": 1.3936,
+      "step": 7935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4652318954467773,
+      "learning_rate": 1.2748538011695908e-05,
+      "loss": 1.5235,
+      "step": 7936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.407318115234375,
+      "learning_rate": 1.274762426900585e-05,
+      "loss": 1.2637,
+      "step": 7937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.550502300262451,
+      "learning_rate": 1.274671052631579e-05,
+      "loss": 1.7044,
+      "step": 7938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.326749324798584,
+      "learning_rate": 1.2745796783625733e-05,
+      "loss": 1.2125,
+      "step": 7939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.530968189239502,
+      "learning_rate": 1.2744883040935673e-05,
+      "loss": 1.3813,
+      "step": 7940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1841249465942383,
+      "learning_rate": 1.2743969298245616e-05,
+      "loss": 1.1457,
+      "step": 7941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.731745958328247,
+      "learning_rate": 1.2743055555555556e-05,
+      "loss": 1.5105,
+      "step": 7942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1177051067352295,
+      "learning_rate": 1.2742141812865498e-05,
+      "loss": 1.2977,
+      "step": 7943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1164302825927734,
+      "learning_rate": 1.274122807017544e-05,
+      "loss": 1.6526,
+      "step": 7944
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9279427528381348,
+      "learning_rate": 1.2740314327485381e-05,
+      "loss": 1.7349,
+      "step": 7945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6292502880096436,
+      "learning_rate": 1.2739400584795323e-05,
+      "loss": 1.509,
+      "step": 7946
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3378946781158447,
+      "learning_rate": 1.2738486842105264e-05,
+      "loss": 1.5933,
+      "step": 7947
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.717121124267578,
+      "learning_rate": 1.2737573099415206e-05,
+      "loss": 1.7956,
+      "step": 7948
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8788561820983887,
+      "learning_rate": 1.2736659356725147e-05,
+      "loss": 1.3787,
+      "step": 7949
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7991552352905273,
+      "learning_rate": 1.2735745614035089e-05,
+      "loss": 1.7032,
+      "step": 7950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6872398853302,
+      "learning_rate": 1.2734831871345031e-05,
+      "loss": 1.4543,
+      "step": 7951
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0946741104125977,
+      "learning_rate": 1.2733918128654971e-05,
+      "loss": 1.1073,
+      "step": 7952
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.295125961303711,
+      "learning_rate": 1.2733004385964914e-05,
+      "loss": 1.6877,
+      "step": 7953
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.416450262069702,
+      "learning_rate": 1.2732090643274854e-05,
+      "loss": 2.1638,
+      "step": 7954
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4748260974884033,
+      "learning_rate": 1.2731176900584796e-05,
+      "loss": 1.4904,
+      "step": 7955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9930813312530518,
+      "learning_rate": 1.2730263157894737e-05,
+      "loss": 1.4639,
+      "step": 7956
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1137306690216064,
+      "learning_rate": 1.272934941520468e-05,
+      "loss": 1.7128,
+      "step": 7957
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7886803150177,
+      "learning_rate": 1.2728435672514621e-05,
+      "loss": 1.6134,
+      "step": 7958
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7528271675109863,
+      "learning_rate": 1.2727521929824562e-05,
+      "loss": 1.6764,
+      "step": 7959
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.113966226577759,
+      "learning_rate": 1.2726608187134504e-05,
+      "loss": 1.714,
+      "step": 7960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6052777767181396,
+      "learning_rate": 1.2725694444444445e-05,
+      "loss": 1.6433,
+      "step": 7961
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7134692668914795,
+      "learning_rate": 1.2724780701754387e-05,
+      "loss": 1.7025,
+      "step": 7962
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.221642017364502,
+      "learning_rate": 1.2723866959064327e-05,
+      "loss": 1.7871,
+      "step": 7963
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6083168983459473,
+      "learning_rate": 1.272295321637427e-05,
+      "loss": 1.5513,
+      "step": 7964
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.96158766746521,
+      "learning_rate": 1.2722039473684212e-05,
+      "loss": 1.9315,
+      "step": 7965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9917166233062744,
+      "learning_rate": 1.2721125730994152e-05,
+      "loss": 1.9093,
+      "step": 7966
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9426791667938232,
+      "learning_rate": 1.2720211988304095e-05,
+      "loss": 2.0316,
+      "step": 7967
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.480055332183838,
+      "learning_rate": 1.2719298245614035e-05,
+      "loss": 1.7546,
+      "step": 7968
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4916470050811768,
+      "learning_rate": 1.2718384502923977e-05,
+      "loss": 1.5348,
+      "step": 7969
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6260547637939453,
+      "learning_rate": 1.2717470760233918e-05,
+      "loss": 1.5232,
+      "step": 7970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.203312635421753,
+      "learning_rate": 1.271655701754386e-05,
+      "loss": 1.9302,
+      "step": 7971
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5584492683410645,
+      "learning_rate": 1.2715643274853802e-05,
+      "loss": 1.6139,
+      "step": 7972
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.790067434310913,
+      "learning_rate": 1.2714729532163743e-05,
+      "loss": 1.6309,
+      "step": 7973
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.443108081817627,
+      "learning_rate": 1.2713815789473685e-05,
+      "loss": 1.4755,
+      "step": 7974
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6067330837249756,
+      "learning_rate": 1.2712902046783626e-05,
+      "loss": 1.5411,
+      "step": 7975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.57759428024292,
+      "learning_rate": 1.2711988304093568e-05,
+      "loss": 1.7954,
+      "step": 7976
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3927836418151855,
+      "learning_rate": 1.2711074561403508e-05,
+      "loss": 1.2961,
+      "step": 7977
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7114555835723877,
+      "learning_rate": 1.271016081871345e-05,
+      "loss": 1.5024,
+      "step": 7978
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0471670627593994,
+      "learning_rate": 1.2709247076023393e-05,
+      "loss": 1.5184,
+      "step": 7979
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0623550415039062,
+      "learning_rate": 1.2708333333333333e-05,
+      "loss": 1.7758,
+      "step": 7980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6445581912994385,
+      "learning_rate": 1.2707419590643276e-05,
+      "loss": 1.4361,
+      "step": 7981
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9131181240081787,
+      "learning_rate": 1.2706505847953216e-05,
+      "loss": 1.8591,
+      "step": 7982
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.867356538772583,
+      "learning_rate": 1.2705592105263158e-05,
+      "loss": 1.574,
+      "step": 7983
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.671541213989258,
+      "learning_rate": 1.2704678362573099e-05,
+      "loss": 1.5777,
+      "step": 7984
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3042845726013184,
+      "learning_rate": 1.2703764619883041e-05,
+      "loss": 1.0987,
+      "step": 7985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.299999713897705,
+      "learning_rate": 1.2702850877192983e-05,
+      "loss": 1.6493,
+      "step": 7986
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4005236625671387,
+      "learning_rate": 1.2701937134502924e-05,
+      "loss": 1.4477,
+      "step": 7987
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9380457401275635,
+      "learning_rate": 1.2701023391812866e-05,
+      "loss": 1.6959,
+      "step": 7988
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5735533237457275,
+      "learning_rate": 1.2700109649122807e-05,
+      "loss": 1.358,
+      "step": 7989
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.82293438911438,
+      "learning_rate": 1.2699195906432749e-05,
+      "loss": 1.7044,
+      "step": 7990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2087671756744385,
+      "learning_rate": 1.269828216374269e-05,
+      "loss": 1.3191,
+      "step": 7991
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1406915187835693,
+      "learning_rate": 1.2697368421052632e-05,
+      "loss": 1.0463,
+      "step": 7992
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1242449283599854,
+      "learning_rate": 1.2696454678362574e-05,
+      "loss": 1.5882,
+      "step": 7993
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.398907423019409,
+      "learning_rate": 1.2695540935672514e-05,
+      "loss": 1.4062,
+      "step": 7994
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0087168216705322,
+      "learning_rate": 1.2694627192982458e-05,
+      "loss": 1.5966,
+      "step": 7995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1626360416412354,
+      "learning_rate": 1.26937134502924e-05,
+      "loss": 1.2829,
+      "step": 7996
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2681665420532227,
+      "learning_rate": 1.2692799707602341e-05,
+      "loss": 1.151,
+      "step": 7997
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9299333095550537,
+      "learning_rate": 1.2691885964912283e-05,
+      "loss": 1.5851,
+      "step": 7998
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9217116832733154,
+      "learning_rate": 1.2690972222222224e-05,
+      "loss": 1.4299,
+      "step": 7999
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.24238920211792,
+      "learning_rate": 1.2690058479532166e-05,
+      "loss": 1.2886,
+      "step": 8000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.730600595474243,
+      "learning_rate": 1.2689144736842106e-05,
+      "loss": 1.7125,
+      "step": 8001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.865980625152588,
+      "learning_rate": 1.2688230994152049e-05,
+      "loss": 1.8155,
+      "step": 8002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.876528024673462,
+      "learning_rate": 1.2687317251461991e-05,
+      "loss": 1.7783,
+      "step": 8003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.039165496826172,
+      "learning_rate": 1.2686403508771931e-05,
+      "loss": 1.1434,
+      "step": 8004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9788851737976074,
+      "learning_rate": 1.2685489766081874e-05,
+      "loss": 1.7909,
+      "step": 8005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7817654609680176,
+      "learning_rate": 1.2684576023391814e-05,
+      "loss": 1.4361,
+      "step": 8006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8887240886688232,
+      "learning_rate": 1.2683662280701756e-05,
+      "loss": 1.5417,
+      "step": 8007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.763201951980591,
+      "learning_rate": 1.2682748538011697e-05,
+      "loss": 1.6129,
+      "step": 8008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.904157876968384,
+      "learning_rate": 1.2681834795321639e-05,
+      "loss": 1.7737,
+      "step": 8009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8967247009277344,
+      "learning_rate": 1.2680921052631581e-05,
+      "loss": 1.823,
+      "step": 8010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7338197231292725,
+      "learning_rate": 1.2680007309941522e-05,
+      "loss": 1.6955,
+      "step": 8011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5422892570495605,
+      "learning_rate": 1.2679093567251464e-05,
+      "loss": 1.416,
+      "step": 8012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0018439292907715,
+      "learning_rate": 1.2678179824561405e-05,
+      "loss": 1.4822,
+      "step": 8013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8083689212799072,
+      "learning_rate": 1.2677266081871347e-05,
+      "loss": 1.399,
+      "step": 8014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0133776664733887,
+      "learning_rate": 1.2676352339181287e-05,
+      "loss": 1.7998,
+      "step": 8015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7361018657684326,
+      "learning_rate": 1.267543859649123e-05,
+      "loss": 1.6236,
+      "step": 8016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0092082023620605,
+      "learning_rate": 1.2674524853801172e-05,
+      "loss": 1.4088,
+      "step": 8017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6941747665405273,
+      "learning_rate": 1.2673611111111112e-05,
+      "loss": 1.6942,
+      "step": 8018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5223491191864014,
+      "learning_rate": 1.2672697368421055e-05,
+      "loss": 1.4245,
+      "step": 8019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5371079444885254,
+      "learning_rate": 1.2671783625730995e-05,
+      "loss": 1.6209,
+      "step": 8020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.130156993865967,
+      "learning_rate": 1.2670869883040937e-05,
+      "loss": 1.1032,
+      "step": 8021
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.461193561553955,
+      "learning_rate": 1.2669956140350878e-05,
+      "loss": 1.5635,
+      "step": 8022
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6828739643096924,
+      "learning_rate": 1.266904239766082e-05,
+      "loss": 1.4445,
+      "step": 8023
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2642219066619873,
+      "learning_rate": 1.2668128654970762e-05,
+      "loss": 1.3019,
+      "step": 8024
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7335424423217773,
+      "learning_rate": 1.2667214912280703e-05,
+      "loss": 1.4978,
+      "step": 8025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2979977130889893,
+      "learning_rate": 1.2666301169590645e-05,
+      "loss": 1.3399,
+      "step": 8026
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.519227981567383,
+      "learning_rate": 1.2665387426900586e-05,
+      "loss": 1.2818,
+      "step": 8027
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2150564193725586,
+      "learning_rate": 1.2664473684210528e-05,
+      "loss": 1.4221,
+      "step": 8028
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.183807373046875,
+      "learning_rate": 1.2663559941520468e-05,
+      "loss": 1.3138,
+      "step": 8029
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.565214157104492,
+      "learning_rate": 1.266264619883041e-05,
+      "loss": 1.3657,
+      "step": 8030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9252374172210693,
+      "learning_rate": 1.2661732456140353e-05,
+      "loss": 1.5091,
+      "step": 8031
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1859569549560547,
+      "learning_rate": 1.2660818713450293e-05,
+      "loss": 2.0627,
+      "step": 8032
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.380427837371826,
+      "learning_rate": 1.2659904970760236e-05,
+      "loss": 1.2112,
+      "step": 8033
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7389307022094727,
+      "learning_rate": 1.2658991228070176e-05,
+      "loss": 1.3321,
+      "step": 8034
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.265239715576172,
+      "learning_rate": 1.2658077485380118e-05,
+      "loss": 1.3055,
+      "step": 8035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8487799167633057,
+      "learning_rate": 1.2657163742690059e-05,
+      "loss": 1.2118,
+      "step": 8036
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9431216716766357,
+      "learning_rate": 1.2656250000000001e-05,
+      "loss": 1.595,
+      "step": 8037
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.52506422996521,
+      "learning_rate": 1.2655336257309943e-05,
+      "loss": 1.6754,
+      "step": 8038
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5174357891082764,
+      "learning_rate": 1.2654422514619884e-05,
+      "loss": 1.533,
+      "step": 8039
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6249189376831055,
+      "learning_rate": 1.2653508771929826e-05,
+      "loss": 1.3572,
+      "step": 8040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8036162853240967,
+      "learning_rate": 1.2652595029239767e-05,
+      "loss": 1.3821,
+      "step": 8041
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5603084564208984,
+      "learning_rate": 1.2651681286549709e-05,
+      "loss": 1.7137,
+      "step": 8042
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.943087577819824,
+      "learning_rate": 1.265076754385965e-05,
+      "loss": 1.6562,
+      "step": 8043
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3972105979919434,
+      "learning_rate": 1.2649853801169591e-05,
+      "loss": 1.6378,
+      "step": 8044
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.274751663208008,
+      "learning_rate": 1.2648940058479534e-05,
+      "loss": 1.3293,
+      "step": 8045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.498004913330078,
+      "learning_rate": 1.2648026315789474e-05,
+      "loss": 1.5456,
+      "step": 8046
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7442626953125,
+      "learning_rate": 1.2647112573099416e-05,
+      "loss": 1.4338,
+      "step": 8047
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7108170986175537,
+      "learning_rate": 1.2646198830409357e-05,
+      "loss": 1.4922,
+      "step": 8048
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5681376457214355,
+      "learning_rate": 1.26452850877193e-05,
+      "loss": 1.5746,
+      "step": 8049
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.496518850326538,
+      "learning_rate": 1.264437134502924e-05,
+      "loss": 1.1552,
+      "step": 8050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.365817070007324,
+      "learning_rate": 1.2643457602339182e-05,
+      "loss": 1.9322,
+      "step": 8051
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7395713329315186,
+      "learning_rate": 1.2642543859649124e-05,
+      "loss": 1.4541,
+      "step": 8052
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7259178161621094,
+      "learning_rate": 1.2641630116959065e-05,
+      "loss": 1.5988,
+      "step": 8053
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.16549015045166,
+      "learning_rate": 1.2640716374269007e-05,
+      "loss": 1.3584,
+      "step": 8054
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.612252950668335,
+      "learning_rate": 1.2639802631578947e-05,
+      "loss": 1.6339,
+      "step": 8055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.819181442260742,
+      "learning_rate": 1.263888888888889e-05,
+      "loss": 1.8161,
+      "step": 8056
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.636239528656006,
+      "learning_rate": 1.263797514619883e-05,
+      "loss": 1.5749,
+      "step": 8057
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4646644592285156,
+      "learning_rate": 1.2637061403508772e-05,
+      "loss": 1.774,
+      "step": 8058
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4194345474243164,
+      "learning_rate": 1.2636147660818715e-05,
+      "loss": 1.3445,
+      "step": 8059
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.675997018814087,
+      "learning_rate": 1.2635233918128655e-05,
+      "loss": 1.6805,
+      "step": 8060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.773066282272339,
+      "learning_rate": 1.2634320175438597e-05,
+      "loss": 1.791,
+      "step": 8061
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.127121686935425,
+      "learning_rate": 1.2633406432748538e-05,
+      "loss": 1.701,
+      "step": 8062
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1441259384155273,
+      "learning_rate": 1.263249269005848e-05,
+      "loss": 1.8445,
+      "step": 8063
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5167126655578613,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 1.5387,
+      "step": 8064
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.271514415740967,
+      "learning_rate": 1.2630665204678363e-05,
+      "loss": 1.5913,
+      "step": 8065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5453391075134277,
+      "learning_rate": 1.2629751461988305e-05,
+      "loss": 1.5401,
+      "step": 8066
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9772098064422607,
+      "learning_rate": 1.2628837719298246e-05,
+      "loss": 1.6775,
+      "step": 8067
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.625892162322998,
+      "learning_rate": 1.2627923976608188e-05,
+      "loss": 1.645,
+      "step": 8068
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.331740379333496,
+      "learning_rate": 1.2627010233918128e-05,
+      "loss": 1.3046,
+      "step": 8069
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4491078853607178,
+      "learning_rate": 1.262609649122807e-05,
+      "loss": 1.3689,
+      "step": 8070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7659945487976074,
+      "learning_rate": 1.2625182748538011e-05,
+      "loss": 1.3847,
+      "step": 8071
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5288543701171875,
+      "learning_rate": 1.2624269005847953e-05,
+      "loss": 1.6372,
+      "step": 8072
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2751591205596924,
+      "learning_rate": 1.2623355263157896e-05,
+      "loss": 1.4268,
+      "step": 8073
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.161489725112915,
+      "learning_rate": 1.2622441520467836e-05,
+      "loss": 1.7856,
+      "step": 8074
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.695178985595703,
+      "learning_rate": 1.2621527777777778e-05,
+      "loss": 1.4171,
+      "step": 8075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6459553241729736,
+      "learning_rate": 1.2620614035087719e-05,
+      "loss": 1.3068,
+      "step": 8076
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7634050846099854,
+      "learning_rate": 1.2619700292397661e-05,
+      "loss": 1.5796,
+      "step": 8077
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1838889122009277,
+      "learning_rate": 1.2618786549707602e-05,
+      "loss": 1.1916,
+      "step": 8078
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1917335987091064,
+      "learning_rate": 1.2617872807017544e-05,
+      "loss": 1.2664,
+      "step": 8079
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6121509075164795,
+      "learning_rate": 1.2616959064327488e-05,
+      "loss": 1.7868,
+      "step": 8080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.779264211654663,
+      "learning_rate": 1.2616045321637428e-05,
+      "loss": 1.5995,
+      "step": 8081
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.389383316040039,
+      "learning_rate": 1.261513157894737e-05,
+      "loss": 1.6626,
+      "step": 8082
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.263948440551758,
+      "learning_rate": 1.2614217836257313e-05,
+      "loss": 1.2265,
+      "step": 8083
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7784440517425537,
+      "learning_rate": 1.2613304093567253e-05,
+      "loss": 1.4235,
+      "step": 8084
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8768696784973145,
+      "learning_rate": 1.2612390350877195e-05,
+      "loss": 1.7505,
+      "step": 8085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.510568380355835,
+      "learning_rate": 1.2611476608187136e-05,
+      "loss": 1.3651,
+      "step": 8086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.483566999435425,
+      "learning_rate": 1.2610562865497078e-05,
+      "loss": 1.3806,
+      "step": 8087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.847687244415283,
+      "learning_rate": 1.2609649122807019e-05,
+      "loss": 1.6668,
+      "step": 8088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1212551593780518,
+      "learning_rate": 1.2608735380116961e-05,
+      "loss": 1.1982,
+      "step": 8089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.859067440032959,
+      "learning_rate": 1.2607821637426903e-05,
+      "loss": 1.4227,
+      "step": 8090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1548290252685547,
+      "learning_rate": 1.2606907894736844e-05,
+      "loss": 1.9563,
+      "step": 8091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.444758415222168,
+      "learning_rate": 1.2605994152046786e-05,
+      "loss": 1.585,
+      "step": 8092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.632415533065796,
+      "learning_rate": 1.2605080409356726e-05,
+      "loss": 1.2004,
+      "step": 8093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.795752763748169,
+      "learning_rate": 1.2604166666666669e-05,
+      "loss": 1.2487,
+      "step": 8094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0725321769714355,
+      "learning_rate": 1.260325292397661e-05,
+      "loss": 1.6249,
+      "step": 8095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4732000827789307,
+      "learning_rate": 1.2602339181286551e-05,
+      "loss": 1.4579,
+      "step": 8096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2738492488861084,
+      "learning_rate": 1.2601425438596494e-05,
+      "loss": 1.2513,
+      "step": 8097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.971439838409424,
+      "learning_rate": 1.2600511695906434e-05,
+      "loss": 1.8602,
+      "step": 8098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.143794298171997,
+      "learning_rate": 1.2599597953216376e-05,
+      "loss": 1.9496,
+      "step": 8099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.173475742340088,
+      "learning_rate": 1.2598684210526317e-05,
+      "loss": 1.5835,
+      "step": 8100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.013767719268799,
+      "learning_rate": 1.2597770467836259e-05,
+      "loss": 1.192,
+      "step": 8101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0521206855773926,
+      "learning_rate": 1.25968567251462e-05,
+      "loss": 1.6087,
+      "step": 8102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.917099952697754,
+      "learning_rate": 1.2595942982456142e-05,
+      "loss": 1.3473,
+      "step": 8103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6072299480438232,
+      "learning_rate": 1.2595029239766084e-05,
+      "loss": 1.1473,
+      "step": 8104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1706976890563965,
+      "learning_rate": 1.2594115497076025e-05,
+      "loss": 1.9509,
+      "step": 8105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.601316452026367,
+      "learning_rate": 1.2593201754385967e-05,
+      "loss": 1.4867,
+      "step": 8106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8731861114501953,
+      "learning_rate": 1.2592288011695907e-05,
+      "loss": 1.6979,
+      "step": 8107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.726813793182373,
+      "learning_rate": 1.259137426900585e-05,
+      "loss": 1.735,
+      "step": 8108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.765336513519287,
+      "learning_rate": 1.259046052631579e-05,
+      "loss": 1.4613,
+      "step": 8109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.988528251647949,
+      "learning_rate": 1.2589546783625732e-05,
+      "loss": 1.6827,
+      "step": 8110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.011444091796875,
+      "learning_rate": 1.2588633040935675e-05,
+      "loss": 1.671,
+      "step": 8111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0310652256011963,
+      "learning_rate": 1.2587719298245615e-05,
+      "loss": 1.8367,
+      "step": 8112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8133139610290527,
+      "learning_rate": 1.2586805555555557e-05,
+      "loss": 1.3675,
+      "step": 8113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1773552894592285,
+      "learning_rate": 1.2585891812865498e-05,
+      "loss": 1.8277,
+      "step": 8114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.048384189605713,
+      "learning_rate": 1.258497807017544e-05,
+      "loss": 1.6037,
+      "step": 8115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7513389587402344,
+      "learning_rate": 1.258406432748538e-05,
+      "loss": 1.5891,
+      "step": 8116
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5722131729125977,
+      "learning_rate": 1.2583150584795323e-05,
+      "loss": 1.3278,
+      "step": 8117
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.590395212173462,
+      "learning_rate": 1.2582236842105265e-05,
+      "loss": 1.2474,
+      "step": 8118
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.648035764694214,
+      "learning_rate": 1.2581323099415206e-05,
+      "loss": 1.5573,
+      "step": 8119
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8429791927337646,
+      "learning_rate": 1.2580409356725148e-05,
+      "loss": 1.4222,
+      "step": 8120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.430262565612793,
+      "learning_rate": 1.2579495614035088e-05,
+      "loss": 1.8721,
+      "step": 8121
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.175792694091797,
+      "learning_rate": 1.257858187134503e-05,
+      "loss": 2.0297,
+      "step": 8122
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0925703048706055,
+      "learning_rate": 1.2577668128654971e-05,
+      "loss": 1.6971,
+      "step": 8123
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1650338172912598,
+      "learning_rate": 1.2576754385964913e-05,
+      "loss": 1.4159,
+      "step": 8124
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7437024116516113,
+      "learning_rate": 1.2575840643274856e-05,
+      "loss": 1.626,
+      "step": 8125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1195871829986572,
+      "learning_rate": 1.2574926900584796e-05,
+      "loss": 1.5523,
+      "step": 8126
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6826717853546143,
+      "learning_rate": 1.2574013157894738e-05,
+      "loss": 1.3541,
+      "step": 8127
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4676218032836914,
+      "learning_rate": 1.2573099415204679e-05,
+      "loss": 1.5754,
+      "step": 8128
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.34503436088562,
+      "learning_rate": 1.2572185672514621e-05,
+      "loss": 1.5244,
+      "step": 8129
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.489835262298584,
+      "learning_rate": 1.2571271929824562e-05,
+      "loss": 1.3673,
+      "step": 8130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5711724758148193,
+      "learning_rate": 1.2570358187134504e-05,
+      "loss": 1.9007,
+      "step": 8131
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6625659465789795,
+      "learning_rate": 1.2569444444444446e-05,
+      "loss": 1.3983,
+      "step": 8132
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.891292095184326,
+      "learning_rate": 1.2568530701754387e-05,
+      "loss": 1.628,
+      "step": 8133
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.966123580932617,
+      "learning_rate": 1.2567616959064329e-05,
+      "loss": 1.5281,
+      "step": 8134
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.674701690673828,
+      "learning_rate": 1.256670321637427e-05,
+      "loss": 1.5938,
+      "step": 8135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.331270217895508,
+      "learning_rate": 1.2565789473684211e-05,
+      "loss": 1.5457,
+      "step": 8136
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9276773929595947,
+      "learning_rate": 1.2564875730994152e-05,
+      "loss": 1.4527,
+      "step": 8137
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9285428524017334,
+      "learning_rate": 1.2563961988304094e-05,
+      "loss": 1.733,
+      "step": 8138
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.997685194015503,
+      "learning_rate": 1.2563048245614036e-05,
+      "loss": 1.7564,
+      "step": 8139
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9908907413482666,
+      "learning_rate": 1.2562134502923977e-05,
+      "loss": 1.7257,
+      "step": 8140
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4235825538635254,
+      "learning_rate": 1.256122076023392e-05,
+      "loss": 1.5293,
+      "step": 8141
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8322317600250244,
+      "learning_rate": 1.256030701754386e-05,
+      "loss": 1.6047,
+      "step": 8142
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1104722023010254,
+      "learning_rate": 1.2559393274853802e-05,
+      "loss": 2.0924,
+      "step": 8143
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.961416482925415,
+      "learning_rate": 1.2558479532163742e-05,
+      "loss": 1.8533,
+      "step": 8144
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8567445278167725,
+      "learning_rate": 1.2557565789473685e-05,
+      "loss": 1.6152,
+      "step": 8145
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6917171478271484,
+      "learning_rate": 1.2556652046783627e-05,
+      "loss": 1.6454,
+      "step": 8146
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.800290584564209,
+      "learning_rate": 1.2555738304093567e-05,
+      "loss": 1.7804,
+      "step": 8147
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.392301559448242,
+      "learning_rate": 1.255482456140351e-05,
+      "loss": 1.2592,
+      "step": 8148
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9851362705230713,
+      "learning_rate": 1.255391081871345e-05,
+      "loss": 1.6074,
+      "step": 8149
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8781893253326416,
+      "learning_rate": 1.2552997076023392e-05,
+      "loss": 1.5105,
+      "step": 8150
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1019880771636963,
+      "learning_rate": 1.2552083333333333e-05,
+      "loss": 1.675,
+      "step": 8151
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6554746627807617,
+      "learning_rate": 1.2551169590643275e-05,
+      "loss": 1.5131,
+      "step": 8152
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.755531072616577,
+      "learning_rate": 1.2550255847953217e-05,
+      "loss": 1.4127,
+      "step": 8153
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.270852565765381,
+      "learning_rate": 1.2549342105263158e-05,
+      "loss": 1.3603,
+      "step": 8154
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8178725242614746,
+      "learning_rate": 1.25484283625731e-05,
+      "loss": 1.6125,
+      "step": 8155
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4719431400299072,
+      "learning_rate": 1.254751461988304e-05,
+      "loss": 1.2712,
+      "step": 8156
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.89611554145813,
+      "learning_rate": 1.2546600877192983e-05,
+      "loss": 1.6599,
+      "step": 8157
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7969696521759033,
+      "learning_rate": 1.2545687134502923e-05,
+      "loss": 1.5325,
+      "step": 8158
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.294933795928955,
+      "learning_rate": 1.2544773391812866e-05,
+      "loss": 1.5209,
+      "step": 8159
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1566944122314453,
+      "learning_rate": 1.2543859649122808e-05,
+      "loss": 1.6644,
+      "step": 8160
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.285681962966919,
+      "learning_rate": 1.2542945906432748e-05,
+      "loss": 2.0415,
+      "step": 8161
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.128843307495117,
+      "learning_rate": 1.254203216374269e-05,
+      "loss": 1.3323,
+      "step": 8162
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0083444118499756,
+      "learning_rate": 1.2541118421052631e-05,
+      "loss": 1.8521,
+      "step": 8163
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8367748260498047,
+      "learning_rate": 1.2540204678362573e-05,
+      "loss": 1.7033,
+      "step": 8164
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.938167095184326,
+      "learning_rate": 1.2539290935672514e-05,
+      "loss": 1.7008,
+      "step": 8165
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0920255184173584,
+      "learning_rate": 1.2538377192982458e-05,
+      "loss": 1.5972,
+      "step": 8166
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5088350772857666,
+      "learning_rate": 1.25374634502924e-05,
+      "loss": 1.5037,
+      "step": 8167
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.655496597290039,
+      "learning_rate": 1.253654970760234e-05,
+      "loss": 1.3199,
+      "step": 8168
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.16664457321167,
+      "learning_rate": 1.2535635964912283e-05,
+      "loss": 1.5133,
+      "step": 8169
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2674560546875,
+      "learning_rate": 1.2534722222222225e-05,
+      "loss": 1.2222,
+      "step": 8170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5368030071258545,
+      "learning_rate": 1.2533808479532166e-05,
+      "loss": 1.6623,
+      "step": 8171
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2279393672943115,
+      "learning_rate": 1.2532894736842108e-05,
+      "loss": 1.3785,
+      "step": 8172
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.938812494277954,
+      "learning_rate": 1.2531980994152048e-05,
+      "loss": 1.7811,
+      "step": 8173
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2656118869781494,
+      "learning_rate": 1.253106725146199e-05,
+      "loss": 1.3112,
+      "step": 8174
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.762216329574585,
+      "learning_rate": 1.2530153508771931e-05,
+      "loss": 1.3784,
+      "step": 8175
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.220674991607666,
+      "learning_rate": 1.2529239766081873e-05,
+      "loss": 1.2042,
+      "step": 8176
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.378579616546631,
+      "learning_rate": 1.2528326023391815e-05,
+      "loss": 1.2731,
+      "step": 8177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.686035394668579,
+      "learning_rate": 1.2527412280701756e-05,
+      "loss": 1.5017,
+      "step": 8178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0370211601257324,
+      "learning_rate": 1.2526498538011698e-05,
+      "loss": 2.0177,
+      "step": 8179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6876869201660156,
+      "learning_rate": 1.2525584795321639e-05,
+      "loss": 1.8107,
+      "step": 8180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.464715003967285,
+      "learning_rate": 1.2524671052631581e-05,
+      "loss": 1.4802,
+      "step": 8181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.732727289199829,
+      "learning_rate": 1.2523757309941521e-05,
+      "loss": 1.6975,
+      "step": 8182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8523600101470947,
+      "learning_rate": 1.2522843567251464e-05,
+      "loss": 1.7097,
+      "step": 8183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.281304359436035,
+      "learning_rate": 1.2521929824561406e-05,
+      "loss": 1.1355,
+      "step": 8184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8979499340057373,
+      "learning_rate": 1.2521016081871346e-05,
+      "loss": 1.5956,
+      "step": 8185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7391114234924316,
+      "learning_rate": 1.2520102339181289e-05,
+      "loss": 1.8242,
+      "step": 8186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2006640434265137,
+      "learning_rate": 1.251918859649123e-05,
+      "loss": 1.303,
+      "step": 8187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0960566997528076,
+      "learning_rate": 1.2518274853801171e-05,
+      "loss": 1.7609,
+      "step": 8188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2436165809631348,
+      "learning_rate": 1.2517361111111112e-05,
+      "loss": 1.2259,
+      "step": 8189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3291897773742676,
+      "learning_rate": 1.2516447368421054e-05,
+      "loss": 1.545,
+      "step": 8190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.637833833694458,
+      "learning_rate": 1.2515533625730996e-05,
+      "loss": 1.4182,
+      "step": 8191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7403409481048584,
+      "learning_rate": 1.2514619883040937e-05,
+      "loss": 1.7128,
+      "step": 8192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.079742670059204,
+      "learning_rate": 1.2513706140350879e-05,
+      "loss": 1.1145,
+      "step": 8193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.425116777420044,
+      "learning_rate": 1.251279239766082e-05,
+      "loss": 1.3481,
+      "step": 8194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6509366035461426,
+      "learning_rate": 1.2511878654970762e-05,
+      "loss": 1.4687,
+      "step": 8195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.745687484741211,
+      "learning_rate": 1.2510964912280702e-05,
+      "loss": 1.678,
+      "step": 8196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6786274909973145,
+      "learning_rate": 1.2510051169590645e-05,
+      "loss": 1.5771,
+      "step": 8197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1271352767944336,
+      "learning_rate": 1.2509137426900587e-05,
+      "loss": 1.5815,
+      "step": 8198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.343531370162964,
+      "learning_rate": 1.2508223684210527e-05,
+      "loss": 1.4353,
+      "step": 8199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0505363941192627,
+      "learning_rate": 1.250730994152047e-05,
+      "loss": 1.553,
+      "step": 8200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4979634284973145,
+      "learning_rate": 1.250639619883041e-05,
+      "loss": 1.75,
+      "step": 8201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5622313022613525,
+      "learning_rate": 1.2505482456140352e-05,
+      "loss": 1.6169,
+      "step": 8202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2149040699005127,
+      "learning_rate": 1.2504568713450293e-05,
+      "loss": 1.3455,
+      "step": 8203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7410008907318115,
+      "learning_rate": 1.2503654970760235e-05,
+      "loss": 1.7131,
+      "step": 8204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9817423820495605,
+      "learning_rate": 1.2502741228070177e-05,
+      "loss": 1.3312,
+      "step": 8205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9102516174316406,
+      "learning_rate": 1.2501827485380118e-05,
+      "loss": 1.62,
+      "step": 8206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.54616117477417,
+      "learning_rate": 1.250091374269006e-05,
+      "loss": 1.4025,
+      "step": 8207
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.729079246520996,
+      "learning_rate": 1.25e-05,
+      "loss": 1.4733,
+      "step": 8208
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5134775638580322,
+      "learning_rate": 1.2499086257309943e-05,
+      "loss": 1.5149,
+      "step": 8209
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.786808967590332,
+      "learning_rate": 1.2498172514619883e-05,
+      "loss": 1.6647,
+      "step": 8210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9137723445892334,
+      "learning_rate": 1.2497258771929826e-05,
+      "loss": 1.4828,
+      "step": 8211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5263054370880127,
+      "learning_rate": 1.2496345029239768e-05,
+      "loss": 1.2763,
+      "step": 8212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.187192440032959,
+      "learning_rate": 1.2495431286549708e-05,
+      "loss": 1.8489,
+      "step": 8213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.654317617416382,
+      "learning_rate": 1.249451754385965e-05,
+      "loss": 1.8966,
+      "step": 8214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8170928955078125,
+      "learning_rate": 1.2493603801169591e-05,
+      "loss": 1.7579,
+      "step": 8215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.361114740371704,
+      "learning_rate": 1.2492690058479533e-05,
+      "loss": 1.4014,
+      "step": 8216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5741517543792725,
+      "learning_rate": 1.2491776315789474e-05,
+      "loss": 1.5859,
+      "step": 8217
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7003283500671387,
+      "learning_rate": 1.2490862573099416e-05,
+      "loss": 1.5541,
+      "step": 8218
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.506391763687134,
+      "learning_rate": 1.2489948830409358e-05,
+      "loss": 1.4534,
+      "step": 8219
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1085643768310547,
+      "learning_rate": 1.2489035087719299e-05,
+      "loss": 1.1963,
+      "step": 8220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5932719707489014,
+      "learning_rate": 1.2488121345029241e-05,
+      "loss": 1.4841,
+      "step": 8221
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.989351987838745,
+      "learning_rate": 1.2487207602339182e-05,
+      "loss": 1.6991,
+      "step": 8222
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2002923488616943,
+      "learning_rate": 1.2486293859649124e-05,
+      "loss": 1.8289,
+      "step": 8223
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.939478874206543,
+      "learning_rate": 1.2485380116959064e-05,
+      "loss": 1.556,
+      "step": 8224
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7453298568725586,
+      "learning_rate": 1.2484466374269007e-05,
+      "loss": 1.7457,
+      "step": 8225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3584816455841064,
+      "learning_rate": 1.2483552631578949e-05,
+      "loss": 1.171,
+      "step": 8226
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.018049716949463,
+      "learning_rate": 1.248263888888889e-05,
+      "loss": 1.2357,
+      "step": 8227
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.779153823852539,
+      "learning_rate": 1.2481725146198831e-05,
+      "loss": 1.5348,
+      "step": 8228
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.460935115814209,
+      "learning_rate": 1.2480811403508772e-05,
+      "loss": 1.3919,
+      "step": 8229
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.734997034072876,
+      "learning_rate": 1.2479897660818714e-05,
+      "loss": 1.482,
+      "step": 8230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5273256301879883,
+      "learning_rate": 1.2478983918128655e-05,
+      "loss": 1.5487,
+      "step": 8231
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0520100593566895,
+      "learning_rate": 1.2478070175438597e-05,
+      "loss": 1.9857,
+      "step": 8232
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2512638568878174,
+      "learning_rate": 1.247715643274854e-05,
+      "loss": 1.9035,
+      "step": 8233
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8189570903778076,
+      "learning_rate": 1.247624269005848e-05,
+      "loss": 1.552,
+      "step": 8234
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.707613468170166,
+      "learning_rate": 1.2475328947368422e-05,
+      "loss": 1.4494,
+      "step": 8235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.892522096633911,
+      "learning_rate": 1.2474415204678362e-05,
+      "loss": 1.6082,
+      "step": 8236
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.621288299560547,
+      "learning_rate": 1.2473501461988305e-05,
+      "loss": 1.6388,
+      "step": 8237
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5469377040863037,
+      "learning_rate": 1.2472587719298245e-05,
+      "loss": 1.3968,
+      "step": 8238
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.546513319015503,
+      "learning_rate": 1.2471673976608187e-05,
+      "loss": 1.4684,
+      "step": 8239
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.836881637573242,
+      "learning_rate": 1.247076023391813e-05,
+      "loss": 1.6576,
+      "step": 8240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.73445200920105,
+      "learning_rate": 1.246984649122807e-05,
+      "loss": 1.7758,
+      "step": 8241
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0668065547943115,
+      "learning_rate": 1.2468932748538012e-05,
+      "loss": 2.044,
+      "step": 8242
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.919370412826538,
+      "learning_rate": 1.2468019005847953e-05,
+      "loss": 1.7123,
+      "step": 8243
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9072751998901367,
+      "learning_rate": 1.2467105263157895e-05,
+      "loss": 1.488,
+      "step": 8244
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1536834239959717,
+      "learning_rate": 1.2466191520467836e-05,
+      "loss": 1.2659,
+      "step": 8245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.810152530670166,
+      "learning_rate": 1.2465277777777778e-05,
+      "loss": 1.7222,
+      "step": 8246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.500800132751465,
+      "learning_rate": 1.246436403508772e-05,
+      "loss": 1.5256,
+      "step": 8247
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.528071403503418,
+      "learning_rate": 1.246345029239766e-05,
+      "loss": 1.5361,
+      "step": 8248
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.481586456298828,
+      "learning_rate": 1.2462536549707603e-05,
+      "loss": 1.5105,
+      "step": 8249
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7070579528808594,
+      "learning_rate": 1.2461622807017543e-05,
+      "loss": 1.6621,
+      "step": 8250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2980964183807373,
+      "learning_rate": 1.2460709064327487e-05,
+      "loss": 1.7715,
+      "step": 8251
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6759960651397705,
+      "learning_rate": 1.245979532163743e-05,
+      "loss": 1.6064,
+      "step": 8252
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.591207981109619,
+      "learning_rate": 1.245888157894737e-05,
+      "loss": 1.8367,
+      "step": 8253
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2183263301849365,
+      "learning_rate": 1.2457967836257312e-05,
+      "loss": 1.7688,
+      "step": 8254
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.487881660461426,
+      "learning_rate": 1.2457054093567253e-05,
+      "loss": 1.4905,
+      "step": 8255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6727852821350098,
+      "learning_rate": 1.2456140350877195e-05,
+      "loss": 1.479,
+      "step": 8256
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.117043972015381,
+      "learning_rate": 1.2455226608187137e-05,
+      "loss": 1.3435,
+      "step": 8257
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4299888610839844,
+      "learning_rate": 1.2454312865497078e-05,
+      "loss": 2.0283,
+      "step": 8258
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.714362859725952,
+      "learning_rate": 1.245339912280702e-05,
+      "loss": 1.9779,
+      "step": 8259
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4504151344299316,
+      "learning_rate": 1.245248538011696e-05,
+      "loss": 1.6729,
+      "step": 8260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.625356912612915,
+      "learning_rate": 1.2451571637426903e-05,
+      "loss": 1.4845,
+      "step": 8261
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8048009872436523,
+      "learning_rate": 1.2450657894736843e-05,
+      "loss": 1.481,
+      "step": 8262
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2848145961761475,
+      "learning_rate": 1.2449744152046786e-05,
+      "loss": 1.8976,
+      "step": 8263
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6034531593322754,
+      "learning_rate": 1.2448830409356728e-05,
+      "loss": 1.4925,
+      "step": 8264
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1117756366729736,
+      "learning_rate": 1.2447916666666668e-05,
+      "loss": 1.5877,
+      "step": 8265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.665708065032959,
+      "learning_rate": 1.244700292397661e-05,
+      "loss": 1.5781,
+      "step": 8266
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5314106941223145,
+      "learning_rate": 1.2446089181286551e-05,
+      "loss": 1.3733,
+      "step": 8267
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6196558475494385,
+      "learning_rate": 1.2445175438596493e-05,
+      "loss": 1.458,
+      "step": 8268
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1132891178131104,
+      "learning_rate": 1.2444261695906434e-05,
+      "loss": 1.6243,
+      "step": 8269
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5395612716674805,
+      "learning_rate": 1.2443347953216376e-05,
+      "loss": 1.5832,
+      "step": 8270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.779588460922241,
+      "learning_rate": 1.2442434210526318e-05,
+      "loss": 1.6753,
+      "step": 8271
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2894322872161865,
+      "learning_rate": 1.2441520467836259e-05,
+      "loss": 1.6344,
+      "step": 8272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.229726791381836,
+      "learning_rate": 1.2440606725146201e-05,
+      "loss": 1.6438,
+      "step": 8273
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8037400245666504,
+      "learning_rate": 1.2439692982456141e-05,
+      "loss": 1.2786,
+      "step": 8274
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.433846950531006,
+      "learning_rate": 1.2438779239766084e-05,
+      "loss": 1.2848,
+      "step": 8275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.942544460296631,
+      "learning_rate": 1.2437865497076024e-05,
+      "loss": 1.4217,
+      "step": 8276
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2198078632354736,
+      "learning_rate": 1.2436951754385966e-05,
+      "loss": 1.7717,
+      "step": 8277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.390033006668091,
+      "learning_rate": 1.2436038011695909e-05,
+      "loss": 1.5346,
+      "step": 8278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4720427989959717,
+      "learning_rate": 1.243512426900585e-05,
+      "loss": 1.5248,
+      "step": 8279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.556692123413086,
+      "learning_rate": 1.2434210526315791e-05,
+      "loss": 2.0198,
+      "step": 8280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2458057403564453,
+      "learning_rate": 1.2433296783625732e-05,
+      "loss": 1.734,
+      "step": 8281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6503660678863525,
+      "learning_rate": 1.2432383040935674e-05,
+      "loss": 1.4708,
+      "step": 8282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.618187189102173,
+      "learning_rate": 1.2431469298245615e-05,
+      "loss": 1.445,
+      "step": 8283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6529407501220703,
+      "learning_rate": 1.2430555555555557e-05,
+      "loss": 1.6894,
+      "step": 8284
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5838804244995117,
+      "learning_rate": 1.2429641812865499e-05,
+      "loss": 2.0896,
+      "step": 8285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.299445629119873,
+      "learning_rate": 1.242872807017544e-05,
+      "loss": 1.5254,
+      "step": 8286
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.733755588531494,
+      "learning_rate": 1.2427814327485382e-05,
+      "loss": 1.5899,
+      "step": 8287
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4291560649871826,
+      "learning_rate": 1.2426900584795322e-05,
+      "loss": 1.3626,
+      "step": 8288
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.893678665161133,
+      "learning_rate": 1.2425986842105265e-05,
+      "loss": 1.5227,
+      "step": 8289
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6884078979492188,
+      "learning_rate": 1.2425073099415205e-05,
+      "loss": 1.6406,
+      "step": 8290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.976762056350708,
+      "learning_rate": 1.2424159356725147e-05,
+      "loss": 1.8177,
+      "step": 8291
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5298609733581543,
+      "learning_rate": 1.242324561403509e-05,
+      "loss": 1.4235,
+      "step": 8292
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8902032375335693,
+      "learning_rate": 1.242233187134503e-05,
+      "loss": 1.6195,
+      "step": 8293
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0950074195861816,
+      "learning_rate": 1.2421418128654972e-05,
+      "loss": 1.3551,
+      "step": 8294
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.984520435333252,
+      "learning_rate": 1.2420504385964913e-05,
+      "loss": 1.4199,
+      "step": 8295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3495547771453857,
+      "learning_rate": 1.2419590643274855e-05,
+      "loss": 1.5815,
+      "step": 8296
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7740373611450195,
+      "learning_rate": 1.2418676900584796e-05,
+      "loss": 1.5532,
+      "step": 8297
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.784376859664917,
+      "learning_rate": 1.2417763157894738e-05,
+      "loss": 1.9055,
+      "step": 8298
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9099924564361572,
+      "learning_rate": 1.241684941520468e-05,
+      "loss": 1.7724,
+      "step": 8299
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.391413688659668,
+      "learning_rate": 1.241593567251462e-05,
+      "loss": 1.2233,
+      "step": 8300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0066535472869873,
+      "learning_rate": 1.2415021929824563e-05,
+      "loss": 2.0611,
+      "step": 8301
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7466330528259277,
+      "learning_rate": 1.2414108187134503e-05,
+      "loss": 1.5579,
+      "step": 8302
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.520087718963623,
+      "learning_rate": 1.2413194444444446e-05,
+      "loss": 1.2256,
+      "step": 8303
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.7850341796875,
+      "learning_rate": 1.2412280701754386e-05,
+      "loss": 1.761,
+      "step": 8304
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7043497562408447,
+      "learning_rate": 1.2411366959064328e-05,
+      "loss": 1.8953,
+      "step": 8305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.392854690551758,
+      "learning_rate": 1.241045321637427e-05,
+      "loss": 1.4334,
+      "step": 8306
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.685856342315674,
+      "learning_rate": 1.2409539473684211e-05,
+      "loss": 1.535,
+      "step": 8307
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.408770799636841,
+      "learning_rate": 1.2408625730994153e-05,
+      "loss": 1.5238,
+      "step": 8308
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.58583927154541,
+      "learning_rate": 1.2407711988304094e-05,
+      "loss": 1.4726,
+      "step": 8309
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8099842071533203,
+      "learning_rate": 1.2406798245614036e-05,
+      "loss": 1.7121,
+      "step": 8310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.814199924468994,
+      "learning_rate": 1.2405884502923977e-05,
+      "loss": 1.4028,
+      "step": 8311
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.646675109863281,
+      "learning_rate": 1.2404970760233919e-05,
+      "loss": 1.5883,
+      "step": 8312
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.139082431793213,
+      "learning_rate": 1.240405701754386e-05,
+      "loss": 1.8131,
+      "step": 8313
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.153630495071411,
+      "learning_rate": 1.2403143274853802e-05,
+      "loss": 1.477,
+      "step": 8314
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.221525192260742,
+      "learning_rate": 1.2402229532163744e-05,
+      "loss": 1.5785,
+      "step": 8315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.30260968208313,
+      "learning_rate": 1.2401315789473684e-05,
+      "loss": 1.62,
+      "step": 8316
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.716554880142212,
+      "learning_rate": 1.2400402046783627e-05,
+      "loss": 1.3784,
+      "step": 8317
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1825053691864014,
+      "learning_rate": 1.2399488304093567e-05,
+      "loss": 1.925,
+      "step": 8318
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.196760892868042,
+      "learning_rate": 1.239857456140351e-05,
+      "loss": 1.2171,
+      "step": 8319
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.839505434036255,
+      "learning_rate": 1.239766081871345e-05,
+      "loss": 1.8135,
+      "step": 8320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.694539785385132,
+      "learning_rate": 1.2396747076023392e-05,
+      "loss": 1.7281,
+      "step": 8321
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.238361358642578,
+      "learning_rate": 1.2395833333333334e-05,
+      "loss": 1.1398,
+      "step": 8322
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.537870168685913,
+      "learning_rate": 1.2394919590643275e-05,
+      "loss": 1.4522,
+      "step": 8323
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8981783390045166,
+      "learning_rate": 1.2394005847953217e-05,
+      "loss": 1.7743,
+      "step": 8324
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0276594161987305,
+      "learning_rate": 1.2393092105263158e-05,
+      "loss": 1.4633,
+      "step": 8325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5003278255462646,
+      "learning_rate": 1.23921783625731e-05,
+      "loss": 1.619,
+      "step": 8326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.453066349029541,
+      "learning_rate": 1.239126461988304e-05,
+      "loss": 1.6076,
+      "step": 8327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7164711952209473,
+      "learning_rate": 1.2390350877192982e-05,
+      "loss": 1.485,
+      "step": 8328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.280590534210205,
+      "learning_rate": 1.2389437134502925e-05,
+      "loss": 1.4679,
+      "step": 8329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0511038303375244,
+      "learning_rate": 1.2388523391812865e-05,
+      "loss": 1.7814,
+      "step": 8330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.484670639038086,
+      "learning_rate": 1.2387609649122807e-05,
+      "loss": 1.5125,
+      "step": 8331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.332934856414795,
+      "learning_rate": 1.2386695906432748e-05,
+      "loss": 1.381,
+      "step": 8332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4766273498535156,
+      "learning_rate": 1.238578216374269e-05,
+      "loss": 1.6002,
+      "step": 8333
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.538241386413574,
+      "learning_rate": 1.238486842105263e-05,
+      "loss": 1.6892,
+      "step": 8334
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3560471534729004,
+      "learning_rate": 1.2383954678362573e-05,
+      "loss": 1.5713,
+      "step": 8335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4278812408447266,
+      "learning_rate": 1.2383040935672515e-05,
+      "loss": 1.1578,
+      "step": 8336
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.054616689682007,
+      "learning_rate": 1.2382127192982459e-05,
+      "loss": 1.7209,
+      "step": 8337
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.879956007003784,
+      "learning_rate": 1.23812134502924e-05,
+      "loss": 1.3852,
+      "step": 8338
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.941581964492798,
+      "learning_rate": 1.2380299707602342e-05,
+      "loss": 1.5869,
+      "step": 8339
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.116544723510742,
+      "learning_rate": 1.2379385964912282e-05,
+      "loss": 1.903,
+      "step": 8340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.523301839828491,
+      "learning_rate": 1.2378472222222225e-05,
+      "loss": 1.4269,
+      "step": 8341
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.946779489517212,
+      "learning_rate": 1.2377558479532165e-05,
+      "loss": 1.6711,
+      "step": 8342
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.859717845916748,
+      "learning_rate": 1.2376644736842107e-05,
+      "loss": 1.7295,
+      "step": 8343
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.232398271560669,
+      "learning_rate": 1.237573099415205e-05,
+      "loss": 1.5671,
+      "step": 8344
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.577430248260498,
+      "learning_rate": 1.237481725146199e-05,
+      "loss": 1.3599,
+      "step": 8345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7558085918426514,
+      "learning_rate": 1.2373903508771932e-05,
+      "loss": 1.4373,
+      "step": 8346
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.740610122680664,
+      "learning_rate": 1.2372989766081873e-05,
+      "loss": 1.6172,
+      "step": 8347
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4598612785339355,
+      "learning_rate": 1.2372076023391815e-05,
+      "loss": 1.442,
+      "step": 8348
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0686733722686768,
+      "learning_rate": 1.2371162280701756e-05,
+      "loss": 1.8145,
+      "step": 8349
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.854588508605957,
+      "learning_rate": 1.2370248538011698e-05,
+      "loss": 1.6331,
+      "step": 8350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.769059419631958,
+      "learning_rate": 1.236933479532164e-05,
+      "loss": 1.9029,
+      "step": 8351
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5196216106414795,
+      "learning_rate": 1.236842105263158e-05,
+      "loss": 1.1819,
+      "step": 8352
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9148900508880615,
+      "learning_rate": 1.2367507309941523e-05,
+      "loss": 1.7945,
+      "step": 8353
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8288493156433105,
+      "learning_rate": 1.2366593567251463e-05,
+      "loss": 1.8562,
+      "step": 8354
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.412243366241455,
+      "learning_rate": 1.2365679824561406e-05,
+      "loss": 1.4748,
+      "step": 8355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9277684688568115,
+      "learning_rate": 1.2364766081871346e-05,
+      "loss": 1.8145,
+      "step": 8356
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.424304246902466,
+      "learning_rate": 1.2363852339181288e-05,
+      "loss": 1.7697,
+      "step": 8357
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1628520488739014,
+      "learning_rate": 1.2362938596491229e-05,
+      "loss": 1.2355,
+      "step": 8358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.62787127494812,
+      "learning_rate": 1.2362024853801171e-05,
+      "loss": 1.4114,
+      "step": 8359
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0173492431640625,
+      "learning_rate": 1.2361111111111113e-05,
+      "loss": 2.0993,
+      "step": 8360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7220568656921387,
+      "learning_rate": 1.2360197368421054e-05,
+      "loss": 1.9017,
+      "step": 8361
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8983638286590576,
+      "learning_rate": 1.2359283625730996e-05,
+      "loss": 1.8692,
+      "step": 8362
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.817873239517212,
+      "learning_rate": 1.2358369883040937e-05,
+      "loss": 1.2466,
+      "step": 8363
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8331615924835205,
+      "learning_rate": 1.2357456140350879e-05,
+      "loss": 1.1258,
+      "step": 8364
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8323609828948975,
+      "learning_rate": 1.235654239766082e-05,
+      "loss": 1.399,
+      "step": 8365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.701016902923584,
+      "learning_rate": 1.2355628654970761e-05,
+      "loss": 1.5689,
+      "step": 8366
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.729029893875122,
+      "learning_rate": 1.2354714912280704e-05,
+      "loss": 1.653,
+      "step": 8367
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.976562023162842,
+      "learning_rate": 1.2353801169590644e-05,
+      "loss": 1.8168,
+      "step": 8368
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7311761379241943,
+      "learning_rate": 1.2352887426900586e-05,
+      "loss": 1.489,
+      "step": 8369
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.160273551940918,
+      "learning_rate": 1.2351973684210527e-05,
+      "loss": 1.0705,
+      "step": 8370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.269576072692871,
+      "learning_rate": 1.235105994152047e-05,
+      "loss": 2.0141,
+      "step": 8371
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5896718502044678,
+      "learning_rate": 1.235014619883041e-05,
+      "loss": 1.3382,
+      "step": 8372
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7918806076049805,
+      "learning_rate": 1.2349232456140352e-05,
+      "loss": 1.8277,
+      "step": 8373
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3013386726379395,
+      "learning_rate": 1.2348318713450294e-05,
+      "loss": 1.3852,
+      "step": 8374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.893671751022339,
+      "learning_rate": 1.2347404970760235e-05,
+      "loss": 1.8097,
+      "step": 8375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3996384143829346,
+      "learning_rate": 1.2346491228070177e-05,
+      "loss": 1.9692,
+      "step": 8376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.613372325897217,
+      "learning_rate": 1.2345577485380117e-05,
+      "loss": 1.5389,
+      "step": 8377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9568471908569336,
+      "learning_rate": 1.234466374269006e-05,
+      "loss": 1.4446,
+      "step": 8378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3548007011413574,
+      "learning_rate": 1.234375e-05,
+      "loss": 1.5636,
+      "step": 8379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5467989444732666,
+      "learning_rate": 1.2342836257309942e-05,
+      "loss": 1.5726,
+      "step": 8380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.294222831726074,
+      "learning_rate": 1.2341922514619885e-05,
+      "loss": 1.4752,
+      "step": 8381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7440288066864014,
+      "learning_rate": 1.2341008771929825e-05,
+      "loss": 1.6545,
+      "step": 8382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.757168769836426,
+      "learning_rate": 1.2340095029239767e-05,
+      "loss": 1.6206,
+      "step": 8383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.731140613555908,
+      "learning_rate": 1.2339181286549708e-05,
+      "loss": 1.5506,
+      "step": 8384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6226255893707275,
+      "learning_rate": 1.233826754385965e-05,
+      "loss": 1.397,
+      "step": 8385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.328517198562622,
+      "learning_rate": 1.233735380116959e-05,
+      "loss": 1.3743,
+      "step": 8386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.520993947982788,
+      "learning_rate": 1.2336440058479533e-05,
+      "loss": 1.7304,
+      "step": 8387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.181886911392212,
+      "learning_rate": 1.2335526315789475e-05,
+      "loss": 1.8896,
+      "step": 8388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.055394172668457,
+      "learning_rate": 1.2334612573099416e-05,
+      "loss": 1.5252,
+      "step": 8389
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.301424980163574,
+      "learning_rate": 1.2333698830409358e-05,
+      "loss": 1.8562,
+      "step": 8390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7815990447998047,
+      "learning_rate": 1.2332785087719298e-05,
+      "loss": 1.672,
+      "step": 8391
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.110629081726074,
+      "learning_rate": 1.233187134502924e-05,
+      "loss": 1.7576,
+      "step": 8392
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.745377779006958,
+      "learning_rate": 1.2330957602339181e-05,
+      "loss": 1.6997,
+      "step": 8393
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.830061197280884,
+      "learning_rate": 1.2330043859649123e-05,
+      "loss": 1.5906,
+      "step": 8394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.935980796813965,
+      "learning_rate": 1.2329130116959066e-05,
+      "loss": 1.835,
+      "step": 8395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2575652599334717,
+      "learning_rate": 1.2328216374269006e-05,
+      "loss": 1.4271,
+      "step": 8396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.05349349975586,
+      "learning_rate": 1.2327302631578948e-05,
+      "loss": 1.6245,
+      "step": 8397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5661067962646484,
+      "learning_rate": 1.2326388888888889e-05,
+      "loss": 1.2239,
+      "step": 8398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4464709758758545,
+      "learning_rate": 1.2325475146198831e-05,
+      "loss": 1.4014,
+      "step": 8399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2520334720611572,
+      "learning_rate": 1.2324561403508772e-05,
+      "loss": 1.3652,
+      "step": 8400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.531224489212036,
+      "learning_rate": 1.2323647660818714e-05,
+      "loss": 1.3211,
+      "step": 8401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.09889817237854,
+      "learning_rate": 1.2322733918128656e-05,
+      "loss": 1.3568,
+      "step": 8402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.214186668395996,
+      "learning_rate": 1.2321820175438597e-05,
+      "loss": 1.6131,
+      "step": 8403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9283080101013184,
+      "learning_rate": 1.2320906432748539e-05,
+      "loss": 1.6234,
+      "step": 8404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7365598678588867,
+      "learning_rate": 1.231999269005848e-05,
+      "loss": 1.5769,
+      "step": 8405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7822697162628174,
+      "learning_rate": 1.2319078947368422e-05,
+      "loss": 1.539,
+      "step": 8406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3766889572143555,
+      "learning_rate": 1.2318165204678362e-05,
+      "loss": 1.5183,
+      "step": 8407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.549492359161377,
+      "learning_rate": 1.2317251461988304e-05,
+      "loss": 1.3386,
+      "step": 8408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0120651721954346,
+      "learning_rate": 1.2316337719298247e-05,
+      "loss": 1.3715,
+      "step": 8409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.741077184677124,
+      "learning_rate": 1.2315423976608187e-05,
+      "loss": 1.6673,
+      "step": 8410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4645845890045166,
+      "learning_rate": 1.231451023391813e-05,
+      "loss": 1.4931,
+      "step": 8411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5063812732696533,
+      "learning_rate": 1.231359649122807e-05,
+      "loss": 1.3665,
+      "step": 8412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.769509792327881,
+      "learning_rate": 1.2312682748538012e-05,
+      "loss": 1.7408,
+      "step": 8413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5520572662353516,
+      "learning_rate": 1.2311769005847953e-05,
+      "loss": 1.5407,
+      "step": 8414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5742766857147217,
+      "learning_rate": 1.2310855263157895e-05,
+      "loss": 1.4357,
+      "step": 8415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7630603313446045,
+      "learning_rate": 1.2309941520467837e-05,
+      "loss": 1.2359,
+      "step": 8416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5684335231781006,
+      "learning_rate": 1.2309027777777777e-05,
+      "loss": 1.5716,
+      "step": 8417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2607431411743164,
+      "learning_rate": 1.230811403508772e-05,
+      "loss": 1.4712,
+      "step": 8418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6076176166534424,
+      "learning_rate": 1.230720029239766e-05,
+      "loss": 1.4425,
+      "step": 8419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.516383647918701,
+      "learning_rate": 1.2306286549707602e-05,
+      "loss": 1.5471,
+      "step": 8420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5937304496765137,
+      "learning_rate": 1.2305372807017543e-05,
+      "loss": 1.4622,
+      "step": 8421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.830813407897949,
+      "learning_rate": 1.2304459064327487e-05,
+      "loss": 1.5551,
+      "step": 8422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.636054039001465,
+      "learning_rate": 1.2303545321637429e-05,
+      "loss": 1.4143,
+      "step": 8423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6406288146972656,
+      "learning_rate": 1.230263157894737e-05,
+      "loss": 1.4592,
+      "step": 8424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3539249897003174,
+      "learning_rate": 1.2301717836257312e-05,
+      "loss": 2.1581,
+      "step": 8425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9829564094543457,
+      "learning_rate": 1.2300804093567254e-05,
+      "loss": 1.8518,
+      "step": 8426
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9893622398376465,
+      "learning_rate": 1.2299890350877195e-05,
+      "loss": 1.409,
+      "step": 8427
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0374748706817627,
+      "learning_rate": 1.2298976608187137e-05,
+      "loss": 2.3124,
+      "step": 8428
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.981292963027954,
+      "learning_rate": 1.2298062865497077e-05,
+      "loss": 1.7006,
+      "step": 8429
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9795186519622803,
+      "learning_rate": 1.229714912280702e-05,
+      "loss": 1.4415,
+      "step": 8430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3471899032592773,
+      "learning_rate": 1.229623538011696e-05,
+      "loss": 1.4674,
+      "step": 8431
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.571550130844116,
+      "learning_rate": 1.2295321637426902e-05,
+      "loss": 1.5906,
+      "step": 8432
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1568727493286133,
+      "learning_rate": 1.2294407894736845e-05,
+      "loss": 1.1866,
+      "step": 8433
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.286224842071533,
+      "learning_rate": 1.2293494152046785e-05,
+      "loss": 2.3867,
+      "step": 8434
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0608415603637695,
+      "learning_rate": 1.2292580409356727e-05,
+      "loss": 1.183,
+      "step": 8435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.016479969024658,
+      "learning_rate": 1.2291666666666668e-05,
+      "loss": 1.431,
+      "step": 8436
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8947389125823975,
+      "learning_rate": 1.229075292397661e-05,
+      "loss": 1.4742,
+      "step": 8437
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0761656761169434,
+      "learning_rate": 1.228983918128655e-05,
+      "loss": 1.9922,
+      "step": 8438
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6026315689086914,
+      "learning_rate": 1.2288925438596493e-05,
+      "loss": 1.6823,
+      "step": 8439
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.571967363357544,
+      "learning_rate": 1.2288011695906435e-05,
+      "loss": 1.5095,
+      "step": 8440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.501300811767578,
+      "learning_rate": 1.2287097953216376e-05,
+      "loss": 1.3143,
+      "step": 8441
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2900185585021973,
+      "learning_rate": 1.2286184210526318e-05,
+      "loss": 1.9237,
+      "step": 8442
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.636836051940918,
+      "learning_rate": 1.2285270467836258e-05,
+      "loss": 1.4437,
+      "step": 8443
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.577727794647217,
+      "learning_rate": 1.22843567251462e-05,
+      "loss": 1.5003,
+      "step": 8444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6624648571014404,
+      "learning_rate": 1.2283442982456141e-05,
+      "loss": 1.5168,
+      "step": 8445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7508981227874756,
+      "learning_rate": 1.2282529239766083e-05,
+      "loss": 1.7447,
+      "step": 8446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7394702434539795,
+      "learning_rate": 1.2281615497076026e-05,
+      "loss": 1.5367,
+      "step": 8447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8474228382110596,
+      "learning_rate": 1.2280701754385966e-05,
+      "loss": 1.8175,
+      "step": 8448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.873133420944214,
+      "learning_rate": 1.2279788011695908e-05,
+      "loss": 1.7582,
+      "step": 8449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.633875846862793,
+      "learning_rate": 1.2278874269005849e-05,
+      "loss": 1.6112,
+      "step": 8450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.003312349319458,
+      "learning_rate": 1.2277960526315791e-05,
+      "loss": 1.7436,
+      "step": 8451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5153894424438477,
+      "learning_rate": 1.2277046783625732e-05,
+      "loss": 1.3972,
+      "step": 8452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.251438617706299,
+      "learning_rate": 1.2276133040935674e-05,
+      "loss": 1.9806,
+      "step": 8453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8104658126831055,
+      "learning_rate": 1.2275219298245616e-05,
+      "loss": 1.4392,
+      "step": 8454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8580069541931152,
+      "learning_rate": 1.2274305555555557e-05,
+      "loss": 1.7107,
+      "step": 8455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2944674491882324,
+      "learning_rate": 1.2273391812865499e-05,
+      "loss": 1.2566,
+      "step": 8456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8830113410949707,
+      "learning_rate": 1.227247807017544e-05,
+      "loss": 1.5348,
+      "step": 8457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.89168643951416,
+      "learning_rate": 1.2271564327485381e-05,
+      "loss": 1.7668,
+      "step": 8458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6202995777130127,
+      "learning_rate": 1.2270650584795322e-05,
+      "loss": 1.4744,
+      "step": 8459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2037997245788574,
+      "learning_rate": 1.2269736842105264e-05,
+      "loss": 1.937,
+      "step": 8460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5132265090942383,
+      "learning_rate": 1.2268823099415206e-05,
+      "loss": 1.3247,
+      "step": 8461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5019121170043945,
+      "learning_rate": 1.2267909356725147e-05,
+      "loss": 1.5292,
+      "step": 8462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.614964485168457,
+      "learning_rate": 1.226699561403509e-05,
+      "loss": 1.539,
+      "step": 8463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.529017210006714,
+      "learning_rate": 1.226608187134503e-05,
+      "loss": 1.5585,
+      "step": 8464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.325038194656372,
+      "learning_rate": 1.2265168128654972e-05,
+      "loss": 1.5383,
+      "step": 8465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3985533714294434,
+      "learning_rate": 1.2264254385964912e-05,
+      "loss": 1.7321,
+      "step": 8466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7014994621276855,
+      "learning_rate": 1.2263340643274855e-05,
+      "loss": 1.3058,
+      "step": 8467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2250022888183594,
+      "learning_rate": 1.2262426900584797e-05,
+      "loss": 1.5854,
+      "step": 8468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6131653785705566,
+      "learning_rate": 1.2261513157894737e-05,
+      "loss": 1.1202,
+      "step": 8469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4606943130493164,
+      "learning_rate": 1.226059941520468e-05,
+      "loss": 1.2285,
+      "step": 8470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8975822925567627,
+      "learning_rate": 1.225968567251462e-05,
+      "loss": 2.0036,
+      "step": 8471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9213998317718506,
+      "learning_rate": 1.2258771929824562e-05,
+      "loss": 2.0549,
+      "step": 8472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0299034118652344,
+      "learning_rate": 1.2257858187134503e-05,
+      "loss": 1.6919,
+      "step": 8473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.169450283050537,
+      "learning_rate": 1.2256944444444445e-05,
+      "loss": 1.6905,
+      "step": 8474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5942585468292236,
+      "learning_rate": 1.2256030701754387e-05,
+      "loss": 1.6962,
+      "step": 8475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3062939643859863,
+      "learning_rate": 1.2255116959064328e-05,
+      "loss": 1.3983,
+      "step": 8476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.218143939971924,
+      "learning_rate": 1.225420321637427e-05,
+      "loss": 1.4012,
+      "step": 8477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0655131340026855,
+      "learning_rate": 1.225328947368421e-05,
+      "loss": 1.7443,
+      "step": 8478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.099442720413208,
+      "learning_rate": 1.2252375730994153e-05,
+      "loss": 1.9722,
+      "step": 8479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3847572803497314,
+      "learning_rate": 1.2251461988304093e-05,
+      "loss": 1.749,
+      "step": 8480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5830674171447754,
+      "learning_rate": 1.2250548245614036e-05,
+      "loss": 1.8959,
+      "step": 8481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1414096355438232,
+      "learning_rate": 1.2249634502923978e-05,
+      "loss": 1.8922,
+      "step": 8482
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8350582122802734,
+      "learning_rate": 1.2248720760233918e-05,
+      "loss": 1.3979,
+      "step": 8483
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5750021934509277,
+      "learning_rate": 1.224780701754386e-05,
+      "loss": 1.5697,
+      "step": 8484
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2619237899780273,
+      "learning_rate": 1.2246893274853801e-05,
+      "loss": 1.0349,
+      "step": 8485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2402093410491943,
+      "learning_rate": 1.2245979532163743e-05,
+      "loss": 1.7081,
+      "step": 8486
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.512481212615967,
+      "learning_rate": 1.2245065789473684e-05,
+      "loss": 1.5483,
+      "step": 8487
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3550667762756348,
+      "learning_rate": 1.2244152046783626e-05,
+      "loss": 1.3732,
+      "step": 8488
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.420217514038086,
+      "learning_rate": 1.2243238304093568e-05,
+      "loss": 1.3853,
+      "step": 8489
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.485011339187622,
+      "learning_rate": 1.2242324561403509e-05,
+      "loss": 1.6924,
+      "step": 8490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9337270259857178,
+      "learning_rate": 1.2241410818713451e-05,
+      "loss": 1.6206,
+      "step": 8491
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6045515537261963,
+      "learning_rate": 1.2240497076023392e-05,
+      "loss": 1.3152,
+      "step": 8492
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1435647010803223,
+      "learning_rate": 1.2239583333333334e-05,
+      "loss": 1.2871,
+      "step": 8493
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.31169056892395,
+      "learning_rate": 1.2238669590643274e-05,
+      "loss": 1.1909,
+      "step": 8494
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6047475337982178,
+      "learning_rate": 1.2237755847953217e-05,
+      "loss": 1.6152,
+      "step": 8495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.051900863647461,
+      "learning_rate": 1.2236842105263159e-05,
+      "loss": 1.651,
+      "step": 8496
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5321743488311768,
+      "learning_rate": 1.22359283625731e-05,
+      "loss": 1.5116,
+      "step": 8497
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4633209705352783,
+      "learning_rate": 1.2235014619883042e-05,
+      "loss": 1.2739,
+      "step": 8498
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0913608074188232,
+      "learning_rate": 1.2234100877192982e-05,
+      "loss": 1.3603,
+      "step": 8499
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4543168544769287,
+      "learning_rate": 1.2233187134502924e-05,
+      "loss": 1.6217,
+      "step": 8500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5691022872924805,
+      "learning_rate": 1.2232273391812865e-05,
+      "loss": 1.4691,
+      "step": 8501
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4130539894104004,
+      "learning_rate": 1.2231359649122807e-05,
+      "loss": 1.3623,
+      "step": 8502
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4379420280456543,
+      "learning_rate": 1.223044590643275e-05,
+      "loss": 2.4388,
+      "step": 8503
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.081915855407715,
+      "learning_rate": 1.222953216374269e-05,
+      "loss": 1.1191,
+      "step": 8504
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.741068124771118,
+      "learning_rate": 1.2228618421052632e-05,
+      "loss": 1.5514,
+      "step": 8505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2179486751556396,
+      "learning_rate": 1.2227704678362573e-05,
+      "loss": 1.2846,
+      "step": 8506
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2624759674072266,
+      "learning_rate": 1.2226790935672515e-05,
+      "loss": 1.3582,
+      "step": 8507
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.942451000213623,
+      "learning_rate": 1.2225877192982459e-05,
+      "loss": 1.7004,
+      "step": 8508
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.661109685897827,
+      "learning_rate": 1.22249634502924e-05,
+      "loss": 1.331,
+      "step": 8509
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6745445728302,
+      "learning_rate": 1.2224049707602341e-05,
+      "loss": 1.4898,
+      "step": 8510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.978161096572876,
+      "learning_rate": 1.2223135964912282e-05,
+      "loss": 1.7101,
+      "step": 8511
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.868424892425537,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 1.6306,
+      "step": 8512
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9437276124954224,
+      "learning_rate": 1.2221308479532166e-05,
+      "loss": 1.3978,
+      "step": 8513
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6901090145111084,
+      "learning_rate": 1.2220394736842107e-05,
+      "loss": 1.4011,
+      "step": 8514
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.10054874420166,
+      "learning_rate": 1.2219480994152049e-05,
+      "loss": 1.1865,
+      "step": 8515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.836554527282715,
+      "learning_rate": 1.221856725146199e-05,
+      "loss": 1.9578,
+      "step": 8516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.284118175506592,
+      "learning_rate": 1.2217653508771932e-05,
+      "loss": 1.5826,
+      "step": 8517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.509032964706421,
+      "learning_rate": 1.2216739766081872e-05,
+      "loss": 1.7497,
+      "step": 8518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8026037216186523,
+      "learning_rate": 1.2215826023391815e-05,
+      "loss": 1.2461,
+      "step": 8519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6534717082977295,
+      "learning_rate": 1.2214912280701757e-05,
+      "loss": 1.6676,
+      "step": 8520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.67067813873291,
+      "learning_rate": 1.2213998538011697e-05,
+      "loss": 1.6701,
+      "step": 8521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6892476081848145,
+      "learning_rate": 1.221308479532164e-05,
+      "loss": 1.7003,
+      "step": 8522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6223392486572266,
+      "learning_rate": 1.221217105263158e-05,
+      "loss": 1.7717,
+      "step": 8523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6238203048706055,
+      "learning_rate": 1.2211257309941522e-05,
+      "loss": 1.6789,
+      "step": 8524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1358354091644287,
+      "learning_rate": 1.2210343567251463e-05,
+      "loss": 1.6531,
+      "step": 8525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5605764389038086,
+      "learning_rate": 1.2209429824561405e-05,
+      "loss": 1.5621,
+      "step": 8526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6712806224823,
+      "learning_rate": 1.2208516081871347e-05,
+      "loss": 1.9593,
+      "step": 8527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8501133918762207,
+      "learning_rate": 1.2207602339181288e-05,
+      "loss": 1.5531,
+      "step": 8528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2800638675689697,
+      "learning_rate": 1.220668859649123e-05,
+      "loss": 1.3435,
+      "step": 8529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8317863941192627,
+      "learning_rate": 1.220577485380117e-05,
+      "loss": 1.5803,
+      "step": 8530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.56097412109375,
+      "learning_rate": 1.2204861111111113e-05,
+      "loss": 1.2505,
+      "step": 8531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.554614543914795,
+      "learning_rate": 1.2203947368421053e-05,
+      "loss": 1.5323,
+      "step": 8532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0161683559417725,
+      "learning_rate": 1.2203033625730996e-05,
+      "loss": 1.6697,
+      "step": 8533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4693405628204346,
+      "learning_rate": 1.2202119883040938e-05,
+      "loss": 1.6078,
+      "step": 8534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.067645788192749,
+      "learning_rate": 1.2201206140350878e-05,
+      "loss": 1.6133,
+      "step": 8535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.693899393081665,
+      "learning_rate": 1.220029239766082e-05,
+      "loss": 1.5438,
+      "step": 8536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8208746910095215,
+      "learning_rate": 1.2199378654970761e-05,
+      "loss": 1.6851,
+      "step": 8537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0047435760498047,
+      "learning_rate": 1.2198464912280703e-05,
+      "loss": 1.6797,
+      "step": 8538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.897794723510742,
+      "learning_rate": 1.2197551169590644e-05,
+      "loss": 1.4249,
+      "step": 8539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5919573307037354,
+      "learning_rate": 1.2196637426900586e-05,
+      "loss": 1.2738,
+      "step": 8540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.426652431488037,
+      "learning_rate": 1.2195723684210528e-05,
+      "loss": 1.4308,
+      "step": 8541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3607053756713867,
+      "learning_rate": 1.2194809941520469e-05,
+      "loss": 1.1847,
+      "step": 8542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3138744831085205,
+      "learning_rate": 1.2193896198830411e-05,
+      "loss": 1.5758,
+      "step": 8543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.595707416534424,
+      "learning_rate": 1.2192982456140352e-05,
+      "loss": 1.6373,
+      "step": 8544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3957748413085938,
+      "learning_rate": 1.2192068713450294e-05,
+      "loss": 1.6011,
+      "step": 8545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1478850841522217,
+      "learning_rate": 1.2191154970760234e-05,
+      "loss": 1.9699,
+      "step": 8546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.651884078979492,
+      "learning_rate": 1.2190241228070177e-05,
+      "loss": 1.5583,
+      "step": 8547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.269721031188965,
+      "learning_rate": 1.2189327485380119e-05,
+      "loss": 1.4312,
+      "step": 8548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0906522274017334,
+      "learning_rate": 1.218841374269006e-05,
+      "loss": 1.5458,
+      "step": 8549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.524120807647705,
+      "learning_rate": 1.2187500000000001e-05,
+      "loss": 1.5822,
+      "step": 8550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.470803737640381,
+      "learning_rate": 1.2186586257309942e-05,
+      "loss": 1.4989,
+      "step": 8551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.391885280609131,
+      "learning_rate": 1.2185672514619884e-05,
+      "loss": 1.293,
+      "step": 8552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.600872039794922,
+      "learning_rate": 1.2184758771929825e-05,
+      "loss": 1.5478,
+      "step": 8553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.577505111694336,
+      "learning_rate": 1.2183845029239767e-05,
+      "loss": 1.5502,
+      "step": 8554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2268383502960205,
+      "learning_rate": 1.218293128654971e-05,
+      "loss": 1.0513,
+      "step": 8555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.566326379776001,
+      "learning_rate": 1.218201754385965e-05,
+      "loss": 1.7501,
+      "step": 8556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5932161808013916,
+      "learning_rate": 1.2181103801169592e-05,
+      "loss": 1.6278,
+      "step": 8557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.645197868347168,
+      "learning_rate": 1.2180190058479532e-05,
+      "loss": 1.5327,
+      "step": 8558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.645798683166504,
+      "learning_rate": 1.2179276315789475e-05,
+      "loss": 1.7871,
+      "step": 8559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7593460083007812,
+      "learning_rate": 1.2178362573099415e-05,
+      "loss": 1.4891,
+      "step": 8560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.518378734588623,
+      "learning_rate": 1.2177448830409357e-05,
+      "loss": 1.6121,
+      "step": 8561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.350693464279175,
+      "learning_rate": 1.21765350877193e-05,
+      "loss": 1.4409,
+      "step": 8562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.415611743927002,
+      "learning_rate": 1.217562134502924e-05,
+      "loss": 1.9125,
+      "step": 8563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2745211124420166,
+      "learning_rate": 1.2174707602339182e-05,
+      "loss": 1.6355,
+      "step": 8564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.693751335144043,
+      "learning_rate": 1.2173793859649123e-05,
+      "loss": 1.3125,
+      "step": 8565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5001614093780518,
+      "learning_rate": 1.2172880116959065e-05,
+      "loss": 1.5524,
+      "step": 8566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.121561288833618,
+      "learning_rate": 1.2171966374269006e-05,
+      "loss": 1.4614,
+      "step": 8567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.734724521636963,
+      "learning_rate": 1.2171052631578948e-05,
+      "loss": 1.523,
+      "step": 8568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.676888942718506,
+      "learning_rate": 1.217013888888889e-05,
+      "loss": 1.5839,
+      "step": 8569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.858177423477173,
+      "learning_rate": 1.216922514619883e-05,
+      "loss": 1.5488,
+      "step": 8570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4699158668518066,
+      "learning_rate": 1.2168311403508773e-05,
+      "loss": 1.439,
+      "step": 8571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2950167655944824,
+      "learning_rate": 1.2167397660818713e-05,
+      "loss": 1.2905,
+      "step": 8572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.641953468322754,
+      "learning_rate": 1.2166483918128656e-05,
+      "loss": 1.6529,
+      "step": 8573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2657086849212646,
+      "learning_rate": 1.2165570175438596e-05,
+      "loss": 1.0429,
+      "step": 8574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0492615699768066,
+      "learning_rate": 1.2164656432748538e-05,
+      "loss": 1.8579,
+      "step": 8575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9011106491088867,
+      "learning_rate": 1.216374269005848e-05,
+      "loss": 1.4073,
+      "step": 8576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.648841381072998,
+      "learning_rate": 1.2162828947368421e-05,
+      "loss": 1.4683,
+      "step": 8577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.325218439102173,
+      "learning_rate": 1.2161915204678363e-05,
+      "loss": 1.4036,
+      "step": 8578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1093835830688477,
+      "learning_rate": 1.2161001461988304e-05,
+      "loss": 1.6855,
+      "step": 8579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5683016777038574,
+      "learning_rate": 1.2160087719298246e-05,
+      "loss": 1.4357,
+      "step": 8580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.263777256011963,
+      "learning_rate": 1.2159173976608187e-05,
+      "loss": 1.2114,
+      "step": 8581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.166360855102539,
+      "learning_rate": 1.2158260233918129e-05,
+      "loss": 1.6516,
+      "step": 8582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.241809844970703,
+      "learning_rate": 1.2157346491228071e-05,
+      "loss": 1.7138,
+      "step": 8583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8126187324523926,
+      "learning_rate": 1.2156432748538012e-05,
+      "loss": 1.435,
+      "step": 8584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.256507396697998,
+      "learning_rate": 1.2155519005847954e-05,
+      "loss": 1.3977,
+      "step": 8585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.513345956802368,
+      "learning_rate": 1.2154605263157894e-05,
+      "loss": 1.6363,
+      "step": 8586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.483220338821411,
+      "learning_rate": 1.2153691520467837e-05,
+      "loss": 1.2876,
+      "step": 8587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.284369707107544,
+      "learning_rate": 1.2152777777777777e-05,
+      "loss": 1.4306,
+      "step": 8588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5808303356170654,
+      "learning_rate": 1.215186403508772e-05,
+      "loss": 1.6705,
+      "step": 8589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0686659812927246,
+      "learning_rate": 1.2150950292397662e-05,
+      "loss": 1.3604,
+      "step": 8590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.699495553970337,
+      "learning_rate": 1.2150036549707602e-05,
+      "loss": 1.4919,
+      "step": 8591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6527698040008545,
+      "learning_rate": 1.2149122807017544e-05,
+      "loss": 1.5479,
+      "step": 8592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.238001823425293,
+      "learning_rate": 1.2148209064327488e-05,
+      "loss": 1.2705,
+      "step": 8593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5782647132873535,
+      "learning_rate": 1.2147295321637429e-05,
+      "loss": 1.5505,
+      "step": 8594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0559747219085693,
+      "learning_rate": 1.2146381578947371e-05,
+      "loss": 1.3592,
+      "step": 8595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9227068424224854,
+      "learning_rate": 1.2145467836257311e-05,
+      "loss": 2.0258,
+      "step": 8596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.366541624069214,
+      "learning_rate": 1.2144554093567254e-05,
+      "loss": 2.0322,
+      "step": 8597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.480907917022705,
+      "learning_rate": 1.2143640350877194e-05,
+      "loss": 1.49,
+      "step": 8598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.727473735809326,
+      "learning_rate": 1.2142726608187136e-05,
+      "loss": 1.5705,
+      "step": 8599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7654335498809814,
+      "learning_rate": 1.2141812865497079e-05,
+      "loss": 1.6213,
+      "step": 8600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2686009407043457,
+      "learning_rate": 1.214089912280702e-05,
+      "loss": 1.6616,
+      "step": 8601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.569378137588501,
+      "learning_rate": 1.2139985380116961e-05,
+      "loss": 1.9126,
+      "step": 8602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5603439807891846,
+      "learning_rate": 1.2139071637426902e-05,
+      "loss": 1.3958,
+      "step": 8603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6197614669799805,
+      "learning_rate": 1.2138157894736844e-05,
+      "loss": 1.6948,
+      "step": 8604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.597019910812378,
+      "learning_rate": 1.2137244152046785e-05,
+      "loss": 1.4305,
+      "step": 8605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5019965171813965,
+      "learning_rate": 1.2136330409356727e-05,
+      "loss": 1.3299,
+      "step": 8606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.511418342590332,
+      "learning_rate": 1.2135416666666669e-05,
+      "loss": 1.3859,
+      "step": 8607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4689085483551025,
+      "learning_rate": 1.213450292397661e-05,
+      "loss": 1.4751,
+      "step": 8608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7158362865448,
+      "learning_rate": 1.2133589181286552e-05,
+      "loss": 1.6105,
+      "step": 8609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4429030418395996,
+      "learning_rate": 1.2132675438596492e-05,
+      "loss": 1.3371,
+      "step": 8610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.270305871963501,
+      "learning_rate": 1.2131761695906435e-05,
+      "loss": 1.3494,
+      "step": 8611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.634014368057251,
+      "learning_rate": 1.2130847953216375e-05,
+      "loss": 1.4101,
+      "step": 8612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2657597064971924,
+      "learning_rate": 1.2129934210526317e-05,
+      "loss": 1.3062,
+      "step": 8613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.248708724975586,
+      "learning_rate": 1.212902046783626e-05,
+      "loss": 1.579,
+      "step": 8614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.943002700805664,
+      "learning_rate": 1.21281067251462e-05,
+      "loss": 1.6239,
+      "step": 8615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1712515354156494,
+      "learning_rate": 1.2127192982456142e-05,
+      "loss": 1.8247,
+      "step": 8616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5009329319000244,
+      "learning_rate": 1.2126279239766083e-05,
+      "loss": 1.6378,
+      "step": 8617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3590242862701416,
+      "learning_rate": 1.2125365497076025e-05,
+      "loss": 1.3983,
+      "step": 8618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.935218572616577,
+      "learning_rate": 1.2124451754385966e-05,
+      "loss": 1.4368,
+      "step": 8619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7591726779937744,
+      "learning_rate": 1.2123538011695908e-05,
+      "loss": 1.496,
+      "step": 8620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9661757946014404,
+      "learning_rate": 1.212262426900585e-05,
+      "loss": 1.8364,
+      "step": 8621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.40572452545166,
+      "learning_rate": 1.212171052631579e-05,
+      "loss": 1.2441,
+      "step": 8622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5860705375671387,
+      "learning_rate": 1.2120796783625733e-05,
+      "loss": 1.4957,
+      "step": 8623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.81170654296875,
+      "learning_rate": 1.2119883040935673e-05,
+      "loss": 1.3823,
+      "step": 8624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1021926403045654,
+      "learning_rate": 1.2118969298245616e-05,
+      "loss": 1.2645,
+      "step": 8625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.900629997253418,
+      "learning_rate": 1.2118055555555556e-05,
+      "loss": 2.0484,
+      "step": 8626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.01824688911438,
+      "learning_rate": 1.2117141812865498e-05,
+      "loss": 1.7966,
+      "step": 8627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2457778453826904,
+      "learning_rate": 1.211622807017544e-05,
+      "loss": 1.1562,
+      "step": 8628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7361249923706055,
+      "learning_rate": 1.2115314327485381e-05,
+      "loss": 1.7242,
+      "step": 8629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5595502853393555,
+      "learning_rate": 1.2114400584795323e-05,
+      "loss": 1.5653,
+      "step": 8630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3655073642730713,
+      "learning_rate": 1.2113486842105264e-05,
+      "loss": 1.6092,
+      "step": 8631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5389339923858643,
+      "learning_rate": 1.2112573099415206e-05,
+      "loss": 1.8587,
+      "step": 8632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.205327033996582,
+      "learning_rate": 1.2111659356725147e-05,
+      "loss": 1.3344,
+      "step": 8633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.939478874206543,
+      "learning_rate": 1.2110745614035089e-05,
+      "loss": 1.6217,
+      "step": 8634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.553743600845337,
+      "learning_rate": 1.2109831871345031e-05,
+      "loss": 1.4613,
+      "step": 8635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0353519916534424,
+      "learning_rate": 1.2108918128654972e-05,
+      "loss": 1.7284,
+      "step": 8636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.915364980697632,
+      "learning_rate": 1.2108004385964914e-05,
+      "loss": 1.6817,
+      "step": 8637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.282564401626587,
+      "learning_rate": 1.2107090643274854e-05,
+      "loss": 1.4361,
+      "step": 8638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6499850749969482,
+      "learning_rate": 1.2106176900584797e-05,
+      "loss": 1.4726,
+      "step": 8639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5668556690216064,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 1.4952,
+      "step": 8640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3308215141296387,
+      "learning_rate": 1.210434941520468e-05,
+      "loss": 1.3245,
+      "step": 8641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5471651554107666,
+      "learning_rate": 1.2103435672514621e-05,
+      "loss": 1.5519,
+      "step": 8642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.214097261428833,
+      "learning_rate": 1.2102521929824562e-05,
+      "loss": 1.8774,
+      "step": 8643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.226969003677368,
+      "learning_rate": 1.2101608187134504e-05,
+      "loss": 1.2656,
+      "step": 8644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.451167106628418,
+      "learning_rate": 1.2100694444444445e-05,
+      "loss": 1.3555,
+      "step": 8645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2650692462921143,
+      "learning_rate": 1.2099780701754387e-05,
+      "loss": 1.5223,
+      "step": 8646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7589988708496094,
+      "learning_rate": 1.2098866959064327e-05,
+      "loss": 1.5768,
+      "step": 8647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2955331802368164,
+      "learning_rate": 1.209795321637427e-05,
+      "loss": 1.5638,
+      "step": 8648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.413053512573242,
+      "learning_rate": 1.2097039473684212e-05,
+      "loss": 1.5021,
+      "step": 8649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5804481506347656,
+      "learning_rate": 1.2096125730994152e-05,
+      "loss": 1.5723,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6290695667266846,
+      "learning_rate": 1.2095211988304095e-05,
+      "loss": 1.6883,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.939422369003296,
+      "learning_rate": 1.2094298245614035e-05,
+      "loss": 1.4383,
+      "step": 8652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.73048996925354,
+      "learning_rate": 1.2093384502923977e-05,
+      "loss": 1.579,
+      "step": 8653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.508409261703491,
+      "learning_rate": 1.2092470760233918e-05,
+      "loss": 1.3684,
+      "step": 8654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.837054967880249,
+      "learning_rate": 1.209155701754386e-05,
+      "loss": 1.5157,
+      "step": 8655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0009498596191406,
+      "learning_rate": 1.2090643274853802e-05,
+      "loss": 1.4067,
+      "step": 8656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9279966354370117,
+      "learning_rate": 1.2089729532163743e-05,
+      "loss": 1.8284,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1559572219848633,
+      "learning_rate": 1.2088815789473685e-05,
+      "loss": 1.2766,
+      "step": 8658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.787538766860962,
+      "learning_rate": 1.2087902046783626e-05,
+      "loss": 1.5639,
+      "step": 8659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5187489986419678,
+      "learning_rate": 1.2086988304093568e-05,
+      "loss": 1.5215,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0003433227539062,
+      "learning_rate": 1.2086074561403508e-05,
+      "loss": 1.6381,
+      "step": 8661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.885406255722046,
+      "learning_rate": 1.208516081871345e-05,
+      "loss": 1.7382,
+      "step": 8662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.689568042755127,
+      "learning_rate": 1.2084247076023393e-05,
+      "loss": 1.2957,
+      "step": 8663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.649449110031128,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 1.5351,
+      "step": 8664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9952000379562378,
+      "learning_rate": 1.2082419590643276e-05,
+      "loss": 1.3645,
+      "step": 8665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.22786283493042,
+      "learning_rate": 1.2081505847953216e-05,
+      "loss": 1.3584,
+      "step": 8666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4250597953796387,
+      "learning_rate": 1.2080592105263158e-05,
+      "loss": 1.4785,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0601086616516113,
+      "learning_rate": 1.2079678362573099e-05,
+      "loss": 1.7241,
+      "step": 8668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8673346042633057,
+      "learning_rate": 1.2078764619883041e-05,
+      "loss": 2.062,
+      "step": 8669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7714715003967285,
+      "learning_rate": 1.2077850877192983e-05,
+      "loss": 1.3871,
+      "step": 8670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.31801438331604,
+      "learning_rate": 1.2076937134502924e-05,
+      "loss": 1.2519,
+      "step": 8671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.330380916595459,
+      "learning_rate": 1.2076023391812866e-05,
+      "loss": 1.7387,
+      "step": 8672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.172409772872925,
+      "learning_rate": 1.2075109649122807e-05,
+      "loss": 1.1861,
+      "step": 8673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.727203845977783,
+      "learning_rate": 1.2074195906432749e-05,
+      "loss": 1.7295,
+      "step": 8674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0203921794891357,
+      "learning_rate": 1.207328216374269e-05,
+      "loss": 1.8677,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.066279411315918,
+      "learning_rate": 1.2072368421052632e-05,
+      "loss": 1.347,
+      "step": 8676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.86340594291687,
+      "learning_rate": 1.2071454678362574e-05,
+      "loss": 1.6787,
+      "step": 8677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.806403160095215,
+      "learning_rate": 1.2070540935672514e-05,
+      "loss": 1.681,
+      "step": 8678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2522056102752686,
+      "learning_rate": 1.2069627192982458e-05,
+      "loss": 1.2177,
+      "step": 8679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3914620876312256,
+      "learning_rate": 1.20687134502924e-05,
+      "loss": 1.4844,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.400886297225952,
+      "learning_rate": 1.2067799707602341e-05,
+      "loss": 1.2459,
+      "step": 8681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.960057258605957,
+      "learning_rate": 1.2066885964912283e-05,
+      "loss": 1.5115,
+      "step": 8682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6526105403900146,
+      "learning_rate": 1.2065972222222224e-05,
+      "loss": 1.4149,
+      "step": 8683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3127195835113525,
+      "learning_rate": 1.2065058479532166e-05,
+      "loss": 2.0852,
+      "step": 8684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7859690189361572,
+      "learning_rate": 1.2064144736842107e-05,
+      "loss": 1.4369,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.877995491027832,
+      "learning_rate": 1.2063230994152049e-05,
+      "loss": 1.3304,
+      "step": 8686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4223029613494873,
+      "learning_rate": 1.2062317251461991e-05,
+      "loss": 1.378,
+      "step": 8687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6745340824127197,
+      "learning_rate": 1.2061403508771931e-05,
+      "loss": 1.7187,
+      "step": 8688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3286406993865967,
+      "learning_rate": 1.2060489766081874e-05,
+      "loss": 2.0474,
+      "step": 8689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.618614673614502,
+      "learning_rate": 1.2059576023391814e-05,
+      "loss": 1.3952,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.61555814743042,
+      "learning_rate": 1.2058662280701756e-05,
+      "loss": 1.4578,
+      "step": 8691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.627248525619507,
+      "learning_rate": 1.2057748538011697e-05,
+      "loss": 1.73,
+      "step": 8692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3292253017425537,
+      "learning_rate": 1.205683479532164e-05,
+      "loss": 1.0894,
+      "step": 8693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.598137855529785,
+      "learning_rate": 1.2055921052631581e-05,
+      "loss": 1.4414,
+      "step": 8694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4588463306427,
+      "learning_rate": 1.2055007309941522e-05,
+      "loss": 1.6179,
+      "step": 8695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.646682024002075,
+      "learning_rate": 1.2054093567251464e-05,
+      "loss": 1.2382,
+      "step": 8696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.26542592048645,
+      "learning_rate": 1.2053179824561405e-05,
+      "loss": 1.2108,
+      "step": 8697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2530362606048584,
+      "learning_rate": 1.2052266081871347e-05,
+      "loss": 1.265,
+      "step": 8698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5486679077148438,
+      "learning_rate": 1.2051352339181287e-05,
+      "loss": 1.6909,
+      "step": 8699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3811895847320557,
+      "learning_rate": 1.205043859649123e-05,
+      "loss": 1.4618,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5433976650238037,
+      "learning_rate": 1.2049524853801172e-05,
+      "loss": 1.3855,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.802234649658203,
+      "learning_rate": 1.2048611111111112e-05,
+      "loss": 1.6658,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0581657886505127,
+      "learning_rate": 1.2047697368421055e-05,
+      "loss": 1.6201,
+      "step": 8703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.238952398300171,
+      "learning_rate": 1.2046783625730995e-05,
+      "loss": 1.8472,
+      "step": 8704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.418368101119995,
+      "learning_rate": 1.2045869883040937e-05,
+      "loss": 1.2359,
+      "step": 8705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3697359561920166,
+      "learning_rate": 1.2044956140350878e-05,
+      "loss": 1.2806,
+      "step": 8706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7979252338409424,
+      "learning_rate": 1.204404239766082e-05,
+      "loss": 1.8322,
+      "step": 8707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.891103744506836,
+      "learning_rate": 1.2043128654970762e-05,
+      "loss": 1.3638,
+      "step": 8708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4565491676330566,
+      "learning_rate": 1.2042214912280703e-05,
+      "loss": 1.4333,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6495742797851562,
+      "learning_rate": 1.2041301169590645e-05,
+      "loss": 1.5256,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7367804050445557,
+      "learning_rate": 1.2040387426900586e-05,
+      "loss": 1.4348,
+      "step": 8711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4651308059692383,
+      "learning_rate": 1.2039473684210528e-05,
+      "loss": 1.51,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5821444988250732,
+      "learning_rate": 1.2038559941520468e-05,
+      "loss": 1.3668,
+      "step": 8713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5104382038116455,
+      "learning_rate": 1.203764619883041e-05,
+      "loss": 1.4369,
+      "step": 8714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4037721157073975,
+      "learning_rate": 1.2036732456140353e-05,
+      "loss": 1.4966,
+      "step": 8715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6705923080444336,
+      "learning_rate": 1.2035818713450293e-05,
+      "loss": 1.7954,
+      "step": 8716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5392212867736816,
+      "learning_rate": 1.2034904970760236e-05,
+      "loss": 1.6639,
+      "step": 8717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4367144107818604,
+      "learning_rate": 1.2033991228070176e-05,
+      "loss": 1.198,
+      "step": 8718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.607896089553833,
+      "learning_rate": 1.2033077485380118e-05,
+      "loss": 1.5675,
+      "step": 8719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2490592002868652,
+      "learning_rate": 1.2032163742690059e-05,
+      "loss": 1.9967,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4795782566070557,
+      "learning_rate": 1.2031250000000001e-05,
+      "loss": 1.4755,
+      "step": 8721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.204775094985962,
+      "learning_rate": 1.2030336257309943e-05,
+      "loss": 1.9067,
+      "step": 8722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8437798023223877,
+      "learning_rate": 1.2029422514619884e-05,
+      "loss": 1.7044,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9596657752990723,
+      "learning_rate": 1.2028508771929826e-05,
+      "loss": 1.7354,
+      "step": 8724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.181011915206909,
+      "learning_rate": 1.2027595029239767e-05,
+      "loss": 1.5208,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6312618255615234,
+      "learning_rate": 1.2026681286549709e-05,
+      "loss": 1.9032,
+      "step": 8726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 69.30879211425781,
+      "learning_rate": 1.202576754385965e-05,
+      "loss": 1.5332,
+      "step": 8727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.093719244003296,
+      "learning_rate": 1.2024853801169592e-05,
+      "loss": 1.8862,
+      "step": 8728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6254141330718994,
+      "learning_rate": 1.2023940058479534e-05,
+      "loss": 1.5483,
+      "step": 8729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.753833293914795,
+      "learning_rate": 1.2023026315789474e-05,
+      "loss": 1.4894,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4564332962036133,
+      "learning_rate": 1.2022112573099417e-05,
+      "loss": 1.6644,
+      "step": 8731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.946953296661377,
+      "learning_rate": 1.2021198830409357e-05,
+      "loss": 1.7997,
+      "step": 8732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5237486362457275,
+      "learning_rate": 1.20202850877193e-05,
+      "loss": 1.5596,
+      "step": 8733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7096407413482666,
+      "learning_rate": 1.201937134502924e-05,
+      "loss": 1.7028,
+      "step": 8734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.43110728263855,
+      "learning_rate": 1.2018457602339182e-05,
+      "loss": 1.2166,
+      "step": 8735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2576494216918945,
+      "learning_rate": 1.2017543859649124e-05,
+      "loss": 1.699,
+      "step": 8736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1773927211761475,
+      "learning_rate": 1.2016630116959065e-05,
+      "loss": 1.0234,
+      "step": 8737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6714048385620117,
+      "learning_rate": 1.2015716374269007e-05,
+      "loss": 1.449,
+      "step": 8738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4845008850097656,
+      "learning_rate": 1.2014802631578947e-05,
+      "loss": 1.4861,
+      "step": 8739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.401824474334717,
+      "learning_rate": 1.201388888888889e-05,
+      "loss": 1.3891,
+      "step": 8740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.755303382873535,
+      "learning_rate": 1.201297514619883e-05,
+      "loss": 1.4477,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.697554588317871,
+      "learning_rate": 1.2012061403508772e-05,
+      "loss": 1.2656,
+      "step": 8742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4860615730285645,
+      "learning_rate": 1.2011147660818715e-05,
+      "loss": 1.5062,
+      "step": 8743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.536647081375122,
+      "learning_rate": 1.2010233918128655e-05,
+      "loss": 1.4488,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0941381454467773,
+      "learning_rate": 1.2009320175438597e-05,
+      "loss": 1.2514,
+      "step": 8745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.330092430114746,
+      "learning_rate": 1.2008406432748538e-05,
+      "loss": 1.3376,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2677133083343506,
+      "learning_rate": 1.200749269005848e-05,
+      "loss": 2.1125,
+      "step": 8747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.384089469909668,
+      "learning_rate": 1.200657894736842e-05,
+      "loss": 1.5115,
+      "step": 8748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9036364555358887,
+      "learning_rate": 1.2005665204678363e-05,
+      "loss": 1.9797,
+      "step": 8749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4903862476348877,
+      "learning_rate": 1.2004751461988305e-05,
+      "loss": 1.5181,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.77388858795166,
+      "learning_rate": 1.2003837719298246e-05,
+      "loss": 1.7883,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.610985279083252,
+      "learning_rate": 1.2002923976608188e-05,
+      "loss": 1.6717,
+      "step": 8752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7282297611236572,
+      "learning_rate": 1.2002010233918128e-05,
+      "loss": 1.5309,
+      "step": 8753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.235891580581665,
+      "learning_rate": 1.200109649122807e-05,
+      "loss": 1.2485,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7209014892578125,
+      "learning_rate": 1.2000182748538011e-05,
+      "loss": 1.6779,
+      "step": 8755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.370464324951172,
+      "learning_rate": 1.1999269005847953e-05,
+      "loss": 1.2512,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4,
+      "eval_train_bleu": 0.22991293446563624,
+      "eval_train_loss": 2.129511833190918,
+      "eval_train_runtime": 24.3493,
+      "eval_train_samples_per_second": 41.11,
+      "eval_train_steps_per_second": 1.314,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1066997051239014,
+      "learning_rate": 1.1998355263157896e-05,
+      "loss": 1.2261,
+      "step": 8757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.605380058288574,
+      "learning_rate": 1.1997441520467836e-05,
+      "loss": 1.5775,
+      "step": 8758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5149343013763428,
+      "learning_rate": 1.1996527777777778e-05,
+      "loss": 1.5972,
+      "step": 8759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7251555919647217,
+      "learning_rate": 1.1995614035087719e-05,
+      "loss": 1.8292,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.574389934539795,
+      "learning_rate": 1.1994700292397661e-05,
+      "loss": 1.4002,
+      "step": 8761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.624279260635376,
+      "learning_rate": 1.1993786549707602e-05,
+      "loss": 1.6027,
+      "step": 8762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5899431705474854,
+      "learning_rate": 1.1992872807017544e-05,
+      "loss": 1.8127,
+      "step": 8763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.004472017288208,
+      "learning_rate": 1.1991959064327488e-05,
+      "loss": 1.7838,
+      "step": 8764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.139751672744751,
+      "learning_rate": 1.1991045321637428e-05,
+      "loss": 1.6613,
+      "step": 8765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.594940185546875,
+      "learning_rate": 1.199013157894737e-05,
+      "loss": 1.3677,
+      "step": 8766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.525623321533203,
+      "learning_rate": 1.1989217836257313e-05,
+      "loss": 1.622,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.035839080810547,
+      "learning_rate": 1.1988304093567253e-05,
+      "loss": 1.4843,
+      "step": 8768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.597825050354004,
+      "learning_rate": 1.1987390350877196e-05,
+      "loss": 1.5768,
+      "step": 8769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1947500705718994,
+      "learning_rate": 1.1986476608187136e-05,
+      "loss": 1.8092,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5929250717163086,
+      "learning_rate": 1.1985562865497078e-05,
+      "loss": 1.5078,
+      "step": 8771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.574040174484253,
+      "learning_rate": 1.1984649122807019e-05,
+      "loss": 1.6802,
+      "step": 8772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.339282989501953,
+      "learning_rate": 1.1983735380116961e-05,
+      "loss": 1.5166,
+      "step": 8773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9718990325927734,
+      "learning_rate": 1.1982821637426903e-05,
+      "loss": 1.3013,
+      "step": 8774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4578747749328613,
+      "learning_rate": 1.1981907894736844e-05,
+      "loss": 1.859,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.561631679534912,
+      "learning_rate": 1.1980994152046786e-05,
+      "loss": 1.3266,
+      "step": 8776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.269880771636963,
+      "learning_rate": 1.1980080409356727e-05,
+      "loss": 1.7219,
+      "step": 8777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.956381320953369,
+      "learning_rate": 1.1979166666666669e-05,
+      "loss": 1.2257,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6090779304504395,
+      "learning_rate": 1.197825292397661e-05,
+      "loss": 1.6083,
+      "step": 8779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2892355918884277,
+      "learning_rate": 1.1977339181286551e-05,
+      "loss": 1.3179,
+      "step": 8780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5862464904785156,
+      "learning_rate": 1.1976425438596494e-05,
+      "loss": 1.1894,
+      "step": 8781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7736477851867676,
+      "learning_rate": 1.1975511695906434e-05,
+      "loss": 1.3463,
+      "step": 8782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.395477533340454,
+      "learning_rate": 1.1974597953216376e-05,
+      "loss": 1.4899,
+      "step": 8783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9338386058807373,
+      "learning_rate": 1.1973684210526317e-05,
+      "loss": 1.7974,
+      "step": 8784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5927209854125977,
+      "learning_rate": 1.197277046783626e-05,
+      "loss": 1.3892,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9174835681915283,
+      "learning_rate": 1.19718567251462e-05,
+      "loss": 1.478,
+      "step": 8786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5669565200805664,
+      "learning_rate": 1.1970942982456142e-05,
+      "loss": 1.6628,
+      "step": 8787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.897641897201538,
+      "learning_rate": 1.1970029239766084e-05,
+      "loss": 1.7154,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3207106590270996,
+      "learning_rate": 1.1969115497076025e-05,
+      "loss": 1.5624,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4437479972839355,
+      "learning_rate": 1.1968201754385967e-05,
+      "loss": 1.437,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1708643436431885,
+      "learning_rate": 1.1967288011695907e-05,
+      "loss": 1.806,
+      "step": 8791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.103610992431641,
+      "learning_rate": 1.196637426900585e-05,
+      "loss": 1.7565,
+      "step": 8792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2474405765533447,
+      "learning_rate": 1.196546052631579e-05,
+      "loss": 1.1959,
+      "step": 8793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5144922733306885,
+      "learning_rate": 1.1964546783625732e-05,
+      "loss": 1.4719,
+      "step": 8794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.796057939529419,
+      "learning_rate": 1.1963633040935675e-05,
+      "loss": 1.4784,
+      "step": 8795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.659515380859375,
+      "learning_rate": 1.1962719298245615e-05,
+      "loss": 1.7054,
+      "step": 8796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0021250247955322,
+      "learning_rate": 1.1961805555555557e-05,
+      "loss": 1.0656,
+      "step": 8797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2421317100524902,
+      "learning_rate": 1.1960891812865498e-05,
+      "loss": 1.6298,
+      "step": 8798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1008918285369873,
+      "learning_rate": 1.195997807017544e-05,
+      "loss": 1.2215,
+      "step": 8799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.774021625518799,
+      "learning_rate": 1.195906432748538e-05,
+      "loss": 1.7035,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0909717082977295,
+      "learning_rate": 1.1958150584795323e-05,
+      "loss": 1.8884,
+      "step": 8801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7024481296539307,
+      "learning_rate": 1.1957236842105265e-05,
+      "loss": 1.535,
+      "step": 8802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9342052936553955,
+      "learning_rate": 1.1956323099415206e-05,
+      "loss": 1.7737,
+      "step": 8803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.049664258956909,
+      "learning_rate": 1.1955409356725148e-05,
+      "loss": 1.7011,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.032135248184204,
+      "learning_rate": 1.1954495614035088e-05,
+      "loss": 1.734,
+      "step": 8805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7989790439605713,
+      "learning_rate": 1.195358187134503e-05,
+      "loss": 1.6995,
+      "step": 8806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.039259672164917,
+      "learning_rate": 1.1952668128654971e-05,
+      "loss": 1.4089,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8699114322662354,
+      "learning_rate": 1.1951754385964913e-05,
+      "loss": 1.9092,
+      "step": 8808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.402068853378296,
+      "learning_rate": 1.1950840643274854e-05,
+      "loss": 1.4221,
+      "step": 8809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.781536340713501,
+      "learning_rate": 1.1949926900584796e-05,
+      "loss": 1.456,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2549471855163574,
+      "learning_rate": 1.1949013157894738e-05,
+      "loss": 1.3899,
+      "step": 8811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.244217872619629,
+      "learning_rate": 1.1948099415204679e-05,
+      "loss": 1.2566,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.124526262283325,
+      "learning_rate": 1.1947185672514621e-05,
+      "loss": 1.9372,
+      "step": 8813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1255478858947754,
+      "learning_rate": 1.1946271929824562e-05,
+      "loss": 1.389,
+      "step": 8814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2788000106811523,
+      "learning_rate": 1.1945358187134504e-05,
+      "loss": 1.3132,
+      "step": 8815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8082211017608643,
+      "learning_rate": 1.1944444444444444e-05,
+      "loss": 1.6909,
+      "step": 8816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.617492914199829,
+      "learning_rate": 1.1943530701754387e-05,
+      "loss": 1.8346,
+      "step": 8817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.320833921432495,
+      "learning_rate": 1.1942616959064329e-05,
+      "loss": 1.3715,
+      "step": 8818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9559757709503174,
+      "learning_rate": 1.194170321637427e-05,
+      "loss": 1.0847,
+      "step": 8819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3141701221466064,
+      "learning_rate": 1.1940789473684212e-05,
+      "loss": 1.3886,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.629235029220581,
+      "learning_rate": 1.1939875730994152e-05,
+      "loss": 1.6717,
+      "step": 8821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7848808765411377,
+      "learning_rate": 1.1938961988304094e-05,
+      "loss": 1.6391,
+      "step": 8822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6500580310821533,
+      "learning_rate": 1.1938048245614035e-05,
+      "loss": 1.4005,
+      "step": 8823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.485553503036499,
+      "learning_rate": 1.1937134502923977e-05,
+      "loss": 1.3736,
+      "step": 8824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.766005039215088,
+      "learning_rate": 1.193622076023392e-05,
+      "loss": 1.4346,
+      "step": 8825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5686373710632324,
+      "learning_rate": 1.193530701754386e-05,
+      "loss": 1.6032,
+      "step": 8826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5501980781555176,
+      "learning_rate": 1.1934393274853802e-05,
+      "loss": 1.3761,
+      "step": 8827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.061967611312866,
+      "learning_rate": 1.1933479532163743e-05,
+      "loss": 1.6915,
+      "step": 8828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.729182004928589,
+      "learning_rate": 1.1932565789473685e-05,
+      "loss": 1.5579,
+      "step": 8829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2631475925445557,
+      "learning_rate": 1.1931652046783625e-05,
+      "loss": 1.4774,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4662539958953857,
+      "learning_rate": 1.1930738304093567e-05,
+      "loss": 1.2326,
+      "step": 8831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9409611225128174,
+      "learning_rate": 1.192982456140351e-05,
+      "loss": 1.5132,
+      "step": 8832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.86983585357666,
+      "learning_rate": 1.192891081871345e-05,
+      "loss": 1.7123,
+      "step": 8833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5236711502075195,
+      "learning_rate": 1.1927997076023392e-05,
+      "loss": 1.5406,
+      "step": 8834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0042121410369873,
+      "learning_rate": 1.1927083333333333e-05,
+      "loss": 1.9137,
+      "step": 8835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.457303285598755,
+      "learning_rate": 1.1926169590643275e-05,
+      "loss": 1.756,
+      "step": 8836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2332637310028076,
+      "learning_rate": 1.1925255847953216e-05,
+      "loss": 1.5045,
+      "step": 8837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6333088874816895,
+      "learning_rate": 1.1924342105263158e-05,
+      "loss": 1.5991,
+      "step": 8838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.230438232421875,
+      "learning_rate": 1.19234283625731e-05,
+      "loss": 1.3429,
+      "step": 8839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7554688453674316,
+      "learning_rate": 1.192251461988304e-05,
+      "loss": 1.9327,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.684488296508789,
+      "learning_rate": 1.1921600877192983e-05,
+      "loss": 1.4275,
+      "step": 8841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8221638202667236,
+      "learning_rate": 1.1920687134502923e-05,
+      "loss": 1.8533,
+      "step": 8842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9352595806121826,
+      "learning_rate": 1.1919773391812866e-05,
+      "loss": 1.9893,
+      "step": 8843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0158982276916504,
+      "learning_rate": 1.1918859649122806e-05,
+      "loss": 1.1685,
+      "step": 8844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3763585090637207,
+      "learning_rate": 1.1917945906432748e-05,
+      "loss": 1.1791,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0732486248016357,
+      "learning_rate": 1.191703216374269e-05,
+      "loss": 1.6609,
+      "step": 8846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5855822563171387,
+      "learning_rate": 1.1916118421052631e-05,
+      "loss": 1.4601,
+      "step": 8847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.600675582885742,
+      "learning_rate": 1.1915204678362573e-05,
+      "loss": 1.6571,
+      "step": 8848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6412057876586914,
+      "learning_rate": 1.1914290935672514e-05,
+      "loss": 1.5245,
+      "step": 8849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.156848192214966,
+      "learning_rate": 1.1913377192982458e-05,
+      "loss": 1.7608,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.58626651763916,
+      "learning_rate": 1.19124634502924e-05,
+      "loss": 1.4445,
+      "step": 8851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.131830930709839,
+      "learning_rate": 1.191154970760234e-05,
+      "loss": 1.9807,
+      "step": 8852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6895627975463867,
+      "learning_rate": 1.1910635964912283e-05,
+      "loss": 1.5114,
+      "step": 8853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7533586025238037,
+      "learning_rate": 1.1909722222222223e-05,
+      "loss": 1.6347,
+      "step": 8854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6908867359161377,
+      "learning_rate": 1.1908808479532166e-05,
+      "loss": 1.6731,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.09757137298584,
+      "learning_rate": 1.1907894736842108e-05,
+      "loss": 1.7564,
+      "step": 8856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5921647548675537,
+      "learning_rate": 1.1906980994152048e-05,
+      "loss": 1.4173,
+      "step": 8857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.411241292953491,
+      "learning_rate": 1.190606725146199e-05,
+      "loss": 1.4773,
+      "step": 8858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3022983074188232,
+      "learning_rate": 1.1905153508771931e-05,
+      "loss": 1.4538,
+      "step": 8859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9970314502716064,
+      "learning_rate": 1.1904239766081873e-05,
+      "loss": 1.6945,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7633562088012695,
+      "learning_rate": 1.1903326023391814e-05,
+      "loss": 1.4606,
+      "step": 8861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8256514072418213,
+      "learning_rate": 1.1902412280701756e-05,
+      "loss": 1.7035,
+      "step": 8862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2826316356658936,
+      "learning_rate": 1.1901498538011698e-05,
+      "loss": 1.5164,
+      "step": 8863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5227770805358887,
+      "learning_rate": 1.1900584795321639e-05,
+      "loss": 1.7361,
+      "step": 8864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.156924247741699,
+      "learning_rate": 1.1899671052631581e-05,
+      "loss": 1.7924,
+      "step": 8865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.859553337097168,
+      "learning_rate": 1.1898757309941522e-05,
+      "loss": 1.5259,
+      "step": 8866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6308298110961914,
+      "learning_rate": 1.1897843567251464e-05,
+      "loss": 1.3503,
+      "step": 8867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.43740177154541,
+      "learning_rate": 1.1896929824561404e-05,
+      "loss": 2.6025,
+      "step": 8868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5816898345947266,
+      "learning_rate": 1.1896016081871347e-05,
+      "loss": 1.4071,
+      "step": 8869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.704275369644165,
+      "learning_rate": 1.1895102339181289e-05,
+      "loss": 1.4544,
+      "step": 8870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.360713481903076,
+      "learning_rate": 1.189418859649123e-05,
+      "loss": 1.3306,
+      "step": 8871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.716484546661377,
+      "learning_rate": 1.1893274853801171e-05,
+      "loss": 1.4626,
+      "step": 8872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4480350017547607,
+      "learning_rate": 1.1892361111111112e-05,
+      "loss": 1.6469,
+      "step": 8873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5633175373077393,
+      "learning_rate": 1.1891447368421054e-05,
+      "loss": 1.4513,
+      "step": 8874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9941134452819824,
+      "learning_rate": 1.1890533625730995e-05,
+      "loss": 1.4651,
+      "step": 8875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.213235378265381,
+      "learning_rate": 1.1889619883040937e-05,
+      "loss": 1.3551,
+      "step": 8876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.668546438217163,
+      "learning_rate": 1.188870614035088e-05,
+      "loss": 1.4129,
+      "step": 8877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3504676818847656,
+      "learning_rate": 1.188779239766082e-05,
+      "loss": 1.5577,
+      "step": 8878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7223520278930664,
+      "learning_rate": 1.1886878654970762e-05,
+      "loss": 1.4954,
+      "step": 8879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6067519187927246,
+      "learning_rate": 1.1885964912280702e-05,
+      "loss": 1.5353,
+      "step": 8880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3687522411346436,
+      "learning_rate": 1.1885051169590645e-05,
+      "loss": 1.5828,
+      "step": 8881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6704535484313965,
+      "learning_rate": 1.1884137426900585e-05,
+      "loss": 1.3135,
+      "step": 8882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.017418622970581,
+      "learning_rate": 1.1883223684210527e-05,
+      "loss": 1.202,
+      "step": 8883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.694859266281128,
+      "learning_rate": 1.188230994152047e-05,
+      "loss": 1.3979,
+      "step": 8884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.291703462600708,
+      "learning_rate": 1.188139619883041e-05,
+      "loss": 1.4651,
+      "step": 8885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7438387870788574,
+      "learning_rate": 1.1880482456140352e-05,
+      "loss": 1.7079,
+      "step": 8886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.721067428588867,
+      "learning_rate": 1.1879568713450293e-05,
+      "loss": 1.6377,
+      "step": 8887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.744957685470581,
+      "learning_rate": 1.1878654970760235e-05,
+      "loss": 1.8778,
+      "step": 8888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.913175344467163,
+      "learning_rate": 1.1877741228070176e-05,
+      "loss": 1.6836,
+      "step": 8889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8604254722595215,
+      "learning_rate": 1.1876827485380118e-05,
+      "loss": 1.5542,
+      "step": 8890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.917757034301758,
+      "learning_rate": 1.187591374269006e-05,
+      "loss": 1.6786,
+      "step": 8891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.40175199508667,
+      "learning_rate": 1.1875e-05,
+      "loss": 1.5285,
+      "step": 8892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.462458848953247,
+      "learning_rate": 1.1874086257309943e-05,
+      "loss": 1.3584,
+      "step": 8893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6753294467926025,
+      "learning_rate": 1.1873172514619883e-05,
+      "loss": 1.248,
+      "step": 8894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8780806064605713,
+      "learning_rate": 1.1872258771929826e-05,
+      "loss": 1.6209,
+      "step": 8895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.187628746032715,
+      "learning_rate": 1.1871345029239766e-05,
+      "loss": 1.0868,
+      "step": 8896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7699108123779297,
+      "learning_rate": 1.1870431286549708e-05,
+      "loss": 1.8196,
+      "step": 8897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3947155475616455,
+      "learning_rate": 1.186951754385965e-05,
+      "loss": 1.7294,
+      "step": 8898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4313294887542725,
+      "learning_rate": 1.1868603801169591e-05,
+      "loss": 1.5189,
+      "step": 8899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9355849027633667,
+      "learning_rate": 1.1867690058479533e-05,
+      "loss": 1.0626,
+      "step": 8900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1572470664978027,
+      "learning_rate": 1.1866776315789474e-05,
+      "loss": 1.7429,
+      "step": 8901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.891238212585449,
+      "learning_rate": 1.1865862573099416e-05,
+      "loss": 1.8863,
+      "step": 8902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.937255620956421,
+      "learning_rate": 1.1864948830409357e-05,
+      "loss": 1.6109,
+      "step": 8903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.719501256942749,
+      "learning_rate": 1.1864035087719299e-05,
+      "loss": 1.6359,
+      "step": 8904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1121721267700195,
+      "learning_rate": 1.1863121345029241e-05,
+      "loss": 2.1789,
+      "step": 8905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.692795515060425,
+      "learning_rate": 1.1862207602339182e-05,
+      "loss": 1.4525,
+      "step": 8906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.357322931289673,
+      "learning_rate": 1.1861293859649124e-05,
+      "loss": 2.0244,
+      "step": 8907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5933053493499756,
+      "learning_rate": 1.1860380116959064e-05,
+      "loss": 1.3617,
+      "step": 8908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3527655601501465,
+      "learning_rate": 1.1859466374269007e-05,
+      "loss": 1.6837,
+      "step": 8909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7120962142944336,
+      "learning_rate": 1.1858552631578947e-05,
+      "loss": 1.5584,
+      "step": 8910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8380656242370605,
+      "learning_rate": 1.185763888888889e-05,
+      "loss": 1.696,
+      "step": 8911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.579371213912964,
+      "learning_rate": 1.1856725146198832e-05,
+      "loss": 1.4171,
+      "step": 8912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.456693172454834,
+      "learning_rate": 1.1855811403508772e-05,
+      "loss": 1.4871,
+      "step": 8913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.482196569442749,
+      "learning_rate": 1.1854897660818714e-05,
+      "loss": 1.362,
+      "step": 8914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.021743059158325,
+      "learning_rate": 1.1853983918128655e-05,
+      "loss": 1.7469,
+      "step": 8915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7239623069763184,
+      "learning_rate": 1.1853070175438597e-05,
+      "loss": 1.2423,
+      "step": 8916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.536306381225586,
+      "learning_rate": 1.1852156432748538e-05,
+      "loss": 1.3267,
+      "step": 8917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3041951656341553,
+      "learning_rate": 1.185124269005848e-05,
+      "loss": 1.4947,
+      "step": 8918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.936046838760376,
+      "learning_rate": 1.1850328947368422e-05,
+      "loss": 1.5706,
+      "step": 8919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9456663131713867,
+      "learning_rate": 1.1849415204678363e-05,
+      "loss": 1.5815,
+      "step": 8920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.829650640487671,
+      "learning_rate": 1.1848501461988305e-05,
+      "loss": 1.4454,
+      "step": 8921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.753697156906128,
+      "learning_rate": 1.1847587719298245e-05,
+      "loss": 1.6565,
+      "step": 8922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5574753284454346,
+      "learning_rate": 1.1846673976608187e-05,
+      "loss": 1.4922,
+      "step": 8923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7039291858673096,
+      "learning_rate": 1.1845760233918128e-05,
+      "loss": 1.2788,
+      "step": 8924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.594778299331665,
+      "learning_rate": 1.184484649122807e-05,
+      "loss": 1.2937,
+      "step": 8925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4908089637756348,
+      "learning_rate": 1.1843932748538012e-05,
+      "loss": 1.4869,
+      "step": 8926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.509678840637207,
+      "learning_rate": 1.1843019005847953e-05,
+      "loss": 1.4774,
+      "step": 8927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3797967433929443,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 1.1742,
+      "step": 8928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2857139110565186,
+      "learning_rate": 1.1841191520467836e-05,
+      "loss": 2.0136,
+      "step": 8929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7774763107299805,
+      "learning_rate": 1.1840277777777778e-05,
+      "loss": 1.5775,
+      "step": 8930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6703178882598877,
+      "learning_rate": 1.1839364035087718e-05,
+      "loss": 1.4694,
+      "step": 8931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.295790433883667,
+      "learning_rate": 1.183845029239766e-05,
+      "loss": 1.2946,
+      "step": 8932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3862202167510986,
+      "learning_rate": 1.1837536549707603e-05,
+      "loss": 1.5164,
+      "step": 8933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.395519256591797,
+      "learning_rate": 1.1836622807017543e-05,
+      "loss": 1.3976,
+      "step": 8934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1038849353790283,
+      "learning_rate": 1.1835709064327487e-05,
+      "loss": 1.4339,
+      "step": 8935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5911457538604736,
+      "learning_rate": 1.183479532163743e-05,
+      "loss": 1.5148,
+      "step": 8936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9891457557678223,
+      "learning_rate": 1.183388157894737e-05,
+      "loss": 1.5701,
+      "step": 8937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4396934509277344,
+      "learning_rate": 1.1832967836257312e-05,
+      "loss": 1.4835,
+      "step": 8938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7545511722564697,
+      "learning_rate": 1.1832054093567253e-05,
+      "loss": 1.9731,
+      "step": 8939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5496788024902344,
+      "learning_rate": 1.1831140350877195e-05,
+      "loss": 1.4524,
+      "step": 8940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6452527046203613,
+      "learning_rate": 1.1830226608187136e-05,
+      "loss": 1.6878,
+      "step": 8941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.045382022857666,
+      "learning_rate": 1.1829312865497078e-05,
+      "loss": 1.927,
+      "step": 8942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9046103954315186,
+      "learning_rate": 1.182839912280702e-05,
+      "loss": 1.6353,
+      "step": 8943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5237925052642822,
+      "learning_rate": 1.182748538011696e-05,
+      "loss": 1.5455,
+      "step": 8944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3349974155426025,
+      "learning_rate": 1.1826571637426903e-05,
+      "loss": 1.2799,
+      "step": 8945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.514871597290039,
+      "learning_rate": 1.1825657894736843e-05,
+      "loss": 1.5719,
+      "step": 8946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5686843395233154,
+      "learning_rate": 1.1824744152046786e-05,
+      "loss": 1.484,
+      "step": 8947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0050923824310303,
+      "learning_rate": 1.1823830409356726e-05,
+      "loss": 1.8106,
+      "step": 8948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.730189085006714,
+      "learning_rate": 1.1822916666666668e-05,
+      "loss": 1.6837,
+      "step": 8949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5875325202941895,
+      "learning_rate": 1.182200292397661e-05,
+      "loss": 1.4672,
+      "step": 8950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9275496006011963,
+      "learning_rate": 1.1821089181286551e-05,
+      "loss": 1.6905,
+      "step": 8951
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7597928047180176,
+      "learning_rate": 1.1820175438596493e-05,
+      "loss": 1.8148,
+      "step": 8952
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.332304000854492,
+      "learning_rate": 1.1819261695906434e-05,
+      "loss": 1.4146,
+      "step": 8953
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1418609619140625,
+      "learning_rate": 1.1818347953216376e-05,
+      "loss": 1.9893,
+      "step": 8954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.50544810295105,
+      "learning_rate": 1.1817434210526317e-05,
+      "loss": 1.4518,
+      "step": 8955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0539233684539795,
+      "learning_rate": 1.1816520467836259e-05,
+      "loss": 2.1231,
+      "step": 8956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4389257431030273,
+      "learning_rate": 1.1815606725146201e-05,
+      "loss": 1.5617,
+      "step": 8957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7879602909088135,
+      "learning_rate": 1.1814692982456142e-05,
+      "loss": 1.3295,
+      "step": 8958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.884474039077759,
+      "learning_rate": 1.1813779239766084e-05,
+      "loss": 1.6948,
+      "step": 8959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.982582688331604,
+      "learning_rate": 1.1812865497076024e-05,
+      "loss": 1.0474,
+      "step": 8960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5021073818206787,
+      "learning_rate": 1.1811951754385967e-05,
+      "loss": 1.5826,
+      "step": 8961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.487870216369629,
+      "learning_rate": 1.1811038011695907e-05,
+      "loss": 1.2762,
+      "step": 8962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.201793670654297,
+      "learning_rate": 1.181012426900585e-05,
+      "loss": 1.3308,
+      "step": 8963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.514094829559326,
+      "learning_rate": 1.1809210526315791e-05,
+      "loss": 1.6223,
+      "step": 8964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4328219890594482,
+      "learning_rate": 1.1808296783625732e-05,
+      "loss": 1.4912,
+      "step": 8965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.732327699661255,
+      "learning_rate": 1.1807383040935674e-05,
+      "loss": 1.6098,
+      "step": 8966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.297302722930908,
+      "learning_rate": 1.1806469298245615e-05,
+      "loss": 1.2829,
+      "step": 8967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.492570638656616,
+      "learning_rate": 1.1805555555555557e-05,
+      "loss": 1.7635,
+      "step": 8968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.399024486541748,
+      "learning_rate": 1.1804641812865497e-05,
+      "loss": 1.5842,
+      "step": 8969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4312400817871094,
+      "learning_rate": 1.180372807017544e-05,
+      "loss": 1.4315,
+      "step": 8970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.680234432220459,
+      "learning_rate": 1.1802814327485382e-05,
+      "loss": 1.3439,
+      "step": 8971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7665343284606934,
+      "learning_rate": 1.1801900584795322e-05,
+      "loss": 1.4938,
+      "step": 8972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.727477550506592,
+      "learning_rate": 1.1800986842105265e-05,
+      "loss": 1.4325,
+      "step": 8973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5414085388183594,
+      "learning_rate": 1.1800073099415205e-05,
+      "loss": 1.3745,
+      "step": 8974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3700733184814453,
+      "learning_rate": 1.1799159356725147e-05,
+      "loss": 1.7017,
+      "step": 8975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3794667720794678,
+      "learning_rate": 1.1798245614035088e-05,
+      "loss": 1.3748,
+      "step": 8976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0699779987335205,
+      "learning_rate": 1.179733187134503e-05,
+      "loss": 2.0355,
+      "step": 8977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2870516777038574,
+      "learning_rate": 1.1796418128654972e-05,
+      "loss": 1.4733,
+      "step": 8978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6391263008117676,
+      "learning_rate": 1.1795504385964913e-05,
+      "loss": 1.5294,
+      "step": 8979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3865416049957275,
+      "learning_rate": 1.1794590643274855e-05,
+      "loss": 1.3817,
+      "step": 8980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.814781665802002,
+      "learning_rate": 1.1793676900584796e-05,
+      "loss": 1.6206,
+      "step": 8981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9641640186309814,
+      "learning_rate": 1.1792763157894738e-05,
+      "loss": 1.5377,
+      "step": 8982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1029510498046875,
+      "learning_rate": 1.1791849415204678e-05,
+      "loss": 1.6967,
+      "step": 8983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1156678199768066,
+      "learning_rate": 1.179093567251462e-05,
+      "loss": 1.9598,
+      "step": 8984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2573137283325195,
+      "learning_rate": 1.1790021929824563e-05,
+      "loss": 1.3272,
+      "step": 8985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6811916828155518,
+      "learning_rate": 1.1789108187134503e-05,
+      "loss": 1.3136,
+      "step": 8986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6124017238616943,
+      "learning_rate": 1.1788194444444446e-05,
+      "loss": 1.5071,
+      "step": 8987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2230350971221924,
+      "learning_rate": 1.1787280701754386e-05,
+      "loss": 1.6837,
+      "step": 8988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1556761264801025,
+      "learning_rate": 1.1786366959064328e-05,
+      "loss": 2.0057,
+      "step": 8989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.854471206665039,
+      "learning_rate": 1.1785453216374269e-05,
+      "loss": 1.6075,
+      "step": 8990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5711677074432373,
+      "learning_rate": 1.1784539473684211e-05,
+      "loss": 1.2284,
+      "step": 8991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.643921375274658,
+      "learning_rate": 1.1783625730994153e-05,
+      "loss": 1.5186,
+      "step": 8992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.279318332672119,
+      "learning_rate": 1.1782711988304094e-05,
+      "loss": 1.4155,
+      "step": 8993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9889156818389893,
+      "learning_rate": 1.1781798245614036e-05,
+      "loss": 1.7643,
+      "step": 8994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4140474796295166,
+      "learning_rate": 1.1780884502923977e-05,
+      "loss": 1.2909,
+      "step": 8995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.218334197998047,
+      "learning_rate": 1.1779970760233919e-05,
+      "loss": 1.0202,
+      "step": 8996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.103243350982666,
+      "learning_rate": 1.177905701754386e-05,
+      "loss": 1.7441,
+      "step": 8997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6945674419403076,
+      "learning_rate": 1.1778143274853802e-05,
+      "loss": 1.2924,
+      "step": 8998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6909782886505127,
+      "learning_rate": 1.1777229532163744e-05,
+      "loss": 1.5743,
+      "step": 8999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9068386554718018,
+      "learning_rate": 1.1776315789473684e-05,
+      "loss": 1.5667,
+      "step": 9000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1799988746643066,
+      "learning_rate": 1.1775402046783627e-05,
+      "loss": 1.8414,
+      "step": 9001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.407949686050415,
+      "learning_rate": 1.1774488304093567e-05,
+      "loss": 1.6152,
+      "step": 9002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8389084339141846,
+      "learning_rate": 1.177357456140351e-05,
+      "loss": 1.5842,
+      "step": 9003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1905903816223145,
+      "learning_rate": 1.177266081871345e-05,
+      "loss": 1.333,
+      "step": 9004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0185744762420654,
+      "learning_rate": 1.1771747076023392e-05,
+      "loss": 1.7567,
+      "step": 9005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5349678993225098,
+      "learning_rate": 1.1770833333333334e-05,
+      "loss": 1.5693,
+      "step": 9006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6541550159454346,
+      "learning_rate": 1.1769919590643275e-05,
+      "loss": 1.5,
+      "step": 9007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2442455291748047,
+      "learning_rate": 1.1769005847953217e-05,
+      "loss": 1.7494,
+      "step": 9008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.867518663406372,
+      "learning_rate": 1.1768092105263158e-05,
+      "loss": 1.9978,
+      "step": 9009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6412558555603027,
+      "learning_rate": 1.17671783625731e-05,
+      "loss": 1.8314,
+      "step": 9010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3225228786468506,
+      "learning_rate": 1.176626461988304e-05,
+      "loss": 1.3206,
+      "step": 9011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3856751918792725,
+      "learning_rate": 1.1765350877192983e-05,
+      "loss": 1.5019,
+      "step": 9012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.58695387840271,
+      "learning_rate": 1.1764437134502925e-05,
+      "loss": 1.339,
+      "step": 9013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7621381282806396,
+      "learning_rate": 1.1763523391812865e-05,
+      "loss": 1.9638,
+      "step": 9014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.764589548110962,
+      "learning_rate": 1.1762609649122807e-05,
+      "loss": 1.5912,
+      "step": 9015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6529135704040527,
+      "learning_rate": 1.1761695906432748e-05,
+      "loss": 1.6571,
+      "step": 9016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4303038120269775,
+      "learning_rate": 1.176078216374269e-05,
+      "loss": 1.7832,
+      "step": 9017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.820768117904663,
+      "learning_rate": 1.175986842105263e-05,
+      "loss": 1.7777,
+      "step": 9018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.617654323577881,
+      "learning_rate": 1.1758954678362573e-05,
+      "loss": 1.5723,
+      "step": 9019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.54986572265625,
+      "learning_rate": 1.1758040935672515e-05,
+      "loss": 1.5128,
+      "step": 9020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9126205444335938,
+      "learning_rate": 1.1757127192982457e-05,
+      "loss": 1.5759,
+      "step": 9021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.162891387939453,
+      "learning_rate": 1.17562134502924e-05,
+      "loss": 1.483,
+      "step": 9022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.454857110977173,
+      "learning_rate": 1.1755299707602342e-05,
+      "loss": 1.3355,
+      "step": 9023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.132641315460205,
+      "learning_rate": 1.1754385964912282e-05,
+      "loss": 1.2278,
+      "step": 9024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.560426950454712,
+      "learning_rate": 1.1753472222222225e-05,
+      "loss": 1.7424,
+      "step": 9025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.459484577178955,
+      "learning_rate": 1.1752558479532165e-05,
+      "loss": 1.4111,
+      "step": 9026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.541802406311035,
+      "learning_rate": 1.1751644736842107e-05,
+      "loss": 1.5665,
+      "step": 9027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.697721004486084,
+      "learning_rate": 1.1750730994152048e-05,
+      "loss": 1.7706,
+      "step": 9028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.509622573852539,
+      "learning_rate": 1.174981725146199e-05,
+      "loss": 1.5044,
+      "step": 9029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.947843313217163,
+      "learning_rate": 1.1748903508771932e-05,
+      "loss": 1.7538,
+      "step": 9030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.785370349884033,
+      "learning_rate": 1.1747989766081873e-05,
+      "loss": 1.5355,
+      "step": 9031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.433964252471924,
+      "learning_rate": 1.1747076023391815e-05,
+      "loss": 1.5834,
+      "step": 9032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5391974449157715,
+      "learning_rate": 1.1746162280701756e-05,
+      "loss": 1.6184,
+      "step": 9033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8980953693389893,
+      "learning_rate": 1.1745248538011698e-05,
+      "loss": 1.8407,
+      "step": 9034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.939052104949951,
+      "learning_rate": 1.1744334795321638e-05,
+      "loss": 1.4079,
+      "step": 9035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0786638259887695,
+      "learning_rate": 1.174342105263158e-05,
+      "loss": 1.1028,
+      "step": 9036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5460939407348633,
+      "learning_rate": 1.1742507309941523e-05,
+      "loss": 1.5546,
+      "step": 9037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.293808698654175,
+      "learning_rate": 1.1741593567251463e-05,
+      "loss": 1.4303,
+      "step": 9038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7564303874969482,
+      "learning_rate": 1.1740679824561406e-05,
+      "loss": 1.7206,
+      "step": 9039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.877074956893921,
+      "learning_rate": 1.1739766081871346e-05,
+      "loss": 1.0417,
+      "step": 9040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6150429248809814,
+      "learning_rate": 1.1738852339181288e-05,
+      "loss": 1.4512,
+      "step": 9041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5337674617767334,
+      "learning_rate": 1.1737938596491229e-05,
+      "loss": 1.5324,
+      "step": 9042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.007753849029541,
+      "learning_rate": 1.1737024853801171e-05,
+      "loss": 1.7618,
+      "step": 9043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0367393493652344,
+      "learning_rate": 1.1736111111111113e-05,
+      "loss": 1.5438,
+      "step": 9044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8219010829925537,
+      "learning_rate": 1.1735197368421054e-05,
+      "loss": 1.6313,
+      "step": 9045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2076735496520996,
+      "learning_rate": 1.1734283625730996e-05,
+      "loss": 1.4361,
+      "step": 9046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2458226680755615,
+      "learning_rate": 1.1733369883040937e-05,
+      "loss": 1.8312,
+      "step": 9047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.910271167755127,
+      "learning_rate": 1.1732456140350879e-05,
+      "loss": 1.6408,
+      "step": 9048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0975584983825684,
+      "learning_rate": 1.173154239766082e-05,
+      "loss": 1.2513,
+      "step": 9049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0828514099121094,
+      "learning_rate": 1.1730628654970762e-05,
+      "loss": 1.6811,
+      "step": 9050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2530133724212646,
+      "learning_rate": 1.1729714912280704e-05,
+      "loss": 1.2574,
+      "step": 9051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8526108264923096,
+      "learning_rate": 1.1728801169590644e-05,
+      "loss": 1.043,
+      "step": 9052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6698360443115234,
+      "learning_rate": 1.1727887426900586e-05,
+      "loss": 1.7589,
+      "step": 9053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0660347938537598,
+      "learning_rate": 1.1726973684210527e-05,
+      "loss": 1.3043,
+      "step": 9054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0034420490264893,
+      "learning_rate": 1.172605994152047e-05,
+      "loss": 1.4936,
+      "step": 9055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.483353853225708,
+      "learning_rate": 1.172514619883041e-05,
+      "loss": 1.4351,
+      "step": 9056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.291778564453125,
+      "learning_rate": 1.1724232456140352e-05,
+      "loss": 1.4653,
+      "step": 9057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3201611042022705,
+      "learning_rate": 1.1723318713450294e-05,
+      "loss": 1.2685,
+      "step": 9058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6112964153289795,
+      "learning_rate": 1.1722404970760235e-05,
+      "loss": 1.7661,
+      "step": 9059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9814577102661133,
+      "learning_rate": 1.1721491228070177e-05,
+      "loss": 1.8358,
+      "step": 9060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7862582206726074,
+      "learning_rate": 1.1720577485380117e-05,
+      "loss": 1.745,
+      "step": 9061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1004700660705566,
+      "learning_rate": 1.171966374269006e-05,
+      "loss": 1.5625,
+      "step": 9062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.397723436355591,
+      "learning_rate": 1.171875e-05,
+      "loss": 1.3378,
+      "step": 9063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8699843883514404,
+      "learning_rate": 1.1717836257309942e-05,
+      "loss": 1.6486,
+      "step": 9064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6841559410095215,
+      "learning_rate": 1.1716922514619885e-05,
+      "loss": 1.7046,
+      "step": 9065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5746240615844727,
+      "learning_rate": 1.1716008771929825e-05,
+      "loss": 1.544,
+      "step": 9066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.455310106277466,
+      "learning_rate": 1.1715095029239767e-05,
+      "loss": 1.2352,
+      "step": 9067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.401076316833496,
+      "learning_rate": 1.1714181286549708e-05,
+      "loss": 1.373,
+      "step": 9068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.282219171524048,
+      "learning_rate": 1.171326754385965e-05,
+      "loss": 1.4571,
+      "step": 9069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8674561977386475,
+      "learning_rate": 1.171235380116959e-05,
+      "loss": 1.6438,
+      "step": 9070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6022744178771973,
+      "learning_rate": 1.1711440058479533e-05,
+      "loss": 1.843,
+      "step": 9071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.738084554672241,
+      "learning_rate": 1.1710526315789475e-05,
+      "loss": 1.7913,
+      "step": 9072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7183172702789307,
+      "learning_rate": 1.1709612573099416e-05,
+      "loss": 1.4631,
+      "step": 9073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6566104888916016,
+      "learning_rate": 1.1708698830409358e-05,
+      "loss": 1.6668,
+      "step": 9074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.768254041671753,
+      "learning_rate": 1.1707785087719298e-05,
+      "loss": 1.52,
+      "step": 9075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.013960838317871,
+      "learning_rate": 1.170687134502924e-05,
+      "loss": 1.8115,
+      "step": 9076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3718409538269043,
+      "learning_rate": 1.1705957602339181e-05,
+      "loss": 1.3068,
+      "step": 9077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6487011909484863,
+      "learning_rate": 1.1705043859649123e-05,
+      "loss": 1.4717,
+      "step": 9078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.988271474838257,
+      "learning_rate": 1.1704130116959066e-05,
+      "loss": 1.4286,
+      "step": 9079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7136359214782715,
+      "learning_rate": 1.1703216374269006e-05,
+      "loss": 1.7566,
+      "step": 9080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.73012638092041,
+      "learning_rate": 1.1702302631578948e-05,
+      "loss": 1.4588,
+      "step": 9081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.642465829849243,
+      "learning_rate": 1.1701388888888889e-05,
+      "loss": 1.3601,
+      "step": 9082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6092145442962646,
+      "learning_rate": 1.1700475146198831e-05,
+      "loss": 1.2897,
+      "step": 9083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6257734298706055,
+      "learning_rate": 1.1699561403508772e-05,
+      "loss": 1.2196,
+      "step": 9084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.256704092025757,
+      "learning_rate": 1.1698647660818714e-05,
+      "loss": 1.6626,
+      "step": 9085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.442920207977295,
+      "learning_rate": 1.1697733918128656e-05,
+      "loss": 1.6719,
+      "step": 9086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4132308959960938,
+      "learning_rate": 1.1696820175438597e-05,
+      "loss": 1.272,
+      "step": 9087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.424104928970337,
+      "learning_rate": 1.1695906432748539e-05,
+      "loss": 1.503,
+      "step": 9088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5190258026123047,
+      "learning_rate": 1.169499269005848e-05,
+      "loss": 1.5067,
+      "step": 9089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7093303203582764,
+      "learning_rate": 1.1694078947368422e-05,
+      "loss": 1.6096,
+      "step": 9090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7261135578155518,
+      "learning_rate": 1.1693165204678362e-05,
+      "loss": 1.5275,
+      "step": 9091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.152250289916992,
+      "learning_rate": 1.1692251461988304e-05,
+      "loss": 1.3274,
+      "step": 9092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.755185842514038,
+      "learning_rate": 1.1691337719298247e-05,
+      "loss": 1.4968,
+      "step": 9093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4448726177215576,
+      "learning_rate": 1.1690423976608187e-05,
+      "loss": 1.5591,
+      "step": 9094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4938149452209473,
+      "learning_rate": 1.168951023391813e-05,
+      "loss": 1.4953,
+      "step": 9095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6330440044403076,
+      "learning_rate": 1.168859649122807e-05,
+      "loss": 1.6318,
+      "step": 9096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.519773483276367,
+      "learning_rate": 1.1687682748538012e-05,
+      "loss": 1.433,
+      "step": 9097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3808491230010986,
+      "learning_rate": 1.1686769005847953e-05,
+      "loss": 1.161,
+      "step": 9098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1534347534179688,
+      "learning_rate": 1.1685855263157895e-05,
+      "loss": 1.399,
+      "step": 9099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.775721311569214,
+      "learning_rate": 1.1684941520467837e-05,
+      "loss": 1.6041,
+      "step": 9100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4281187057495117,
+      "learning_rate": 1.1684027777777778e-05,
+      "loss": 1.267,
+      "step": 9101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.613696813583374,
+      "learning_rate": 1.168311403508772e-05,
+      "loss": 1.6376,
+      "step": 9102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.826735734939575,
+      "learning_rate": 1.168220029239766e-05,
+      "loss": 1.3133,
+      "step": 9103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8502910137176514,
+      "learning_rate": 1.1681286549707603e-05,
+      "loss": 1.7194,
+      "step": 9104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3988864421844482,
+      "learning_rate": 1.1680372807017543e-05,
+      "loss": 1.5105,
+      "step": 9105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.51981520652771,
+      "learning_rate": 1.1679459064327487e-05,
+      "loss": 1.0236,
+      "step": 9106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9104397296905518,
+      "learning_rate": 1.167854532163743e-05,
+      "loss": 1.4354,
+      "step": 9107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3524816036224365,
+      "learning_rate": 1.167763157894737e-05,
+      "loss": 1.1869,
+      "step": 9108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4529924392700195,
+      "learning_rate": 1.1676717836257312e-05,
+      "loss": 1.4421,
+      "step": 9109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.828883647918701,
+      "learning_rate": 1.1675804093567254e-05,
+      "loss": 1.4322,
+      "step": 9110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9048538208007812,
+      "learning_rate": 1.1674890350877195e-05,
+      "loss": 1.4239,
+      "step": 9111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7220354080200195,
+      "learning_rate": 1.1673976608187137e-05,
+      "loss": 1.5395,
+      "step": 9112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.609447717666626,
+      "learning_rate": 1.1673062865497077e-05,
+      "loss": 1.4775,
+      "step": 9113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7602689266204834,
+      "learning_rate": 1.167214912280702e-05,
+      "loss": 1.4874,
+      "step": 9114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9582126140594482,
+      "learning_rate": 1.167123538011696e-05,
+      "loss": 1.5211,
+      "step": 9115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6626670360565186,
+      "learning_rate": 1.1670321637426902e-05,
+      "loss": 1.5186,
+      "step": 9116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.97023606300354,
+      "learning_rate": 1.1669407894736845e-05,
+      "loss": 1.5272,
+      "step": 9117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.597947359085083,
+      "learning_rate": 1.1668494152046785e-05,
+      "loss": 1.4995,
+      "step": 9118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5168628692626953,
+      "learning_rate": 1.1667580409356727e-05,
+      "loss": 1.8942,
+      "step": 9119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6031715869903564,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.2365,
+      "step": 9120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6313068866729736,
+      "learning_rate": 1.166575292397661e-05,
+      "loss": 1.6046,
+      "step": 9121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3794965744018555,
+      "learning_rate": 1.166483918128655e-05,
+      "loss": 1.5603,
+      "step": 9122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9202306270599365,
+      "learning_rate": 1.1663925438596493e-05,
+      "loss": 1.812,
+      "step": 9123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6658411026000977,
+      "learning_rate": 1.1663011695906435e-05,
+      "loss": 1.6226,
+      "step": 9124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8217971324920654,
+      "learning_rate": 1.1662097953216376e-05,
+      "loss": 1.8109,
+      "step": 9125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4469077587127686,
+      "learning_rate": 1.1661184210526318e-05,
+      "loss": 1.477,
+      "step": 9126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.079043388366699,
+      "learning_rate": 1.1660270467836258e-05,
+      "loss": 1.5601,
+      "step": 9127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.946125030517578,
+      "learning_rate": 1.16593567251462e-05,
+      "loss": 1.9447,
+      "step": 9128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.80202317237854,
+      "learning_rate": 1.1658442982456141e-05,
+      "loss": 1.3168,
+      "step": 9129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1341288089752197,
+      "learning_rate": 1.1657529239766083e-05,
+      "loss": 1.1495,
+      "step": 9130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9176483154296875,
+      "learning_rate": 1.1656615497076026e-05,
+      "loss": 1.5991,
+      "step": 9131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3462703227996826,
+      "learning_rate": 1.1655701754385966e-05,
+      "loss": 1.4557,
+      "step": 9132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6561427116394043,
+      "learning_rate": 1.1654788011695908e-05,
+      "loss": 1.4698,
+      "step": 9133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0619912147521973,
+      "learning_rate": 1.1653874269005849e-05,
+      "loss": 1.6781,
+      "step": 9134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6857142448425293,
+      "learning_rate": 1.1652960526315791e-05,
+      "loss": 1.8372,
+      "step": 9135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.907365083694458,
+      "learning_rate": 1.1652046783625732e-05,
+      "loss": 1.4951,
+      "step": 9136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3731191158294678,
+      "learning_rate": 1.1651133040935674e-05,
+      "loss": 1.3773,
+      "step": 9137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7966785430908203,
+      "learning_rate": 1.1650219298245616e-05,
+      "loss": 1.544,
+      "step": 9138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4808595180511475,
+      "learning_rate": 1.1649305555555557e-05,
+      "loss": 1.3792,
+      "step": 9139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.263568162918091,
+      "learning_rate": 1.1648391812865499e-05,
+      "loss": 1.4648,
+      "step": 9140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.413710832595825,
+      "learning_rate": 1.164747807017544e-05,
+      "loss": 1.5816,
+      "step": 9141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.316293239593506,
+      "learning_rate": 1.1646564327485382e-05,
+      "loss": 2.0395,
+      "step": 9142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6045939922332764,
+      "learning_rate": 1.1645650584795322e-05,
+      "loss": 1.6157,
+      "step": 9143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6347291469573975,
+      "learning_rate": 1.1644736842105264e-05,
+      "loss": 1.4721,
+      "step": 9144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.983980178833008,
+      "learning_rate": 1.1643823099415206e-05,
+      "loss": 1.1957,
+      "step": 9145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5814502239227295,
+      "learning_rate": 1.1642909356725147e-05,
+      "loss": 1.5718,
+      "step": 9146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.421022891998291,
+      "learning_rate": 1.164199561403509e-05,
+      "loss": 1.3511,
+      "step": 9147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5480237007141113,
+      "learning_rate": 1.164108187134503e-05,
+      "loss": 1.5281,
+      "step": 9148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5712172985076904,
+      "learning_rate": 1.1640168128654972e-05,
+      "loss": 1.4071,
+      "step": 9149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.689406156539917,
+      "learning_rate": 1.1639254385964913e-05,
+      "loss": 1.3679,
+      "step": 9150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.422909736633301,
+      "learning_rate": 1.1638340643274855e-05,
+      "loss": 1.3714,
+      "step": 9151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7184228897094727,
+      "learning_rate": 1.1637426900584797e-05,
+      "loss": 1.7768,
+      "step": 9152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1879782676696777,
+      "learning_rate": 1.1636513157894737e-05,
+      "loss": 1.6426,
+      "step": 9153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7492289543151855,
+      "learning_rate": 1.163559941520468e-05,
+      "loss": 1.4188,
+      "step": 9154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4956037998199463,
+      "learning_rate": 1.163468567251462e-05,
+      "loss": 1.3511,
+      "step": 9155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8053808212280273,
+      "learning_rate": 1.1633771929824562e-05,
+      "loss": 1.6437,
+      "step": 9156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1185314655303955,
+      "learning_rate": 1.1632858187134503e-05,
+      "loss": 1.1736,
+      "step": 9157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.164787530899048,
+      "learning_rate": 1.1631944444444445e-05,
+      "loss": 1.2977,
+      "step": 9158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5365209579467773,
+      "learning_rate": 1.1631030701754387e-05,
+      "loss": 1.4171,
+      "step": 9159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.864055633544922,
+      "learning_rate": 1.1630116959064328e-05,
+      "loss": 1.7185,
+      "step": 9160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.722062349319458,
+      "learning_rate": 1.162920321637427e-05,
+      "loss": 1.4809,
+      "step": 9161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9503135681152344,
+      "learning_rate": 1.162828947368421e-05,
+      "loss": 1.4287,
+      "step": 9162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2110822200775146,
+      "learning_rate": 1.1627375730994153e-05,
+      "loss": 1.1445,
+      "step": 9163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6428470611572266,
+      "learning_rate": 1.1626461988304093e-05,
+      "loss": 1.0628,
+      "step": 9164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4025824069976807,
+      "learning_rate": 1.1625548245614036e-05,
+      "loss": 1.0984,
+      "step": 9165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.875511646270752,
+      "learning_rate": 1.1624634502923978e-05,
+      "loss": 1.7639,
+      "step": 9166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.418741464614868,
+      "learning_rate": 1.1623720760233918e-05,
+      "loss": 2.033,
+      "step": 9167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7317795753479004,
+      "learning_rate": 1.162280701754386e-05,
+      "loss": 1.3082,
+      "step": 9168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.556497097015381,
+      "learning_rate": 1.1621893274853801e-05,
+      "loss": 1.5573,
+      "step": 9169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2008955478668213,
+      "learning_rate": 1.1620979532163743e-05,
+      "loss": 1.2682,
+      "step": 9170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.870394706726074,
+      "learning_rate": 1.1620065789473684e-05,
+      "loss": 1.4157,
+      "step": 9171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.524177312850952,
+      "learning_rate": 1.1619152046783626e-05,
+      "loss": 1.5344,
+      "step": 9172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3316097259521484,
+      "learning_rate": 1.1618238304093568e-05,
+      "loss": 1.5051,
+      "step": 9173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4607157707214355,
+      "learning_rate": 1.1617324561403509e-05,
+      "loss": 1.526,
+      "step": 9174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7905709743499756,
+      "learning_rate": 1.1616410818713451e-05,
+      "loss": 1.3141,
+      "step": 9175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3705849647521973,
+      "learning_rate": 1.1615497076023392e-05,
+      "loss": 1.5967,
+      "step": 9176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0225110054016113,
+      "learning_rate": 1.1614583333333334e-05,
+      "loss": 1.5994,
+      "step": 9177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.012685775756836,
+      "learning_rate": 1.1613669590643274e-05,
+      "loss": 1.7424,
+      "step": 9178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.800238847732544,
+      "learning_rate": 1.1612755847953217e-05,
+      "loss": 1.4375,
+      "step": 9179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6135218143463135,
+      "learning_rate": 1.1611842105263159e-05,
+      "loss": 1.6499,
+      "step": 9180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.546855926513672,
+      "learning_rate": 1.16109283625731e-05,
+      "loss": 1.4668,
+      "step": 9181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7396697998046875,
+      "learning_rate": 1.1610014619883042e-05,
+      "loss": 1.7407,
+      "step": 9182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8102471828460693,
+      "learning_rate": 1.1609100877192982e-05,
+      "loss": 1.7019,
+      "step": 9183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.663358688354492,
+      "learning_rate": 1.1608187134502924e-05,
+      "loss": 1.5058,
+      "step": 9184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.583542585372925,
+      "learning_rate": 1.1607273391812865e-05,
+      "loss": 1.7927,
+      "step": 9185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6924452781677246,
+      "learning_rate": 1.1606359649122807e-05,
+      "loss": 1.6608,
+      "step": 9186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8047685623168945,
+      "learning_rate": 1.160544590643275e-05,
+      "loss": 1.6985,
+      "step": 9187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.627220869064331,
+      "learning_rate": 1.160453216374269e-05,
+      "loss": 1.5438,
+      "step": 9188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8431334495544434,
+      "learning_rate": 1.1603618421052632e-05,
+      "loss": 2.0121,
+      "step": 9189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.778745412826538,
+      "learning_rate": 1.1602704678362573e-05,
+      "loss": 1.4999,
+      "step": 9190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.607210159301758,
+      "learning_rate": 1.1601790935672515e-05,
+      "loss": 1.4799,
+      "step": 9191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5602259635925293,
+      "learning_rate": 1.1600877192982459e-05,
+      "loss": 2.0697,
+      "step": 9192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.656566619873047,
+      "learning_rate": 1.15999634502924e-05,
+      "loss": 1.518,
+      "step": 9193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8943750858306885,
+      "learning_rate": 1.1599049707602341e-05,
+      "loss": 1.6426,
+      "step": 9194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.411055088043213,
+      "learning_rate": 1.1598135964912282e-05,
+      "loss": 1.5029,
+      "step": 9195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9147592782974243,
+      "learning_rate": 1.1597222222222224e-05,
+      "loss": 1.1737,
+      "step": 9196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.58246111869812,
+      "learning_rate": 1.1596308479532166e-05,
+      "loss": 1.3322,
+      "step": 9197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0504636764526367,
+      "learning_rate": 1.1595394736842107e-05,
+      "loss": 1.8184,
+      "step": 9198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5513482093811035,
+      "learning_rate": 1.159448099415205e-05,
+      "loss": 1.2185,
+      "step": 9199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.157555103302002,
+      "learning_rate": 1.159356725146199e-05,
+      "loss": 1.8333,
+      "step": 9200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2609498500823975,
+      "learning_rate": 1.1592653508771932e-05,
+      "loss": 1.1978,
+      "step": 9201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7643754482269287,
+      "learning_rate": 1.1591739766081872e-05,
+      "loss": 1.7893,
+      "step": 9202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.610469102859497,
+      "learning_rate": 1.1590826023391815e-05,
+      "loss": 1.6191,
+      "step": 9203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5850327014923096,
+      "learning_rate": 1.1589912280701757e-05,
+      "loss": 1.405,
+      "step": 9204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2246809005737305,
+      "learning_rate": 1.1588998538011697e-05,
+      "loss": 1.0484,
+      "step": 9205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.358023166656494,
+      "learning_rate": 1.158808479532164e-05,
+      "loss": 1.2439,
+      "step": 9206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3762028217315674,
+      "learning_rate": 1.158717105263158e-05,
+      "loss": 1.6202,
+      "step": 9207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8729019165039062,
+      "learning_rate": 1.1586257309941522e-05,
+      "loss": 1.5193,
+      "step": 9208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.59641695022583,
+      "learning_rate": 1.1585343567251463e-05,
+      "loss": 1.4363,
+      "step": 9209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.727447032928467,
+      "learning_rate": 1.1584429824561405e-05,
+      "loss": 1.7745,
+      "step": 9210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1695468425750732,
+      "learning_rate": 1.1583516081871347e-05,
+      "loss": 1.6227,
+      "step": 9211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.179668664932251,
+      "learning_rate": 1.1582602339181288e-05,
+      "loss": 1.2225,
+      "step": 9212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.196295976638794,
+      "learning_rate": 1.158168859649123e-05,
+      "loss": 1.8412,
+      "step": 9213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.709207773208618,
+      "learning_rate": 1.158077485380117e-05,
+      "loss": 1.6772,
+      "step": 9214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5421676635742188,
+      "learning_rate": 1.1579861111111113e-05,
+      "loss": 1.5554,
+      "step": 9215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.793771743774414,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 1.7272,
+      "step": 9216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2659945487976074,
+      "learning_rate": 1.1578033625730996e-05,
+      "loss": 1.522,
+      "step": 9217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.527883768081665,
+      "learning_rate": 1.1577119883040938e-05,
+      "loss": 1.4438,
+      "step": 9218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0465664863586426,
+      "learning_rate": 1.1576206140350878e-05,
+      "loss": 1.2382,
+      "step": 9219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7513837814331055,
+      "learning_rate": 1.157529239766082e-05,
+      "loss": 1.2742,
+      "step": 9220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7528927326202393,
+      "learning_rate": 1.1574378654970761e-05,
+      "loss": 1.7083,
+      "step": 9221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7609567642211914,
+      "learning_rate": 1.1573464912280703e-05,
+      "loss": 1.4347,
+      "step": 9222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4672484397888184,
+      "learning_rate": 1.1572551169590644e-05,
+      "loss": 1.2963,
+      "step": 9223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.244495391845703,
+      "learning_rate": 1.1571637426900586e-05,
+      "loss": 1.2622,
+      "step": 9224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.507012367248535,
+      "learning_rate": 1.1570723684210528e-05,
+      "loss": 1.324,
+      "step": 9225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.015136957168579,
+      "learning_rate": 1.1569809941520469e-05,
+      "loss": 1.9427,
+      "step": 9226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.328848123550415,
+      "learning_rate": 1.1568896198830411e-05,
+      "loss": 1.3588,
+      "step": 9227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.79724383354187,
+      "learning_rate": 1.1567982456140352e-05,
+      "loss": 1.7382,
+      "step": 9228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.684563398361206,
+      "learning_rate": 1.1567068713450294e-05,
+      "loss": 1.705,
+      "step": 9229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.522634506225586,
+      "learning_rate": 1.1566154970760234e-05,
+      "loss": 1.5501,
+      "step": 9230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.017043352127075,
+      "learning_rate": 1.1565241228070177e-05,
+      "loss": 1.4336,
+      "step": 9231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0150954723358154,
+      "learning_rate": 1.1564327485380119e-05,
+      "loss": 1.5397,
+      "step": 9232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.782973527908325,
+      "learning_rate": 1.156341374269006e-05,
+      "loss": 1.2782,
+      "step": 9233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.102226972579956,
+      "learning_rate": 1.1562500000000002e-05,
+      "loss": 1.7701,
+      "step": 9234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1064398288726807,
+      "learning_rate": 1.1561586257309942e-05,
+      "loss": 1.1815,
+      "step": 9235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6564879417419434,
+      "learning_rate": 1.1560672514619884e-05,
+      "loss": 1.3085,
+      "step": 9236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6337544918060303,
+      "learning_rate": 1.1559758771929825e-05,
+      "loss": 1.4678,
+      "step": 9237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.833226203918457,
+      "learning_rate": 1.1558845029239767e-05,
+      "loss": 1.8372,
+      "step": 9238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.747025728225708,
+      "learning_rate": 1.155793128654971e-05,
+      "loss": 1.6461,
+      "step": 9239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.81048583984375,
+      "learning_rate": 1.155701754385965e-05,
+      "loss": 1.7184,
+      "step": 9240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.782581329345703,
+      "learning_rate": 1.1556103801169592e-05,
+      "loss": 1.749,
+      "step": 9241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4858975410461426,
+      "learning_rate": 1.1555190058479533e-05,
+      "loss": 1.4062,
+      "step": 9242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.902463912963867,
+      "learning_rate": 1.1554276315789475e-05,
+      "loss": 1.796,
+      "step": 9243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1019692420959473,
+      "learning_rate": 1.1553362573099415e-05,
+      "loss": 1.7348,
+      "step": 9244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.306407928466797,
+      "learning_rate": 1.1552448830409357e-05,
+      "loss": 1.8045,
+      "step": 9245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.387644052505493,
+      "learning_rate": 1.15515350877193e-05,
+      "loss": 1.2436,
+      "step": 9246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0543363094329834,
+      "learning_rate": 1.155062134502924e-05,
+      "loss": 1.701,
+      "step": 9247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7866742610931396,
+      "learning_rate": 1.1549707602339182e-05,
+      "loss": 1.7635,
+      "step": 9248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6527795791625977,
+      "learning_rate": 1.1548793859649123e-05,
+      "loss": 1.625,
+      "step": 9249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.592907428741455,
+      "learning_rate": 1.1547880116959065e-05,
+      "loss": 1.1899,
+      "step": 9250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8552794456481934,
+      "learning_rate": 1.1546966374269006e-05,
+      "loss": 1.9137,
+      "step": 9251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.577019691467285,
+      "learning_rate": 1.1546052631578948e-05,
+      "loss": 1.6827,
+      "step": 9252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4954614639282227,
+      "learning_rate": 1.154513888888889e-05,
+      "loss": 1.8034,
+      "step": 9253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.783738374710083,
+      "learning_rate": 1.154422514619883e-05,
+      "loss": 1.7451,
+      "step": 9254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6899263858795166,
+      "learning_rate": 1.1543311403508773e-05,
+      "loss": 1.4721,
+      "step": 9255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8609817028045654,
+      "learning_rate": 1.1542397660818713e-05,
+      "loss": 1.3594,
+      "step": 9256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.489534854888916,
+      "learning_rate": 1.1541483918128656e-05,
+      "loss": 1.4059,
+      "step": 9257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.908658027648926,
+      "learning_rate": 1.1540570175438596e-05,
+      "loss": 1.5483,
+      "step": 9258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7145512104034424,
+      "learning_rate": 1.1539656432748538e-05,
+      "loss": 1.5416,
+      "step": 9259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8549582958221436,
+      "learning_rate": 1.1538742690058479e-05,
+      "loss": 1.5007,
+      "step": 9260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8400163650512695,
+      "learning_rate": 1.1537828947368421e-05,
+      "loss": 1.2668,
+      "step": 9261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8555378913879395,
+      "learning_rate": 1.1536915204678363e-05,
+      "loss": 1.519,
+      "step": 9262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.809999704360962,
+      "learning_rate": 1.1536001461988304e-05,
+      "loss": 1.9104,
+      "step": 9263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4358584880828857,
+      "learning_rate": 1.1535087719298246e-05,
+      "loss": 1.3955,
+      "step": 9264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6467082500457764,
+      "learning_rate": 1.1534173976608187e-05,
+      "loss": 1.5816,
+      "step": 9265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.511387825012207,
+      "learning_rate": 1.1533260233918129e-05,
+      "loss": 1.4035,
+      "step": 9266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1009228229522705,
+      "learning_rate": 1.153234649122807e-05,
+      "loss": 1.955,
+      "step": 9267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.477221965789795,
+      "learning_rate": 1.1531432748538012e-05,
+      "loss": 1.4421,
+      "step": 9268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.193577289581299,
+      "learning_rate": 1.1530519005847954e-05,
+      "loss": 1.8558,
+      "step": 9269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7860255241394043,
+      "learning_rate": 1.1529605263157894e-05,
+      "loss": 1.753,
+      "step": 9270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7835516929626465,
+      "learning_rate": 1.1528691520467837e-05,
+      "loss": 1.5096,
+      "step": 9271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.743309259414673,
+      "learning_rate": 1.1527777777777777e-05,
+      "loss": 1.5759,
+      "step": 9272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2452104091644287,
+      "learning_rate": 1.152686403508772e-05,
+      "loss": 2.1482,
+      "step": 9273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.320422887802124,
+      "learning_rate": 1.152595029239766e-05,
+      "loss": 1.4776,
+      "step": 9274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6949779987335205,
+      "learning_rate": 1.1525036549707602e-05,
+      "loss": 1.5561,
+      "step": 9275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.650895595550537,
+      "learning_rate": 1.1524122807017544e-05,
+      "loss": 1.6574,
+      "step": 9276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5771074295043945,
+      "learning_rate": 1.1523209064327488e-05,
+      "loss": 1.3528,
+      "step": 9277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.017883777618408,
+      "learning_rate": 1.1522295321637429e-05,
+      "loss": 1.6667,
+      "step": 9278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9973456859588623,
+      "learning_rate": 1.1521381578947371e-05,
+      "loss": 1.0605,
+      "step": 9279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0587823390960693,
+      "learning_rate": 1.1520467836257312e-05,
+      "loss": 1.4286,
+      "step": 9280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.279106616973877,
+      "learning_rate": 1.1519554093567254e-05,
+      "loss": 1.4778,
+      "step": 9281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1473753452301025,
+      "learning_rate": 1.1518640350877194e-05,
+      "loss": 1.5602,
+      "step": 9282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.800908327102661,
+      "learning_rate": 1.1517726608187136e-05,
+      "loss": 1.447,
+      "step": 9283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.854557514190674,
+      "learning_rate": 1.1516812865497079e-05,
+      "loss": 1.5292,
+      "step": 9284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.565795421600342,
+      "learning_rate": 1.151589912280702e-05,
+      "loss": 1.4151,
+      "step": 9285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.573317289352417,
+      "learning_rate": 1.1514985380116961e-05,
+      "loss": 1.5056,
+      "step": 9286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.592122793197632,
+      "learning_rate": 1.1514071637426902e-05,
+      "loss": 1.3834,
+      "step": 9287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.090343475341797,
+      "learning_rate": 1.1513157894736844e-05,
+      "loss": 1.4123,
+      "step": 9288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.878970146179199,
+      "learning_rate": 1.1512244152046785e-05,
+      "loss": 2.013,
+      "step": 9289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.284437894821167,
+      "learning_rate": 1.1511330409356727e-05,
+      "loss": 1.2232,
+      "step": 9290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1654586791992188,
+      "learning_rate": 1.151041666666667e-05,
+      "loss": 1.3056,
+      "step": 9291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.515129566192627,
+      "learning_rate": 1.150950292397661e-05,
+      "loss": 1.2818,
+      "step": 9292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4445993900299072,
+      "learning_rate": 1.1508589181286552e-05,
+      "loss": 1.6607,
+      "step": 9293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.983405113220215,
+      "learning_rate": 1.1507675438596492e-05,
+      "loss": 1.5144,
+      "step": 9294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.550469160079956,
+      "learning_rate": 1.1506761695906435e-05,
+      "loss": 1.4875,
+      "step": 9295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.842400550842285,
+      "learning_rate": 1.1505847953216375e-05,
+      "loss": 1.552,
+      "step": 9296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5395591259002686,
+      "learning_rate": 1.1504934210526317e-05,
+      "loss": 1.4499,
+      "step": 9297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1431596279144287,
+      "learning_rate": 1.150402046783626e-05,
+      "loss": 1.7415,
+      "step": 9298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.674745559692383,
+      "learning_rate": 1.15031067251462e-05,
+      "loss": 1.4879,
+      "step": 9299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6668777465820312,
+      "learning_rate": 1.1502192982456142e-05,
+      "loss": 1.5587,
+      "step": 9300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.011876106262207,
+      "learning_rate": 1.1501279239766083e-05,
+      "loss": 1.7839,
+      "step": 9301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.590364933013916,
+      "learning_rate": 1.1500365497076025e-05,
+      "loss": 1.7096,
+      "step": 9302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.193294048309326,
+      "learning_rate": 1.1499451754385966e-05,
+      "loss": 1.269,
+      "step": 9303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.498154640197754,
+      "learning_rate": 1.1498538011695908e-05,
+      "loss": 1.4331,
+      "step": 9304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.789665460586548,
+      "learning_rate": 1.1497624269005848e-05,
+      "loss": 1.6625,
+      "step": 9305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5863492488861084,
+      "learning_rate": 1.149671052631579e-05,
+      "loss": 1.6297,
+      "step": 9306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6568961143493652,
+      "learning_rate": 1.1495796783625733e-05,
+      "loss": 1.4002,
+      "step": 9307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7105979919433594,
+      "learning_rate": 1.1494883040935673e-05,
+      "loss": 1.5289,
+      "step": 9308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9016029834747314,
+      "learning_rate": 1.1493969298245616e-05,
+      "loss": 1.3683,
+      "step": 9309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.459077835083008,
+      "learning_rate": 1.1493055555555556e-05,
+      "loss": 1.4501,
+      "step": 9310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.088013172149658,
+      "learning_rate": 1.1492141812865498e-05,
+      "loss": 1.7416,
+      "step": 9311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5679805278778076,
+      "learning_rate": 1.1491228070175439e-05,
+      "loss": 1.5069,
+      "step": 9312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6231024265289307,
+      "learning_rate": 1.1490314327485381e-05,
+      "loss": 1.4892,
+      "step": 9313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.906074047088623,
+      "learning_rate": 1.1489400584795323e-05,
+      "loss": 1.6636,
+      "step": 9314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.235518455505371,
+      "learning_rate": 1.1488486842105264e-05,
+      "loss": 2.0815,
+      "step": 9315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8914480209350586,
+      "learning_rate": 1.1487573099415206e-05,
+      "loss": 1.7275,
+      "step": 9316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6670820713043213,
+      "learning_rate": 1.1486659356725147e-05,
+      "loss": 1.6492,
+      "step": 9317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9389712810516357,
+      "learning_rate": 1.1485745614035089e-05,
+      "loss": 1.8931,
+      "step": 9318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6613755226135254,
+      "learning_rate": 1.148483187134503e-05,
+      "loss": 1.2586,
+      "step": 9319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5776546001434326,
+      "learning_rate": 1.1483918128654972e-05,
+      "loss": 1.5442,
+      "step": 9320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1331679821014404,
+      "learning_rate": 1.1483004385964914e-05,
+      "loss": 1.5087,
+      "step": 9321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.178839683532715,
+      "learning_rate": 1.1482090643274854e-05,
+      "loss": 1.1479,
+      "step": 9322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9274837970733643,
+      "learning_rate": 1.1481176900584797e-05,
+      "loss": 1.9606,
+      "step": 9323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3794634342193604,
+      "learning_rate": 1.1480263157894737e-05,
+      "loss": 1.2447,
+      "step": 9324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3180301189422607,
+      "learning_rate": 1.147934941520468e-05,
+      "loss": 1.8215,
+      "step": 9325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.600334644317627,
+      "learning_rate": 1.147843567251462e-05,
+      "loss": 1.5214,
+      "step": 9326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3498942852020264,
+      "learning_rate": 1.1477521929824562e-05,
+      "loss": 1.4158,
+      "step": 9327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1977953910827637,
+      "learning_rate": 1.1476608187134504e-05,
+      "loss": 1.3496,
+      "step": 9328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5009942054748535,
+      "learning_rate": 1.1475694444444445e-05,
+      "loss": 1.3126,
+      "step": 9329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9467854499816895,
+      "learning_rate": 1.1474780701754387e-05,
+      "loss": 1.5756,
+      "step": 9330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4913976192474365,
+      "learning_rate": 1.1473866959064328e-05,
+      "loss": 1.4152,
+      "step": 9331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5913197994232178,
+      "learning_rate": 1.147295321637427e-05,
+      "loss": 2.0288,
+      "step": 9332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.958204507827759,
+      "learning_rate": 1.147203947368421e-05,
+      "loss": 1.7321,
+      "step": 9333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3090577125549316,
+      "learning_rate": 1.1471125730994153e-05,
+      "loss": 1.344,
+      "step": 9334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.932736396789551,
+      "learning_rate": 1.1470211988304095e-05,
+      "loss": 1.9192,
+      "step": 9335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.481438159942627,
+      "learning_rate": 1.1469298245614035e-05,
+      "loss": 1.3249,
+      "step": 9336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.314232110977173,
+      "learning_rate": 1.1468384502923977e-05,
+      "loss": 1.8931,
+      "step": 9337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.24916410446167,
+      "learning_rate": 1.1467470760233918e-05,
+      "loss": 1.1858,
+      "step": 9338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8267645835876465,
+      "learning_rate": 1.146655701754386e-05,
+      "loss": 1.0118,
+      "step": 9339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2329089641571045,
+      "learning_rate": 1.14656432748538e-05,
+      "loss": 1.8717,
+      "step": 9340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.798649787902832,
+      "learning_rate": 1.1464729532163743e-05,
+      "loss": 1.6648,
+      "step": 9341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.525657892227173,
+      "learning_rate": 1.1463815789473685e-05,
+      "loss": 1.4597,
+      "step": 9342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7496330738067627,
+      "learning_rate": 1.1462902046783626e-05,
+      "loss": 1.7053,
+      "step": 9343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.268096446990967,
+      "learning_rate": 1.1461988304093568e-05,
+      "loss": 1.5237,
+      "step": 9344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.239105701446533,
+      "learning_rate": 1.1461074561403508e-05,
+      "loss": 1.306,
+      "step": 9345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6649744510650635,
+      "learning_rate": 1.146016081871345e-05,
+      "loss": 1.506,
+      "step": 9346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1981687545776367,
+      "learning_rate": 1.1459247076023391e-05,
+      "loss": 2.0555,
+      "step": 9347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1049067974090576,
+      "learning_rate": 1.1458333333333333e-05,
+      "loss": 1.8568,
+      "step": 9348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6860859394073486,
+      "learning_rate": 1.1457419590643276e-05,
+      "loss": 1.6447,
+      "step": 9349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.419734477996826,
+      "learning_rate": 1.1456505847953216e-05,
+      "loss": 1.4273,
+      "step": 9350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5953099727630615,
+      "learning_rate": 1.1455592105263158e-05,
+      "loss": 1.7423,
+      "step": 9351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.604315757751465,
+      "learning_rate": 1.1454678362573099e-05,
+      "loss": 1.5095,
+      "step": 9352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.76505446434021,
+      "learning_rate": 1.1453764619883041e-05,
+      "loss": 1.5003,
+      "step": 9353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6602697372436523,
+      "learning_rate": 1.1452850877192982e-05,
+      "loss": 1.7373,
+      "step": 9354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.252549886703491,
+      "learning_rate": 1.1451937134502924e-05,
+      "loss": 1.5863,
+      "step": 9355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.367999792098999,
+      "learning_rate": 1.1451023391812866e-05,
+      "loss": 1.3984,
+      "step": 9356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.50091814994812,
+      "learning_rate": 1.1450109649122807e-05,
+      "loss": 1.5754,
+      "step": 9357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.284823417663574,
+      "learning_rate": 1.1449195906432749e-05,
+      "loss": 1.2852,
+      "step": 9358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9039769172668457,
+      "learning_rate": 1.144828216374269e-05,
+      "loss": 1.7263,
+      "step": 9359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.548137664794922,
+      "learning_rate": 1.1447368421052632e-05,
+      "loss": 1.2648,
+      "step": 9360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9825373888015747,
+      "learning_rate": 1.1446454678362572e-05,
+      "loss": 0.9703,
+      "step": 9361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.055208444595337,
+      "learning_rate": 1.1445540935672514e-05,
+      "loss": 1.9113,
+      "step": 9362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.612914800643921,
+      "learning_rate": 1.1444627192982458e-05,
+      "loss": 1.547,
+      "step": 9363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.940857887268066,
+      "learning_rate": 1.1443713450292399e-05,
+      "loss": 2.3135,
+      "step": 9364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.707573652267456,
+      "learning_rate": 1.1442799707602341e-05,
+      "loss": 1.4627,
+      "step": 9365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1845147609710693,
+      "learning_rate": 1.1441885964912283e-05,
+      "loss": 1.3742,
+      "step": 9366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7335128784179688,
+      "learning_rate": 1.1440972222222224e-05,
+      "loss": 1.3674,
+      "step": 9367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1873281002044678,
+      "learning_rate": 1.1440058479532166e-05,
+      "loss": 1.9162,
+      "step": 9368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5344464778900146,
+      "learning_rate": 1.1439144736842107e-05,
+      "loss": 1.5244,
+      "step": 9369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0108261108398438,
+      "learning_rate": 1.1438230994152049e-05,
+      "loss": 1.5999,
+      "step": 9370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2691266536712646,
+      "learning_rate": 1.143731725146199e-05,
+      "loss": 1.477,
+      "step": 9371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9313886165618896,
+      "learning_rate": 1.1436403508771932e-05,
+      "loss": 1.7585,
+      "step": 9372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8669633865356445,
+      "learning_rate": 1.1435489766081874e-05,
+      "loss": 1.4688,
+      "step": 9373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7328357696533203,
+      "learning_rate": 1.1434576023391814e-05,
+      "loss": 1.3522,
+      "step": 9374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.394609212875366,
+      "learning_rate": 1.1433662280701756e-05,
+      "loss": 1.5198,
+      "step": 9375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.028860569000244,
+      "learning_rate": 1.1432748538011697e-05,
+      "loss": 1.5344,
+      "step": 9376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.629276990890503,
+      "learning_rate": 1.143183479532164e-05,
+      "loss": 1.3556,
+      "step": 9377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9927494525909424,
+      "learning_rate": 1.143092105263158e-05,
+      "loss": 1.9536,
+      "step": 9378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6672985553741455,
+      "learning_rate": 1.1430007309941522e-05,
+      "loss": 1.7726,
+      "step": 9379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2701926231384277,
+      "learning_rate": 1.1429093567251464e-05,
+      "loss": 1.3471,
+      "step": 9380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.04905104637146,
+      "learning_rate": 1.1428179824561405e-05,
+      "loss": 1.777,
+      "step": 9381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7615113258361816,
+      "learning_rate": 1.1427266081871347e-05,
+      "loss": 1.9821,
+      "step": 9382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1545956134796143,
+      "learning_rate": 1.1426352339181287e-05,
+      "loss": 1.2641,
+      "step": 9383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1183059215545654,
+      "learning_rate": 1.142543859649123e-05,
+      "loss": 1.6376,
+      "step": 9384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5301384925842285,
+      "learning_rate": 1.142452485380117e-05,
+      "loss": 1.3807,
+      "step": 9385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7387917041778564,
+      "learning_rate": 1.1423611111111112e-05,
+      "loss": 1.632,
+      "step": 9386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6873130798339844,
+      "learning_rate": 1.1422697368421055e-05,
+      "loss": 1.3314,
+      "step": 9387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.359804391860962,
+      "learning_rate": 1.1421783625730995e-05,
+      "loss": 1.2698,
+      "step": 9388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.783313751220703,
+      "learning_rate": 1.1420869883040937e-05,
+      "loss": 1.6543,
+      "step": 9389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.736961603164673,
+      "learning_rate": 1.1419956140350878e-05,
+      "loss": 1.5953,
+      "step": 9390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7821803092956543,
+      "learning_rate": 1.141904239766082e-05,
+      "loss": 1.4071,
+      "step": 9391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5353755950927734,
+      "learning_rate": 1.141812865497076e-05,
+      "loss": 1.2399,
+      "step": 9392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.346102237701416,
+      "learning_rate": 1.1417214912280703e-05,
+      "loss": 1.463,
+      "step": 9393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.929447650909424,
+      "learning_rate": 1.1416301169590645e-05,
+      "loss": 1.7191,
+      "step": 9394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1136302947998047,
+      "learning_rate": 1.1415387426900586e-05,
+      "loss": 1.3231,
+      "step": 9395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.591324806213379,
+      "learning_rate": 1.1414473684210528e-05,
+      "loss": 1.4369,
+      "step": 9396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.438650131225586,
+      "learning_rate": 1.1413559941520468e-05,
+      "loss": 1.3478,
+      "step": 9397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.838744878768921,
+      "learning_rate": 1.141264619883041e-05,
+      "loss": 1.8042,
+      "step": 9398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 277.0063171386719,
+      "learning_rate": 1.1411732456140351e-05,
+      "loss": 1.6138,
+      "step": 9399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.836117744445801,
+      "learning_rate": 1.1410818713450293e-05,
+      "loss": 1.906,
+      "step": 9400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6181769371032715,
+      "learning_rate": 1.1409904970760236e-05,
+      "loss": 1.4417,
+      "step": 9401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.064582109451294,
+      "learning_rate": 1.1408991228070176e-05,
+      "loss": 1.5477,
+      "step": 9402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2859368324279785,
+      "learning_rate": 1.1408077485380118e-05,
+      "loss": 1.2364,
+      "step": 9403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2901723384857178,
+      "learning_rate": 1.1407163742690059e-05,
+      "loss": 1.4479,
+      "step": 9404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8220674991607666,
+      "learning_rate": 1.1406250000000001e-05,
+      "loss": 1.4456,
+      "step": 9405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.697289228439331,
+      "learning_rate": 1.1405336257309942e-05,
+      "loss": 1.6181,
+      "step": 9406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.801027774810791,
+      "learning_rate": 1.1404422514619884e-05,
+      "loss": 1.5865,
+      "step": 9407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.163809299468994,
+      "learning_rate": 1.1403508771929826e-05,
+      "loss": 1.7613,
+      "step": 9408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.228818416595459,
+      "learning_rate": 1.1402595029239767e-05,
+      "loss": 1.6399,
+      "step": 9409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6032114028930664,
+      "learning_rate": 1.1401681286549709e-05,
+      "loss": 1.6855,
+      "step": 9410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.48960280418396,
+      "learning_rate": 1.140076754385965e-05,
+      "loss": 1.612,
+      "step": 9411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0580217838287354,
+      "learning_rate": 1.1399853801169592e-05,
+      "loss": 1.1657,
+      "step": 9412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.103161573410034,
+      "learning_rate": 1.1398940058479532e-05,
+      "loss": 1.3678,
+      "step": 9413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.030747413635254,
+      "learning_rate": 1.1398026315789474e-05,
+      "loss": 1.2257,
+      "step": 9414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3045847415924072,
+      "learning_rate": 1.1397112573099417e-05,
+      "loss": 1.1865,
+      "step": 9415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.880384683609009,
+      "learning_rate": 1.1396198830409357e-05,
+      "loss": 1.8776,
+      "step": 9416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8192853927612305,
+      "learning_rate": 1.13952850877193e-05,
+      "loss": 1.6223,
+      "step": 9417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7396559715270996,
+      "learning_rate": 1.139437134502924e-05,
+      "loss": 1.6368,
+      "step": 9418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8256819248199463,
+      "learning_rate": 1.1393457602339182e-05,
+      "loss": 1.653,
+      "step": 9419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.785385847091675,
+      "learning_rate": 1.1392543859649123e-05,
+      "loss": 1.7219,
+      "step": 9420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.648909330368042,
+      "learning_rate": 1.1391630116959065e-05,
+      "loss": 1.3698,
+      "step": 9421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.06107497215271,
+      "learning_rate": 1.1390716374269007e-05,
+      "loss": 1.3205,
+      "step": 9422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.166029453277588,
+      "learning_rate": 1.1389802631578948e-05,
+      "loss": 1.8816,
+      "step": 9423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.733483076095581,
+      "learning_rate": 1.138888888888889e-05,
+      "loss": 1.2878,
+      "step": 9424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.887002944946289,
+      "learning_rate": 1.138797514619883e-05,
+      "loss": 1.8142,
+      "step": 9425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1104979515075684,
+      "learning_rate": 1.1387061403508773e-05,
+      "loss": 1.5253,
+      "step": 9426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.484961986541748,
+      "learning_rate": 1.1386147660818713e-05,
+      "loss": 1.2327,
+      "step": 9427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7798328399658203,
+      "learning_rate": 1.1385233918128655e-05,
+      "loss": 1.4409,
+      "step": 9428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1129024028778076,
+      "learning_rate": 1.1384320175438597e-05,
+      "loss": 1.5438,
+      "step": 9429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3570199012756348,
+      "learning_rate": 1.1383406432748538e-05,
+      "loss": 1.3778,
+      "step": 9430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0855536460876465,
+      "learning_rate": 1.138249269005848e-05,
+      "loss": 1.8058,
+      "step": 9431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5934717655181885,
+      "learning_rate": 1.138157894736842e-05,
+      "loss": 1.5173,
+      "step": 9432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.798175573348999,
+      "learning_rate": 1.1380665204678363e-05,
+      "loss": 1.701,
+      "step": 9433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3630423545837402,
+      "learning_rate": 1.1379751461988304e-05,
+      "loss": 1.6073,
+      "step": 9434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.511634111404419,
+      "learning_rate": 1.1378837719298246e-05,
+      "loss": 1.3794,
+      "step": 9435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.782404899597168,
+      "learning_rate": 1.1377923976608188e-05,
+      "loss": 1.7933,
+      "step": 9436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5214052200317383,
+      "learning_rate": 1.1377010233918128e-05,
+      "loss": 1.2896,
+      "step": 9437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.869969367980957,
+      "learning_rate": 1.137609649122807e-05,
+      "loss": 1.8051,
+      "step": 9438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9343955516815186,
+      "learning_rate": 1.1375182748538011e-05,
+      "loss": 1.3475,
+      "step": 9439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.999945640563965,
+      "learning_rate": 1.1374269005847953e-05,
+      "loss": 1.3547,
+      "step": 9440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4516913890838623,
+      "learning_rate": 1.1373355263157894e-05,
+      "loss": 1.8276,
+      "step": 9441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.676626682281494,
+      "learning_rate": 1.1372441520467836e-05,
+      "loss": 1.4537,
+      "step": 9442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5887742042541504,
+      "learning_rate": 1.1371527777777778e-05,
+      "loss": 1.3188,
+      "step": 9443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5469741821289062,
+      "learning_rate": 1.1370614035087719e-05,
+      "loss": 1.2711,
+      "step": 9444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9889750480651855,
+      "learning_rate": 1.1369700292397661e-05,
+      "loss": 1.7234,
+      "step": 9445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6735708713531494,
+      "learning_rate": 1.1368786549707602e-05,
+      "loss": 1.542,
+      "step": 9446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1999242305755615,
+      "learning_rate": 1.1367872807017544e-05,
+      "loss": 1.8671,
+      "step": 9447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6748344898223877,
+      "learning_rate": 1.1366959064327488e-05,
+      "loss": 1.2111,
+      "step": 9448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.505721092224121,
+      "learning_rate": 1.1366045321637428e-05,
+      "loss": 1.5814,
+      "step": 9449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.452677011489868,
+      "learning_rate": 1.136513157894737e-05,
+      "loss": 1.1883,
+      "step": 9450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4930639266967773,
+      "learning_rate": 1.1364217836257311e-05,
+      "loss": 1.6223,
+      "step": 9451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.188244342803955,
+      "learning_rate": 1.1363304093567253e-05,
+      "loss": 1.8769,
+      "step": 9452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6835885047912598,
+      "learning_rate": 1.1362390350877196e-05,
+      "loss": 1.4309,
+      "step": 9453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.337261199951172,
+      "learning_rate": 1.1361476608187136e-05,
+      "loss": 1.3558,
+      "step": 9454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.54168438911438,
+      "learning_rate": 1.1360562865497078e-05,
+      "loss": 1.645,
+      "step": 9455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2598631381988525,
+      "learning_rate": 1.1359649122807019e-05,
+      "loss": 1.6639,
+      "step": 9456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.891138792037964,
+      "learning_rate": 1.1358735380116961e-05,
+      "loss": 1.5439,
+      "step": 9457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.921473264694214,
+      "learning_rate": 1.1357821637426902e-05,
+      "loss": 1.5455,
+      "step": 9458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4052274227142334,
+      "learning_rate": 1.1356907894736844e-05,
+      "loss": 1.3777,
+      "step": 9459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.349055290222168,
+      "learning_rate": 1.1355994152046786e-05,
+      "loss": 1.1737,
+      "step": 9460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.537802219390869,
+      "learning_rate": 1.1355080409356727e-05,
+      "loss": 1.3842,
+      "step": 9461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3849539756774902,
+      "learning_rate": 1.1354166666666669e-05,
+      "loss": 1.1251,
+      "step": 9462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2942733764648438,
+      "learning_rate": 1.135325292397661e-05,
+      "loss": 1.3486,
+      "step": 9463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2288312911987305,
+      "learning_rate": 1.1352339181286552e-05,
+      "loss": 1.2648,
+      "step": 9464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5018310546875,
+      "learning_rate": 1.1351425438596492e-05,
+      "loss": 1.2968,
+      "step": 9465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5957980155944824,
+      "learning_rate": 1.1350511695906434e-05,
+      "loss": 1.4107,
+      "step": 9466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7890493869781494,
+      "learning_rate": 1.1349597953216376e-05,
+      "loss": 1.378,
+      "step": 9467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0185604095458984,
+      "learning_rate": 1.1348684210526317e-05,
+      "loss": 1.8448,
+      "step": 9468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.076341390609741,
+      "learning_rate": 1.134777046783626e-05,
+      "loss": 1.8427,
+      "step": 9469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7745654582977295,
+      "learning_rate": 1.13468567251462e-05,
+      "loss": 1.4101,
+      "step": 9470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9000368118286133,
+      "learning_rate": 1.1345942982456142e-05,
+      "loss": 1.7799,
+      "step": 9471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.949589252471924,
+      "learning_rate": 1.1345029239766083e-05,
+      "loss": 1.6247,
+      "step": 9472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0993707180023193,
+      "learning_rate": 1.1344115497076025e-05,
+      "loss": 1.9003,
+      "step": 9473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.000727653503418,
+      "learning_rate": 1.1343201754385967e-05,
+      "loss": 1.4289,
+      "step": 9474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4958012104034424,
+      "learning_rate": 1.1342288011695907e-05,
+      "loss": 1.5568,
+      "step": 9475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.397010326385498,
+      "learning_rate": 1.134137426900585e-05,
+      "loss": 1.4219,
+      "step": 9476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.605595350265503,
+      "learning_rate": 1.134046052631579e-05,
+      "loss": 1.3305,
+      "step": 9477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3904948234558105,
+      "learning_rate": 1.1339546783625732e-05,
+      "loss": 1.2365,
+      "step": 9478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.386197805404663,
+      "learning_rate": 1.1338633040935673e-05,
+      "loss": 1.6272,
+      "step": 9479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.146627187728882,
+      "learning_rate": 1.1337719298245615e-05,
+      "loss": 1.3678,
+      "step": 9480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.444962978363037,
+      "learning_rate": 1.1336805555555557e-05,
+      "loss": 1.4816,
+      "step": 9481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5885143280029297,
+      "learning_rate": 1.1335891812865498e-05,
+      "loss": 1.4673,
+      "step": 9482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9012255668640137,
+      "learning_rate": 1.133497807017544e-05,
+      "loss": 1.5653,
+      "step": 9483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0564959049224854,
+      "learning_rate": 1.133406432748538e-05,
+      "loss": 1.2678,
+      "step": 9484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.286308526992798,
+      "learning_rate": 1.1333150584795323e-05,
+      "loss": 1.2097,
+      "step": 9485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.614325761795044,
+      "learning_rate": 1.1332236842105263e-05,
+      "loss": 1.5662,
+      "step": 9486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5465140342712402,
+      "learning_rate": 1.1331323099415206e-05,
+      "loss": 1.8251,
+      "step": 9487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.751556873321533,
+      "learning_rate": 1.1330409356725148e-05,
+      "loss": 1.892,
+      "step": 9488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4038872718811035,
+      "learning_rate": 1.1329495614035088e-05,
+      "loss": 1.2803,
+      "step": 9489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.259783983230591,
+      "learning_rate": 1.132858187134503e-05,
+      "loss": 1.0236,
+      "step": 9490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.701564311981201,
+      "learning_rate": 1.1327668128654971e-05,
+      "loss": 1.5715,
+      "step": 9491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.502913475036621,
+      "learning_rate": 1.1326754385964913e-05,
+      "loss": 1.7144,
+      "step": 9492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.566331624984741,
+      "learning_rate": 1.1325840643274854e-05,
+      "loss": 1.5033,
+      "step": 9493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9861528873443604,
+      "learning_rate": 1.1324926900584796e-05,
+      "loss": 1.5667,
+      "step": 9494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5463621616363525,
+      "learning_rate": 1.1324013157894738e-05,
+      "loss": 1.4732,
+      "step": 9495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.164853096008301,
+      "learning_rate": 1.1323099415204679e-05,
+      "loss": 1.8709,
+      "step": 9496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.581226110458374,
+      "learning_rate": 1.1322185672514621e-05,
+      "loss": 1.8418,
+      "step": 9497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0216031074523926,
+      "learning_rate": 1.1321271929824562e-05,
+      "loss": 1.7318,
+      "step": 9498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1154401302337646,
+      "learning_rate": 1.1320358187134504e-05,
+      "loss": 1.7075,
+      "step": 9499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.70473575592041,
+      "learning_rate": 1.1319444444444444e-05,
+      "loss": 1.711,
+      "step": 9500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7820522785186768,
+      "learning_rate": 1.1318530701754387e-05,
+      "loss": 1.3039,
+      "step": 9501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.477299928665161,
+      "learning_rate": 1.1317616959064329e-05,
+      "loss": 1.3566,
+      "step": 9502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8051090240478516,
+      "learning_rate": 1.131670321637427e-05,
+      "loss": 1.5621,
+      "step": 9503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.656787157058716,
+      "learning_rate": 1.1315789473684212e-05,
+      "loss": 1.4862,
+      "step": 9504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5611953735351562,
+      "learning_rate": 1.1314875730994152e-05,
+      "loss": 1.2978,
+      "step": 9505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.311537027359009,
+      "learning_rate": 1.1313961988304094e-05,
+      "loss": 1.5564,
+      "step": 9506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.314016819000244,
+      "learning_rate": 1.1313048245614035e-05,
+      "loss": 1.4307,
+      "step": 9507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 62.84480285644531,
+      "learning_rate": 1.1312134502923977e-05,
+      "loss": 1.6586,
+      "step": 9508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.459941864013672,
+      "learning_rate": 1.131122076023392e-05,
+      "loss": 1.3689,
+      "step": 9509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8605289459228516,
+      "learning_rate": 1.131030701754386e-05,
+      "loss": 1.6923,
+      "step": 9510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.091702461242676,
+      "learning_rate": 1.1309393274853802e-05,
+      "loss": 1.5729,
+      "step": 9511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.90584397315979,
+      "learning_rate": 1.1308479532163743e-05,
+      "loss": 1.6669,
+      "step": 9512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.107379198074341,
+      "learning_rate": 1.1307565789473685e-05,
+      "loss": 1.553,
+      "step": 9513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7746522426605225,
+      "learning_rate": 1.1306652046783625e-05,
+      "loss": 1.6172,
+      "step": 9514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8588647842407227,
+      "learning_rate": 1.1305738304093568e-05,
+      "loss": 1.5245,
+      "step": 9515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7836265563964844,
+      "learning_rate": 1.130482456140351e-05,
+      "loss": 2.0245,
+      "step": 9516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.457847833633423,
+      "learning_rate": 1.130391081871345e-05,
+      "loss": 1.5338,
+      "step": 9517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.696758270263672,
+      "learning_rate": 1.1302997076023393e-05,
+      "loss": 1.4518,
+      "step": 9518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9986518621444702,
+      "learning_rate": 1.1302083333333333e-05,
+      "loss": 0.9862,
+      "step": 9519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2237002849578857,
+      "learning_rate": 1.1301169590643275e-05,
+      "loss": 1.2214,
+      "step": 9520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7078070640563965,
+      "learning_rate": 1.1300255847953216e-05,
+      "loss": 1.4789,
+      "step": 9521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.686258554458618,
+      "learning_rate": 1.1299342105263158e-05,
+      "loss": 1.692,
+      "step": 9522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.778909683227539,
+      "learning_rate": 1.12984283625731e-05,
+      "loss": 1.6508,
+      "step": 9523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.449007034301758,
+      "learning_rate": 1.129751461988304e-05,
+      "loss": 1.7333,
+      "step": 9524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8293800354003906,
+      "learning_rate": 1.1296600877192983e-05,
+      "loss": 1.7836,
+      "step": 9525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.438492774963379,
+      "learning_rate": 1.1295687134502924e-05,
+      "loss": 1.3832,
+      "step": 9526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9946489334106445,
+      "learning_rate": 1.1294773391812866e-05,
+      "loss": 1.5679,
+      "step": 9527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4153411388397217,
+      "learning_rate": 1.1293859649122806e-05,
+      "loss": 1.5838,
+      "step": 9528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.902174949645996,
+      "learning_rate": 1.1292945906432748e-05,
+      "loss": 1.704,
+      "step": 9529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8977410793304443,
+      "learning_rate": 1.129203216374269e-05,
+      "loss": 1.5027,
+      "step": 9530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.317422866821289,
+      "learning_rate": 1.1291118421052631e-05,
+      "loss": 1.6211,
+      "step": 9531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.861469268798828,
+      "learning_rate": 1.1290204678362573e-05,
+      "loss": 1.5932,
+      "step": 9532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.056915521621704,
+      "learning_rate": 1.1289290935672514e-05,
+      "loss": 1.9601,
+      "step": 9533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8292183876037598,
+      "learning_rate": 1.1288377192982458e-05,
+      "loss": 1.7563,
+      "step": 9534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.711709499359131,
+      "learning_rate": 1.12874634502924e-05,
+      "loss": 1.5496,
+      "step": 9535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.765131950378418,
+      "learning_rate": 1.128654970760234e-05,
+      "loss": 1.4679,
+      "step": 9536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4012997150421143,
+      "learning_rate": 1.1285635964912283e-05,
+      "loss": 1.5863,
+      "step": 9537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.791614055633545,
+      "learning_rate": 1.1284722222222223e-05,
+      "loss": 1.7038,
+      "step": 9538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8066534996032715,
+      "learning_rate": 1.1283808479532166e-05,
+      "loss": 1.0665,
+      "step": 9539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0944786071777344,
+      "learning_rate": 1.1282894736842108e-05,
+      "loss": 1.1646,
+      "step": 9540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.331721305847168,
+      "learning_rate": 1.1281980994152048e-05,
+      "loss": 1.392,
+      "step": 9541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6636927127838135,
+      "learning_rate": 1.128106725146199e-05,
+      "loss": 1.5258,
+      "step": 9542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0774948596954346,
+      "learning_rate": 1.1280153508771931e-05,
+      "loss": 1.8049,
+      "step": 9543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9485390186309814,
+      "learning_rate": 1.1279239766081873e-05,
+      "loss": 1.5407,
+      "step": 9544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7435715198516846,
+      "learning_rate": 1.1278326023391814e-05,
+      "loss": 1.729,
+      "step": 9545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.823122978210449,
+      "learning_rate": 1.1277412280701756e-05,
+      "loss": 1.5084,
+      "step": 9546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6931416988372803,
+      "learning_rate": 1.1276498538011698e-05,
+      "loss": 1.4874,
+      "step": 9547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.91941499710083,
+      "learning_rate": 1.1275584795321639e-05,
+      "loss": 1.5473,
+      "step": 9548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.806882619857788,
+      "learning_rate": 1.1274671052631581e-05,
+      "loss": 1.4403,
+      "step": 9549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.669208288192749,
+      "learning_rate": 1.1273757309941522e-05,
+      "loss": 1.3679,
+      "step": 9550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.465696096420288,
+      "learning_rate": 1.1272843567251464e-05,
+      "loss": 1.4296,
+      "step": 9551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4980645179748535,
+      "learning_rate": 1.1271929824561404e-05,
+      "loss": 1.5371,
+      "step": 9552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.428622007369995,
+      "learning_rate": 1.1271016081871347e-05,
+      "loss": 1.2504,
+      "step": 9553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4135727882385254,
+      "learning_rate": 1.1270102339181289e-05,
+      "loss": 1.4322,
+      "step": 9554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7027535438537598,
+      "learning_rate": 1.126918859649123e-05,
+      "loss": 1.8829,
+      "step": 9555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.651057243347168,
+      "learning_rate": 1.1268274853801172e-05,
+      "loss": 1.6707,
+      "step": 9556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.613319158554077,
+      "learning_rate": 1.1267361111111112e-05,
+      "loss": 1.499,
+      "step": 9557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.530015468597412,
+      "learning_rate": 1.1266447368421054e-05,
+      "loss": 1.6645,
+      "step": 9558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9162867069244385,
+      "learning_rate": 1.1265533625730995e-05,
+      "loss": 1.7785,
+      "step": 9559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.741621494293213,
+      "learning_rate": 1.1264619883040937e-05,
+      "loss": 1.3523,
+      "step": 9560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.115745782852173,
+      "learning_rate": 1.126370614035088e-05,
+      "loss": 1.3405,
+      "step": 9561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.369626522064209,
+      "learning_rate": 1.126279239766082e-05,
+      "loss": 2.0811,
+      "step": 9562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.371509552001953,
+      "learning_rate": 1.1261878654970762e-05,
+      "loss": 1.2904,
+      "step": 9563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.839387893676758,
+      "learning_rate": 1.1260964912280703e-05,
+      "loss": 1.6444,
+      "step": 9564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.469205856323242,
+      "learning_rate": 1.1260051169590645e-05,
+      "loss": 1.5896,
+      "step": 9565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0982296466827393,
+      "learning_rate": 1.1259137426900585e-05,
+      "loss": 1.4887,
+      "step": 9566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7981302738189697,
+      "learning_rate": 1.1258223684210527e-05,
+      "loss": 1.7613,
+      "step": 9567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.045440196990967,
+      "learning_rate": 1.125730994152047e-05,
+      "loss": 1.1105,
+      "step": 9568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.816499948501587,
+      "learning_rate": 1.125639619883041e-05,
+      "loss": 1.5296,
+      "step": 9569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.824330806732178,
+      "learning_rate": 1.1255482456140352e-05,
+      "loss": 1.6181,
+      "step": 9570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1749651432037354,
+      "learning_rate": 1.1254568713450293e-05,
+      "loss": 1.9604,
+      "step": 9571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4617464542388916,
+      "learning_rate": 1.1253654970760235e-05,
+      "loss": 1.4589,
+      "step": 9572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4059479236602783,
+      "learning_rate": 1.1252741228070176e-05,
+      "loss": 1.4867,
+      "step": 9573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.421092987060547,
+      "learning_rate": 1.1251827485380118e-05,
+      "loss": 1.2113,
+      "step": 9574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.217577934265137,
+      "learning_rate": 1.125091374269006e-05,
+      "loss": 2.0458,
+      "step": 9575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8476626873016357,
+      "learning_rate": 1.125e-05,
+      "loss": 1.7919,
+      "step": 9576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.996443271636963,
+      "learning_rate": 1.1249086257309943e-05,
+      "loss": 1.8682,
+      "step": 9577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.381464719772339,
+      "learning_rate": 1.1248172514619883e-05,
+      "loss": 1.264,
+      "step": 9578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.473670482635498,
+      "learning_rate": 1.1247258771929826e-05,
+      "loss": 1.4675,
+      "step": 9579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.817103624343872,
+      "learning_rate": 1.1246345029239766e-05,
+      "loss": 1.5337,
+      "step": 9580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.257473945617676,
+      "learning_rate": 1.1245431286549708e-05,
+      "loss": 1.7461,
+      "step": 9581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3314759731292725,
+      "learning_rate": 1.124451754385965e-05,
+      "loss": 1.6036,
+      "step": 9582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.974040985107422,
+      "learning_rate": 1.1243603801169591e-05,
+      "loss": 1.5789,
+      "step": 9583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.532538890838623,
+      "learning_rate": 1.1242690058479533e-05,
+      "loss": 1.9855,
+      "step": 9584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.994760274887085,
+      "learning_rate": 1.1241776315789474e-05,
+      "loss": 1.6031,
+      "step": 9585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3355813026428223,
+      "learning_rate": 1.1240862573099416e-05,
+      "loss": 1.2789,
+      "step": 9586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7274930477142334,
+      "learning_rate": 1.1239948830409357e-05,
+      "loss": 1.4019,
+      "step": 9587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0164167881011963,
+      "learning_rate": 1.1239035087719299e-05,
+      "loss": 1.5268,
+      "step": 9588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7461886405944824,
+      "learning_rate": 1.1238121345029241e-05,
+      "loss": 1.4803,
+      "step": 9589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5858101844787598,
+      "learning_rate": 1.1237207602339182e-05,
+      "loss": 1.7044,
+      "step": 9590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.785740613937378,
+      "learning_rate": 1.1236293859649124e-05,
+      "loss": 1.5013,
+      "step": 9591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9523956775665283,
+      "learning_rate": 1.1235380116959064e-05,
+      "loss": 1.8335,
+      "step": 9592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.065568685531616,
+      "learning_rate": 1.1234466374269007e-05,
+      "loss": 1.6043,
+      "step": 9593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4677538871765137,
+      "learning_rate": 1.1233552631578947e-05,
+      "loss": 1.207,
+      "step": 9594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.305384874343872,
+      "learning_rate": 1.123263888888889e-05,
+      "loss": 1.4748,
+      "step": 9595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.01324725151062,
+      "learning_rate": 1.1231725146198832e-05,
+      "loss": 2.1836,
+      "step": 9596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.136692523956299,
+      "learning_rate": 1.1230811403508772e-05,
+      "loss": 1.2193,
+      "step": 9597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.297374963760376,
+      "learning_rate": 1.1229897660818714e-05,
+      "loss": 1.3032,
+      "step": 9598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.615563154220581,
+      "learning_rate": 1.1228983918128655e-05,
+      "loss": 1.5511,
+      "step": 9599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3934662342071533,
+      "learning_rate": 1.1228070175438597e-05,
+      "loss": 1.1746,
+      "step": 9600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2089719772338867,
+      "learning_rate": 1.1227156432748538e-05,
+      "loss": 1.713,
+      "step": 9601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.830095052719116,
+      "learning_rate": 1.122624269005848e-05,
+      "loss": 1.4541,
+      "step": 9602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6920981407165527,
+      "learning_rate": 1.1225328947368422e-05,
+      "loss": 1.7476,
+      "step": 9603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.595305919647217,
+      "learning_rate": 1.1224415204678363e-05,
+      "loss": 1.522,
+      "step": 9604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.145911693572998,
+      "learning_rate": 1.1223501461988305e-05,
+      "loss": 1.4502,
+      "step": 9605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6589314937591553,
+      "learning_rate": 1.1222587719298245e-05,
+      "loss": 1.6672,
+      "step": 9606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.501877546310425,
+      "learning_rate": 1.1221673976608188e-05,
+      "loss": 1.4427,
+      "step": 9607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.288607120513916,
+      "learning_rate": 1.1220760233918128e-05,
+      "loss": 1.5953,
+      "step": 9608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.818981885910034,
+      "learning_rate": 1.121984649122807e-05,
+      "loss": 1.4111,
+      "step": 9609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.675051212310791,
+      "learning_rate": 1.1218932748538013e-05,
+      "loss": 1.7579,
+      "step": 9610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8557851314544678,
+      "learning_rate": 1.1218019005847953e-05,
+      "loss": 1.3532,
+      "step": 9611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.704833984375,
+      "learning_rate": 1.1217105263157895e-05,
+      "loss": 1.4354,
+      "step": 9612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.707841634750366,
+      "learning_rate": 1.1216191520467836e-05,
+      "loss": 1.8786,
+      "step": 9613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.688852548599243,
+      "learning_rate": 1.1215277777777778e-05,
+      "loss": 1.5713,
+      "step": 9614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.865304946899414,
+      "learning_rate": 1.1214364035087719e-05,
+      "loss": 1.3684,
+      "step": 9615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8905224800109863,
+      "learning_rate": 1.121345029239766e-05,
+      "loss": 2.0448,
+      "step": 9616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7298009395599365,
+      "learning_rate": 1.1212536549707603e-05,
+      "loss": 1.6875,
+      "step": 9617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.690854072570801,
+      "learning_rate": 1.1211622807017544e-05,
+      "loss": 1.5848,
+      "step": 9618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.639108180999756,
+      "learning_rate": 1.1210709064327487e-05,
+      "loss": 1.7021,
+      "step": 9619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3659911155700684,
+      "learning_rate": 1.120979532163743e-05,
+      "loss": 1.3167,
+      "step": 9620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.293260097503662,
+      "learning_rate": 1.120888157894737e-05,
+      "loss": 1.3258,
+      "step": 9621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.540086030960083,
+      "learning_rate": 1.1207967836257312e-05,
+      "loss": 1.2704,
+      "step": 9622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3385069370269775,
+      "learning_rate": 1.1207054093567253e-05,
+      "loss": 1.1684,
+      "step": 9623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.508188247680664,
+      "learning_rate": 1.1206140350877195e-05,
+      "loss": 1.2412,
+      "step": 9624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.635985851287842,
+      "learning_rate": 1.1205226608187136e-05,
+      "loss": 1.3477,
+      "step": 9625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.315077304840088,
+      "learning_rate": 1.1204312865497078e-05,
+      "loss": 1.3713,
+      "step": 9626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.272148609161377,
+      "learning_rate": 1.120339912280702e-05,
+      "loss": 1.4174,
+      "step": 9627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8213183879852295,
+      "learning_rate": 1.120248538011696e-05,
+      "loss": 1.6981,
+      "step": 9628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.516660213470459,
+      "learning_rate": 1.1201571637426903e-05,
+      "loss": 1.6709,
+      "step": 9629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.387101411819458,
+      "learning_rate": 1.1200657894736843e-05,
+      "loss": 1.4883,
+      "step": 9630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.438143014907837,
+      "learning_rate": 1.1199744152046786e-05,
+      "loss": 1.2288,
+      "step": 9631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2559621334075928,
+      "learning_rate": 1.1198830409356726e-05,
+      "loss": 1.2539,
+      "step": 9632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.441545009613037,
+      "learning_rate": 1.1197916666666668e-05,
+      "loss": 1.5715,
+      "step": 9633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.35353684425354,
+      "learning_rate": 1.119700292397661e-05,
+      "loss": 1.244,
+      "step": 9634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5711495876312256,
+      "learning_rate": 1.1196089181286551e-05,
+      "loss": 1.1928,
+      "step": 9635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.286475896835327,
+      "learning_rate": 1.1195175438596493e-05,
+      "loss": 1.3132,
+      "step": 9636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8986105918884277,
+      "learning_rate": 1.1194261695906434e-05,
+      "loss": 1.6764,
+      "step": 9637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.053532123565674,
+      "learning_rate": 1.1193347953216376e-05,
+      "loss": 1.2896,
+      "step": 9638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.822110891342163,
+      "learning_rate": 1.1192434210526317e-05,
+      "loss": 1.7077,
+      "step": 9639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.274567127227783,
+      "learning_rate": 1.1191520467836259e-05,
+      "loss": 1.2793,
+      "step": 9640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.960721969604492,
+      "learning_rate": 1.1190606725146201e-05,
+      "loss": 1.3849,
+      "step": 9641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.670614242553711,
+      "learning_rate": 1.1189692982456142e-05,
+      "loss": 1.6665,
+      "step": 9642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5461888313293457,
+      "learning_rate": 1.1188779239766084e-05,
+      "loss": 1.6114,
+      "step": 9643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.225525379180908,
+      "learning_rate": 1.1187865497076024e-05,
+      "loss": 1.8357,
+      "step": 9644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.487752914428711,
+      "learning_rate": 1.1186951754385967e-05,
+      "loss": 1.0831,
+      "step": 9645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5583035945892334,
+      "learning_rate": 1.1186038011695907e-05,
+      "loss": 1.6803,
+      "step": 9646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.58981990814209,
+      "learning_rate": 1.118512426900585e-05,
+      "loss": 1.4278,
+      "step": 9647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4362149238586426,
+      "learning_rate": 1.1184210526315792e-05,
+      "loss": 1.2827,
+      "step": 9648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9210283756256104,
+      "learning_rate": 1.1183296783625732e-05,
+      "loss": 2.3428,
+      "step": 9649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5778822898864746,
+      "learning_rate": 1.1182383040935674e-05,
+      "loss": 1.5933,
+      "step": 9650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4583897590637207,
+      "learning_rate": 1.1181469298245615e-05,
+      "loss": 1.2168,
+      "step": 9651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3741400241851807,
+      "learning_rate": 1.1180555555555557e-05,
+      "loss": 1.0302,
+      "step": 9652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.494946002960205,
+      "learning_rate": 1.1179641812865498e-05,
+      "loss": 1.3409,
+      "step": 9653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1018338203430176,
+      "learning_rate": 1.117872807017544e-05,
+      "loss": 1.7285,
+      "step": 9654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6677026748657227,
+      "learning_rate": 1.1177814327485382e-05,
+      "loss": 1.1662,
+      "step": 9655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.60851788520813,
+      "learning_rate": 1.1176900584795323e-05,
+      "loss": 1.5093,
+      "step": 9656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.07255482673645,
+      "learning_rate": 1.1175986842105265e-05,
+      "loss": 1.0814,
+      "step": 9657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6939663887023926,
+      "learning_rate": 1.1175073099415205e-05,
+      "loss": 1.4638,
+      "step": 9658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.829153060913086,
+      "learning_rate": 1.1174159356725147e-05,
+      "loss": 1.6379,
+      "step": 9659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0936360359191895,
+      "learning_rate": 1.1173245614035088e-05,
+      "loss": 1.7742,
+      "step": 9660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3606040477752686,
+      "learning_rate": 1.117233187134503e-05,
+      "loss": 1.1628,
+      "step": 9661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9548981189727783,
+      "learning_rate": 1.1171418128654972e-05,
+      "loss": 1.3888,
+      "step": 9662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6704342365264893,
+      "learning_rate": 1.1170504385964913e-05,
+      "loss": 1.7269,
+      "step": 9663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6013598442077637,
+      "learning_rate": 1.1169590643274855e-05,
+      "loss": 1.463,
+      "step": 9664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3295106887817383,
+      "learning_rate": 1.1168676900584796e-05,
+      "loss": 1.4732,
+      "step": 9665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8548622131347656,
+      "learning_rate": 1.1167763157894738e-05,
+      "loss": 1.1824,
+      "step": 9666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1566433906555176,
+      "learning_rate": 1.1166849415204678e-05,
+      "loss": 1.4068,
+      "step": 9667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.633870840072632,
+      "learning_rate": 1.116593567251462e-05,
+      "loss": 1.5229,
+      "step": 9668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.082521677017212,
+      "learning_rate": 1.1165021929824563e-05,
+      "loss": 1.4725,
+      "step": 9669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.800905704498291,
+      "learning_rate": 1.1164108187134503e-05,
+      "loss": 1.2698,
+      "step": 9670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2560813426971436,
+      "learning_rate": 1.1163194444444446e-05,
+      "loss": 1.438,
+      "step": 9671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2166380882263184,
+      "learning_rate": 1.1162280701754386e-05,
+      "loss": 1.1794,
+      "step": 9672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.587149143218994,
+      "learning_rate": 1.1161366959064328e-05,
+      "loss": 1.6091,
+      "step": 9673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6229114532470703,
+      "learning_rate": 1.1160453216374269e-05,
+      "loss": 1.4596,
+      "step": 9674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6657485961914062,
+      "learning_rate": 1.1159539473684211e-05,
+      "loss": 1.6031,
+      "step": 9675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0176122188568115,
+      "learning_rate": 1.1158625730994153e-05,
+      "loss": 1.4833,
+      "step": 9676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9169812202453613,
+      "learning_rate": 1.1157711988304094e-05,
+      "loss": 1.6392,
+      "step": 9677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1992006301879883,
+      "learning_rate": 1.1156798245614036e-05,
+      "loss": 1.5832,
+      "step": 9678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0086376667022705,
+      "learning_rate": 1.1155884502923977e-05,
+      "loss": 1.8429,
+      "step": 9679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0063133239746094,
+      "learning_rate": 1.1154970760233919e-05,
+      "loss": 1.5727,
+      "step": 9680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5841166973114014,
+      "learning_rate": 1.115405701754386e-05,
+      "loss": 1.3502,
+      "step": 9681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9755983352661133,
+      "learning_rate": 1.1153143274853802e-05,
+      "loss": 1.666,
+      "step": 9682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.362294912338257,
+      "learning_rate": 1.1152229532163744e-05,
+      "loss": 1.2026,
+      "step": 9683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.101444959640503,
+      "learning_rate": 1.1151315789473684e-05,
+      "loss": 1.2765,
+      "step": 9684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.644017219543457,
+      "learning_rate": 1.1150402046783627e-05,
+      "loss": 1.3111,
+      "step": 9685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8937549591064453,
+      "learning_rate": 1.1149488304093567e-05,
+      "loss": 1.6642,
+      "step": 9686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.296100378036499,
+      "learning_rate": 1.114857456140351e-05,
+      "loss": 1.2952,
+      "step": 9687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5308332443237305,
+      "learning_rate": 1.114766081871345e-05,
+      "loss": 1.6623,
+      "step": 9688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.434265613555908,
+      "learning_rate": 1.1146747076023392e-05,
+      "loss": 1.3476,
+      "step": 9689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9839553833007812,
+      "learning_rate": 1.1145833333333334e-05,
+      "loss": 1.4346,
+      "step": 9690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.522915840148926,
+      "learning_rate": 1.1144919590643275e-05,
+      "loss": 1.3138,
+      "step": 9691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7342159748077393,
+      "learning_rate": 1.1144005847953217e-05,
+      "loss": 1.362,
+      "step": 9692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5744855403900146,
+      "learning_rate": 1.1143092105263158e-05,
+      "loss": 1.6003,
+      "step": 9693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7957592010498047,
+      "learning_rate": 1.11421783625731e-05,
+      "loss": 1.4045,
+      "step": 9694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.298362970352173,
+      "learning_rate": 1.114126461988304e-05,
+      "loss": 1.6114,
+      "step": 9695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2978951930999756,
+      "learning_rate": 1.1140350877192983e-05,
+      "loss": 1.0523,
+      "step": 9696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.136030435562134,
+      "learning_rate": 1.1139437134502925e-05,
+      "loss": 1.4811,
+      "step": 9697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.612651824951172,
+      "learning_rate": 1.1138523391812865e-05,
+      "loss": 1.3566,
+      "step": 9698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1067469120025635,
+      "learning_rate": 1.1137609649122808e-05,
+      "loss": 1.8525,
+      "step": 9699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5159404277801514,
+      "learning_rate": 1.1136695906432748e-05,
+      "loss": 1.4886,
+      "step": 9700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.415036916732788,
+      "learning_rate": 1.113578216374269e-05,
+      "loss": 1.2762,
+      "step": 9701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.513202667236328,
+      "learning_rate": 1.113486842105263e-05,
+      "loss": 1.4572,
+      "step": 9702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3901188373565674,
+      "learning_rate": 1.1133954678362573e-05,
+      "loss": 1.4089,
+      "step": 9703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.327740430831909,
+      "learning_rate": 1.1133040935672515e-05,
+      "loss": 1.2538,
+      "step": 9704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.428762197494507,
+      "learning_rate": 1.1132127192982457e-05,
+      "loss": 1.4553,
+      "step": 9705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.544790506362915,
+      "learning_rate": 1.11312134502924e-05,
+      "loss": 1.5708,
+      "step": 9706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9239511489868164,
+      "learning_rate": 1.1130299707602342e-05,
+      "loss": 2.0061,
+      "step": 9707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6651103496551514,
+      "learning_rate": 1.1129385964912282e-05,
+      "loss": 1.5536,
+      "step": 9708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.204678773880005,
+      "learning_rate": 1.1128472222222225e-05,
+      "loss": 1.9336,
+      "step": 9709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5390610694885254,
+      "learning_rate": 1.1127558479532165e-05,
+      "loss": 1.4238,
+      "step": 9710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6447064876556396,
+      "learning_rate": 1.1126644736842107e-05,
+      "loss": 1.3332,
+      "step": 9711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2375741004943848,
+      "learning_rate": 1.1125730994152048e-05,
+      "loss": 1.3287,
+      "step": 9712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5394654273986816,
+      "learning_rate": 1.112481725146199e-05,
+      "loss": 1.6098,
+      "step": 9713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7807044982910156,
+      "learning_rate": 1.1123903508771932e-05,
+      "loss": 1.5982,
+      "step": 9714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2873642444610596,
+      "learning_rate": 1.1122989766081873e-05,
+      "loss": 1.3733,
+      "step": 9715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5945489406585693,
+      "learning_rate": 1.1122076023391815e-05,
+      "loss": 1.5292,
+      "step": 9716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.198418378829956,
+      "learning_rate": 1.1121162280701756e-05,
+      "loss": 1.9449,
+      "step": 9717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.926652669906616,
+      "learning_rate": 1.1120248538011698e-05,
+      "loss": 2.059,
+      "step": 9718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2381365299224854,
+      "learning_rate": 1.1119334795321638e-05,
+      "loss": 1.404,
+      "step": 9719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.557034492492676,
+      "learning_rate": 1.111842105263158e-05,
+      "loss": 1.4987,
+      "step": 9720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5550143718719482,
+      "learning_rate": 1.1117507309941523e-05,
+      "loss": 1.7604,
+      "step": 9721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.064356565475464,
+      "learning_rate": 1.1116593567251463e-05,
+      "loss": 2.025,
+      "step": 9722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9479118585586548,
+      "learning_rate": 1.1115679824561406e-05,
+      "loss": 1.1773,
+      "step": 9723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6783275604248047,
+      "learning_rate": 1.1114766081871346e-05,
+      "loss": 1.536,
+      "step": 9724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9891867637634277,
+      "learning_rate": 1.1113852339181288e-05,
+      "loss": 1.6602,
+      "step": 9725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2724192142486572,
+      "learning_rate": 1.1112938596491229e-05,
+      "loss": 1.2566,
+      "step": 9726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.959263563156128,
+      "learning_rate": 1.1112024853801171e-05,
+      "loss": 1.4504,
+      "step": 9727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6732053756713867,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 1.3058,
+      "step": 9728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.851288318634033,
+      "learning_rate": 1.1110197368421054e-05,
+      "loss": 2.1,
+      "step": 9729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.756263256072998,
+      "learning_rate": 1.1109283625730996e-05,
+      "loss": 1.7862,
+      "step": 9730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.794184923171997,
+      "learning_rate": 1.1108369883040937e-05,
+      "loss": 1.3832,
+      "step": 9731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.701063871383667,
+      "learning_rate": 1.1107456140350879e-05,
+      "loss": 1.4503,
+      "step": 9732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.363670587539673,
+      "learning_rate": 1.110654239766082e-05,
+      "loss": 1.4135,
+      "step": 9733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.604490280151367,
+      "learning_rate": 1.1105628654970762e-05,
+      "loss": 1.6349,
+      "step": 9734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3027796745300293,
+      "learning_rate": 1.1104714912280704e-05,
+      "loss": 1.9483,
+      "step": 9735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0785868167877197,
+      "learning_rate": 1.1103801169590644e-05,
+      "loss": 1.8249,
+      "step": 9736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.03056263923645,
+      "learning_rate": 1.1102887426900587e-05,
+      "loss": 1.12,
+      "step": 9737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.88258957862854,
+      "learning_rate": 1.1101973684210527e-05,
+      "loss": 1.9307,
+      "step": 9738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6713767051696777,
+      "learning_rate": 1.110105994152047e-05,
+      "loss": 1.6045,
+      "step": 9739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2489418983459473,
+      "learning_rate": 1.110014619883041e-05,
+      "loss": 2.0098,
+      "step": 9740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3857014179229736,
+      "learning_rate": 1.1099232456140352e-05,
+      "loss": 1.4447,
+      "step": 9741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5369935035705566,
+      "learning_rate": 1.1098318713450294e-05,
+      "loss": 1.4309,
+      "step": 9742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.190391778945923,
+      "learning_rate": 1.1097404970760235e-05,
+      "loss": 1.7858,
+      "step": 9743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6385185718536377,
+      "learning_rate": 1.1096491228070177e-05,
+      "loss": 1.7788,
+      "step": 9744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5848798751831055,
+      "learning_rate": 1.1095577485380118e-05,
+      "loss": 1.4214,
+      "step": 9745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7005350589752197,
+      "learning_rate": 1.109466374269006e-05,
+      "loss": 2.0221,
+      "step": 9746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.790050983428955,
+      "learning_rate": 1.109375e-05,
+      "loss": 1.799,
+      "step": 9747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.783445119857788,
+      "learning_rate": 1.1092836257309943e-05,
+      "loss": 1.5025,
+      "step": 9748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2111575603485107,
+      "learning_rate": 1.1091922514619885e-05,
+      "loss": 1.4974,
+      "step": 9749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3347156047821045,
+      "learning_rate": 1.1091008771929825e-05,
+      "loss": 1.3392,
+      "step": 9750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.820126533508301,
+      "learning_rate": 1.1090095029239767e-05,
+      "loss": 1.7389,
+      "step": 9751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.665761709213257,
+      "learning_rate": 1.1089181286549708e-05,
+      "loss": 1.5568,
+      "step": 9752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9390339851379395,
+      "learning_rate": 1.108826754385965e-05,
+      "loss": 1.4989,
+      "step": 9753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4673094749450684,
+      "learning_rate": 1.108735380116959e-05,
+      "loss": 1.4432,
+      "step": 9754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5760443210601807,
+      "learning_rate": 1.1086440058479533e-05,
+      "loss": 1.5835,
+      "step": 9755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.707911968231201,
+      "learning_rate": 1.1085526315789474e-05,
+      "loss": 1.7308,
+      "step": 9756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5515520572662354,
+      "learning_rate": 1.1084612573099416e-05,
+      "loss": 1.4211,
+      "step": 9757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6077206134796143,
+      "learning_rate": 1.1083698830409358e-05,
+      "loss": 1.7255,
+      "step": 9758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.076549768447876,
+      "learning_rate": 1.1082785087719298e-05,
+      "loss": 1.8711,
+      "step": 9759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5700254440307617,
+      "learning_rate": 1.108187134502924e-05,
+      "loss": 1.2654,
+      "step": 9760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2259814739227295,
+      "learning_rate": 1.1080957602339181e-05,
+      "loss": 1.2422,
+      "step": 9761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.607025623321533,
+      "learning_rate": 1.1080043859649123e-05,
+      "loss": 1.6838,
+      "step": 9762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9153342247009277,
+      "learning_rate": 1.1079130116959064e-05,
+      "loss": 1.5627,
+      "step": 9763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2254602909088135,
+      "learning_rate": 1.1078216374269006e-05,
+      "loss": 2.127,
+      "step": 9764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6504507064819336,
+      "learning_rate": 1.1077302631578948e-05,
+      "loss": 1.5614,
+      "step": 9765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8436832427978516,
+      "learning_rate": 1.1076388888888889e-05,
+      "loss": 1.5644,
+      "step": 9766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.86016583442688,
+      "learning_rate": 1.1075475146198831e-05,
+      "loss": 1.7581,
+      "step": 9767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.665804147720337,
+      "learning_rate": 1.1074561403508772e-05,
+      "loss": 1.6592,
+      "step": 9768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.346411943435669,
+      "learning_rate": 1.1073647660818714e-05,
+      "loss": 1.3564,
+      "step": 9769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1829757690429688,
+      "learning_rate": 1.1072733918128654e-05,
+      "loss": 1.6609,
+      "step": 9770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8059487342834473,
+      "learning_rate": 1.1071820175438597e-05,
+      "loss": 1.4641,
+      "step": 9771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4118072986602783,
+      "learning_rate": 1.1070906432748539e-05,
+      "loss": 1.4377,
+      "step": 9772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.433532238006592,
+      "learning_rate": 1.106999269005848e-05,
+      "loss": 1.5951,
+      "step": 9773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0973188877105713,
+      "learning_rate": 1.1069078947368422e-05,
+      "loss": 1.9821,
+      "step": 9774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.852968454360962,
+      "learning_rate": 1.1068165204678362e-05,
+      "loss": 1.6378,
+      "step": 9775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.583998203277588,
+      "learning_rate": 1.1067251461988304e-05,
+      "loss": 1.7296,
+      "step": 9776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.371426582336426,
+      "learning_rate": 1.1066337719298245e-05,
+      "loss": 1.1274,
+      "step": 9777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.565481662750244,
+      "learning_rate": 1.1065423976608187e-05,
+      "loss": 1.476,
+      "step": 9778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8184571266174316,
+      "learning_rate": 1.106451023391813e-05,
+      "loss": 1.7836,
+      "step": 9779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2401180267333984,
+      "learning_rate": 1.106359649122807e-05,
+      "loss": 1.7145,
+      "step": 9780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.694441795349121,
+      "learning_rate": 1.1062682748538012e-05,
+      "loss": 1.6464,
+      "step": 9781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3190693855285645,
+      "learning_rate": 1.1061769005847953e-05,
+      "loss": 1.3097,
+      "step": 9782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.259063482284546,
+      "learning_rate": 1.1060855263157895e-05,
+      "loss": 1.426,
+      "step": 9783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.901202917098999,
+      "learning_rate": 1.1059941520467835e-05,
+      "loss": 1.5236,
+      "step": 9784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7138664722442627,
+      "learning_rate": 1.1059027777777778e-05,
+      "loss": 1.3361,
+      "step": 9785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9757070541381836,
+      "learning_rate": 1.105811403508772e-05,
+      "loss": 1.1183,
+      "step": 9786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.976346969604492,
+      "learning_rate": 1.105720029239766e-05,
+      "loss": 1.7268,
+      "step": 9787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3335511684417725,
+      "learning_rate": 1.1056286549707603e-05,
+      "loss": 1.4228,
+      "step": 9788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6090900897979736,
+      "learning_rate": 1.1055372807017543e-05,
+      "loss": 1.2518,
+      "step": 9789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4181506633758545,
+      "learning_rate": 1.1054459064327487e-05,
+      "loss": 1.3298,
+      "step": 9790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.595921516418457,
+      "learning_rate": 1.105354532163743e-05,
+      "loss": 1.5978,
+      "step": 9791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7492470741271973,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 1.362,
+      "step": 9792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.61427903175354,
+      "learning_rate": 1.1051717836257312e-05,
+      "loss": 1.6118,
+      "step": 9793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5993990898132324,
+      "learning_rate": 1.1050804093567253e-05,
+      "loss": 1.8041,
+      "step": 9794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1326498985290527,
+      "learning_rate": 1.1049890350877195e-05,
+      "loss": 1.5958,
+      "step": 9795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5540549755096436,
+      "learning_rate": 1.1048976608187137e-05,
+      "loss": 1.3383,
+      "step": 9796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5055289268493652,
+      "learning_rate": 1.1048062865497077e-05,
+      "loss": 1.3218,
+      "step": 9797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.646050214767456,
+      "learning_rate": 1.104714912280702e-05,
+      "loss": 1.085,
+      "step": 9798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.430284023284912,
+      "learning_rate": 1.104623538011696e-05,
+      "loss": 1.4162,
+      "step": 9799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.562641143798828,
+      "learning_rate": 1.1045321637426902e-05,
+      "loss": 1.6559,
+      "step": 9800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5393073558807373,
+      "learning_rate": 1.1044407894736843e-05,
+      "loss": 1.6554,
+      "step": 9801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3392512798309326,
+      "learning_rate": 1.1043494152046785e-05,
+      "loss": 1.3001,
+      "step": 9802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.455434799194336,
+      "learning_rate": 1.1042580409356727e-05,
+      "loss": 1.505,
+      "step": 9803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.57106351852417,
+      "learning_rate": 1.1041666666666668e-05,
+      "loss": 1.648,
+      "step": 9804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.451188564300537,
+      "learning_rate": 1.104075292397661e-05,
+      "loss": 1.3893,
+      "step": 9805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.606489419937134,
+      "learning_rate": 1.103983918128655e-05,
+      "loss": 1.4164,
+      "step": 9806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.196939468383789,
+      "learning_rate": 1.1038925438596493e-05,
+      "loss": 1.9306,
+      "step": 9807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5832912921905518,
+      "learning_rate": 1.1038011695906433e-05,
+      "loss": 1.3083,
+      "step": 9808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.691352128982544,
+      "learning_rate": 1.1037097953216376e-05,
+      "loss": 1.7186,
+      "step": 9809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.507270574569702,
+      "learning_rate": 1.1036184210526318e-05,
+      "loss": 1.5428,
+      "step": 9810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7997047901153564,
+      "learning_rate": 1.1035270467836258e-05,
+      "loss": 1.6935,
+      "step": 9811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.86716628074646,
+      "learning_rate": 1.10343567251462e-05,
+      "loss": 0.9849,
+      "step": 9812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6059234142303467,
+      "learning_rate": 1.1033442982456141e-05,
+      "loss": 1.6122,
+      "step": 9813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3886072635650635,
+      "learning_rate": 1.1032529239766083e-05,
+      "loss": 1.3346,
+      "step": 9814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.323136806488037,
+      "learning_rate": 1.1031615497076024e-05,
+      "loss": 1.4056,
+      "step": 9815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4449636936187744,
+      "learning_rate": 1.1030701754385966e-05,
+      "loss": 1.6397,
+      "step": 9816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5636515617370605,
+      "learning_rate": 1.1029788011695908e-05,
+      "loss": 1.2172,
+      "step": 9817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.310286521911621,
+      "learning_rate": 1.1028874269005849e-05,
+      "loss": 1.3607,
+      "step": 9818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9496545791625977,
+      "learning_rate": 1.1027960526315791e-05,
+      "loss": 1.6466,
+      "step": 9819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6047897338867188,
+      "learning_rate": 1.1027046783625732e-05,
+      "loss": 1.5658,
+      "step": 9820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1742258071899414,
+      "learning_rate": 1.1026133040935674e-05,
+      "loss": 1.3758,
+      "step": 9821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.91062355041504,
+      "learning_rate": 1.1025219298245614e-05,
+      "loss": 1.8235,
+      "step": 9822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8426554203033447,
+      "learning_rate": 1.1024305555555557e-05,
+      "loss": 1.5546,
+      "step": 9823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6677865982055664,
+      "learning_rate": 1.1023391812865499e-05,
+      "loss": 1.3467,
+      "step": 9824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8030970096588135,
+      "learning_rate": 1.102247807017544e-05,
+      "loss": 1.6735,
+      "step": 9825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.325239419937134,
+      "learning_rate": 1.1021564327485382e-05,
+      "loss": 1.2355,
+      "step": 9826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6619412899017334,
+      "learning_rate": 1.1020650584795322e-05,
+      "loss": 1.5488,
+      "step": 9827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.99230694770813,
+      "learning_rate": 1.1019736842105264e-05,
+      "loss": 1.6912,
+      "step": 9828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.680972099304199,
+      "learning_rate": 1.1018823099415205e-05,
+      "loss": 1.4693,
+      "step": 9829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0670864582061768,
+      "learning_rate": 1.1017909356725147e-05,
+      "loss": 1.933,
+      "step": 9830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.815315008163452,
+      "learning_rate": 1.101699561403509e-05,
+      "loss": 1.4794,
+      "step": 9831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.583993911743164,
+      "learning_rate": 1.101608187134503e-05,
+      "loss": 1.3742,
+      "step": 9832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1746582984924316,
+      "learning_rate": 1.1015168128654972e-05,
+      "loss": 1.7426,
+      "step": 9833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.625457286834717,
+      "learning_rate": 1.1014254385964913e-05,
+      "loss": 1.5269,
+      "step": 9834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7204315662384033,
+      "learning_rate": 1.1013340643274855e-05,
+      "loss": 1.9054,
+      "step": 9835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4184813499450684,
+      "learning_rate": 1.1012426900584795e-05,
+      "loss": 1.5466,
+      "step": 9836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0246596336364746,
+      "learning_rate": 1.1011513157894738e-05,
+      "loss": 1.8769,
+      "step": 9837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.406310796737671,
+      "learning_rate": 1.101059941520468e-05,
+      "loss": 1.5599,
+      "step": 9838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.343745708465576,
+      "learning_rate": 1.100968567251462e-05,
+      "loss": 1.4224,
+      "step": 9839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7889654636383057,
+      "learning_rate": 1.1008771929824563e-05,
+      "loss": 1.3355,
+      "step": 9840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.242400884628296,
+      "learning_rate": 1.1007858187134503e-05,
+      "loss": 1.1402,
+      "step": 9841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3455770015716553,
+      "learning_rate": 1.1006944444444445e-05,
+      "loss": 1.6883,
+      "step": 9842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6499242782592773,
+      "learning_rate": 1.1006030701754386e-05,
+      "loss": 1.3921,
+      "step": 9843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.365283966064453,
+      "learning_rate": 1.1005116959064328e-05,
+      "loss": 1.2119,
+      "step": 9844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.34448504447937,
+      "learning_rate": 1.100420321637427e-05,
+      "loss": 1.1277,
+      "step": 9845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9475133419036865,
+      "learning_rate": 1.100328947368421e-05,
+      "loss": 1.625,
+      "step": 9846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0489988327026367,
+      "learning_rate": 1.1002375730994153e-05,
+      "loss": 1.1456,
+      "step": 9847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5839619636535645,
+      "learning_rate": 1.1001461988304094e-05,
+      "loss": 1.8096,
+      "step": 9848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.713813304901123,
+      "learning_rate": 1.1000548245614036e-05,
+      "loss": 1.6468,
+      "step": 9849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2844648361206055,
+      "learning_rate": 1.0999634502923976e-05,
+      "loss": 1.4007,
+      "step": 9850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2697947025299072,
+      "learning_rate": 1.0998720760233918e-05,
+      "loss": 1.4587,
+      "step": 9851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.288593053817749,
+      "learning_rate": 1.099780701754386e-05,
+      "loss": 1.7014,
+      "step": 9852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.854557752609253,
+      "learning_rate": 1.0996893274853801e-05,
+      "loss": 1.332,
+      "step": 9853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.627572774887085,
+      "learning_rate": 1.0995979532163743e-05,
+      "loss": 1.6194,
+      "step": 9854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.091506242752075,
+      "learning_rate": 1.0995065789473684e-05,
+      "loss": 1.8334,
+      "step": 9855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5800065994262695,
+      "learning_rate": 1.0994152046783626e-05,
+      "loss": 1.3991,
+      "step": 9856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.940423011779785,
+      "learning_rate": 1.0993238304093567e-05,
+      "loss": 1.6272,
+      "step": 9857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.125572681427002,
+      "learning_rate": 1.0992324561403509e-05,
+      "loss": 1.6891,
+      "step": 9858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5649471282958984,
+      "learning_rate": 1.0991410818713451e-05,
+      "loss": 1.3544,
+      "step": 9859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8680474758148193,
+      "learning_rate": 1.0990497076023392e-05,
+      "loss": 1.6096,
+      "step": 9860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6659607887268066,
+      "learning_rate": 1.0989583333333334e-05,
+      "loss": 1.5019,
+      "step": 9861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2115509510040283,
+      "learning_rate": 1.0988669590643274e-05,
+      "loss": 1.2151,
+      "step": 9862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.610567569732666,
+      "learning_rate": 1.0987755847953217e-05,
+      "loss": 1.6774,
+      "step": 9863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.615344524383545,
+      "learning_rate": 1.0986842105263157e-05,
+      "loss": 1.354,
+      "step": 9864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2650821208953857,
+      "learning_rate": 1.09859283625731e-05,
+      "loss": 1.7101,
+      "step": 9865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.888773798942566,
+      "learning_rate": 1.0985014619883042e-05,
+      "loss": 1.1623,
+      "step": 9866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5076723098754883,
+      "learning_rate": 1.0984100877192982e-05,
+      "loss": 1.3318,
+      "step": 9867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7246620655059814,
+      "learning_rate": 1.0983187134502924e-05,
+      "loss": 1.6521,
+      "step": 9868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5169296264648438,
+      "learning_rate": 1.0982273391812865e-05,
+      "loss": 1.57,
+      "step": 9869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2566041946411133,
+      "learning_rate": 1.0981359649122807e-05,
+      "loss": 1.1504,
+      "step": 9870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9057767391204834,
+      "learning_rate": 1.0980445906432748e-05,
+      "loss": 1.6401,
+      "step": 9871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0247581005096436,
+      "learning_rate": 1.097953216374269e-05,
+      "loss": 1.7492,
+      "step": 9872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.900747299194336,
+      "learning_rate": 1.0978618421052632e-05,
+      "loss": 1.4631,
+      "step": 9873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7577898502349854,
+      "learning_rate": 1.0977704678362573e-05,
+      "loss": 1.4376,
+      "step": 9874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7541043758392334,
+      "learning_rate": 1.0976790935672515e-05,
+      "loss": 1.626,
+      "step": 9875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.651143789291382,
+      "learning_rate": 1.0975877192982459e-05,
+      "loss": 1.4932,
+      "step": 9876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1494851112365723,
+      "learning_rate": 1.09749634502924e-05,
+      "loss": 1.5252,
+      "step": 9877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.544188976287842,
+      "learning_rate": 1.0974049707602342e-05,
+      "loss": 1.4965,
+      "step": 9878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.65946102142334,
+      "learning_rate": 1.0973135964912282e-05,
+      "loss": 1.5422,
+      "step": 9879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8377346992492676,
+      "learning_rate": 1.0972222222222224e-05,
+      "loss": 1.5112,
+      "step": 9880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.486320734024048,
+      "learning_rate": 1.0971308479532165e-05,
+      "loss": 1.5364,
+      "step": 9881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.444396495819092,
+      "learning_rate": 1.0970394736842107e-05,
+      "loss": 1.4291,
+      "step": 9882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3729257583618164,
+      "learning_rate": 1.096948099415205e-05,
+      "loss": 1.4319,
+      "step": 9883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6764283180236816,
+      "learning_rate": 1.096856725146199e-05,
+      "loss": 1.7468,
+      "step": 9884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.468991756439209,
+      "learning_rate": 1.0967653508771932e-05,
+      "loss": 1.7965,
+      "step": 9885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0924441814422607,
+      "learning_rate": 1.0966739766081873e-05,
+      "loss": 1.7688,
+      "step": 9886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7918341159820557,
+      "learning_rate": 1.0965826023391815e-05,
+      "loss": 1.6951,
+      "step": 9887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.77706241607666,
+      "learning_rate": 1.0964912280701755e-05,
+      "loss": 1.5811,
+      "step": 9888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.691323757171631,
+      "learning_rate": 1.0963998538011697e-05,
+      "loss": 1.6959,
+      "step": 9889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.971327543258667,
+      "learning_rate": 1.096308479532164e-05,
+      "loss": 1.4637,
+      "step": 9890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6225039958953857,
+      "learning_rate": 1.096217105263158e-05,
+      "loss": 1.7243,
+      "step": 9891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0934746265411377,
+      "learning_rate": 1.0961257309941522e-05,
+      "loss": 1.8562,
+      "step": 9892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5268447399139404,
+      "learning_rate": 1.0960343567251463e-05,
+      "loss": 1.3195,
+      "step": 9893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9353299140930176,
+      "learning_rate": 1.0959429824561405e-05,
+      "loss": 1.5992,
+      "step": 9894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3087310791015625,
+      "learning_rate": 1.0958516081871346e-05,
+      "loss": 1.9817,
+      "step": 9895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.794126272201538,
+      "learning_rate": 1.0957602339181288e-05,
+      "loss": 1.659,
+      "step": 9896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5496654510498047,
+      "learning_rate": 1.095668859649123e-05,
+      "loss": 1.6161,
+      "step": 9897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7599966526031494,
+      "learning_rate": 1.095577485380117e-05,
+      "loss": 1.6029,
+      "step": 9898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6287670135498047,
+      "learning_rate": 1.0954861111111113e-05,
+      "loss": 1.3568,
+      "step": 9899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5099503993988037,
+      "learning_rate": 1.0953947368421053e-05,
+      "loss": 1.5245,
+      "step": 9900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3575785160064697,
+      "learning_rate": 1.0953033625730996e-05,
+      "loss": 1.3652,
+      "step": 9901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.726499557495117,
+      "learning_rate": 1.0952119883040936e-05,
+      "loss": 1.8886,
+      "step": 9902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.685127019882202,
+      "learning_rate": 1.0951206140350878e-05,
+      "loss": 1.9292,
+      "step": 9903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6062867641448975,
+      "learning_rate": 1.095029239766082e-05,
+      "loss": 1.7362,
+      "step": 9904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7972311973571777,
+      "learning_rate": 1.0949378654970761e-05,
+      "loss": 1.3409,
+      "step": 9905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1230788230895996,
+      "learning_rate": 1.0948464912280703e-05,
+      "loss": 1.3744,
+      "step": 9906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3131916522979736,
+      "learning_rate": 1.0947551169590644e-05,
+      "loss": 1.2515,
+      "step": 9907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.481555700302124,
+      "learning_rate": 1.0946637426900586e-05,
+      "loss": 1.4635,
+      "step": 9908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2867271900177,
+      "learning_rate": 1.0945723684210527e-05,
+      "loss": 1.2799,
+      "step": 9909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9708714485168457,
+      "learning_rate": 1.0944809941520469e-05,
+      "loss": 1.7333,
+      "step": 9910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.525275707244873,
+      "learning_rate": 1.0943896198830411e-05,
+      "loss": 1.4359,
+      "step": 9911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.421882152557373,
+      "learning_rate": 1.0942982456140352e-05,
+      "loss": 1.6116,
+      "step": 9912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5892679691314697,
+      "learning_rate": 1.0942068713450294e-05,
+      "loss": 1.5778,
+      "step": 9913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6963250637054443,
+      "learning_rate": 1.0941154970760234e-05,
+      "loss": 1.6861,
+      "step": 9914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5856807231903076,
+      "learning_rate": 1.0940241228070177e-05,
+      "loss": 1.2847,
+      "step": 9915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4366378784179688,
+      "learning_rate": 1.0939327485380117e-05,
+      "loss": 1.4019,
+      "step": 9916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.478844165802002,
+      "learning_rate": 1.093841374269006e-05,
+      "loss": 1.4561,
+      "step": 9917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5616631507873535,
+      "learning_rate": 1.0937500000000002e-05,
+      "loss": 1.5907,
+      "step": 9918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.100088596343994,
+      "learning_rate": 1.0936586257309942e-05,
+      "loss": 1.2073,
+      "step": 9919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3964853286743164,
+      "learning_rate": 1.0935672514619884e-05,
+      "loss": 1.599,
+      "step": 9920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8921003341674805,
+      "learning_rate": 1.0934758771929825e-05,
+      "loss": 1.5319,
+      "step": 9921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4931278228759766,
+      "learning_rate": 1.0933845029239767e-05,
+      "loss": 1.3108,
+      "step": 9922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2583346366882324,
+      "learning_rate": 1.0932931286549708e-05,
+      "loss": 1.4246,
+      "step": 9923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6641016006469727,
+      "learning_rate": 1.093201754385965e-05,
+      "loss": 1.1602,
+      "step": 9924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8822131156921387,
+      "learning_rate": 1.0931103801169592e-05,
+      "loss": 1.7043,
+      "step": 9925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.945839524269104,
+      "learning_rate": 1.0930190058479533e-05,
+      "loss": 1.1385,
+      "step": 9926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2576584815979004,
+      "learning_rate": 1.0929276315789475e-05,
+      "loss": 1.9626,
+      "step": 9927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6681630611419678,
+      "learning_rate": 1.0928362573099415e-05,
+      "loss": 1.5762,
+      "step": 9928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.852587938308716,
+      "learning_rate": 1.0927448830409358e-05,
+      "loss": 1.6017,
+      "step": 9929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4095635414123535,
+      "learning_rate": 1.0926535087719298e-05,
+      "loss": 1.5382,
+      "step": 9930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.355836868286133,
+      "learning_rate": 1.092562134502924e-05,
+      "loss": 1.2819,
+      "step": 9931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.821681499481201,
+      "learning_rate": 1.0924707602339183e-05,
+      "loss": 1.5805,
+      "step": 9932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1666765213012695,
+      "learning_rate": 1.0923793859649123e-05,
+      "loss": 1.1788,
+      "step": 9933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7485320568084717,
+      "learning_rate": 1.0922880116959065e-05,
+      "loss": 1.4844,
+      "step": 9934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0483548641204834,
+      "learning_rate": 1.0921966374269006e-05,
+      "loss": 2.0794,
+      "step": 9935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.066667079925537,
+      "learning_rate": 1.0921052631578948e-05,
+      "loss": 1.9951,
+      "step": 9936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5242536067962646,
+      "learning_rate": 1.0920138888888889e-05,
+      "loss": 1.3777,
+      "step": 9937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4715754985809326,
+      "learning_rate": 1.091922514619883e-05,
+      "loss": 1.4889,
+      "step": 9938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5842161178588867,
+      "learning_rate": 1.0918311403508773e-05,
+      "loss": 1.6613,
+      "step": 9939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.609795331954956,
+      "learning_rate": 1.0917397660818714e-05,
+      "loss": 1.437,
+      "step": 9940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0238521099090576,
+      "learning_rate": 1.0916483918128656e-05,
+      "loss": 1.5495,
+      "step": 9941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5666072368621826,
+      "learning_rate": 1.0915570175438596e-05,
+      "loss": 1.4313,
+      "step": 9942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.546274423599243,
+      "learning_rate": 1.0914656432748538e-05,
+      "loss": 1.4024,
+      "step": 9943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6107053756713867,
+      "learning_rate": 1.0913742690058479e-05,
+      "loss": 1.3124,
+      "step": 9944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.375370740890503,
+      "learning_rate": 1.0912828947368421e-05,
+      "loss": 1.6224,
+      "step": 9945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.155345916748047,
+      "learning_rate": 1.0911915204678363e-05,
+      "loss": 2.3811,
+      "step": 9946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6399567127227783,
+      "learning_rate": 1.0911001461988304e-05,
+      "loss": 1.1943,
+      "step": 9947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.992550849914551,
+      "learning_rate": 1.0910087719298246e-05,
+      "loss": 1.68,
+      "step": 9948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5381691455841064,
+      "learning_rate": 1.0909173976608187e-05,
+      "loss": 1.407,
+      "step": 9949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.036240339279175,
+      "learning_rate": 1.0908260233918129e-05,
+      "loss": 1.9063,
+      "step": 9950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5374107360839844,
+      "learning_rate": 1.090734649122807e-05,
+      "loss": 1.4255,
+      "step": 9951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5394113063812256,
+      "learning_rate": 1.0906432748538012e-05,
+      "loss": 1.4349,
+      "step": 9952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2521512508392334,
+      "learning_rate": 1.0905519005847954e-05,
+      "loss": 1.9612,
+      "step": 9953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0790927410125732,
+      "learning_rate": 1.0904605263157894e-05,
+      "loss": 1.5114,
+      "step": 9954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.862679958343506,
+      "learning_rate": 1.0903691520467837e-05,
+      "loss": 1.7762,
+      "step": 9955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.270125150680542,
+      "learning_rate": 1.0902777777777777e-05,
+      "loss": 1.902,
+      "step": 9956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7407591342926025,
+      "learning_rate": 1.090186403508772e-05,
+      "loss": 1.6653,
+      "step": 9957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6788690090179443,
+      "learning_rate": 1.090095029239766e-05,
+      "loss": 1.6557,
+      "step": 9958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6146953105926514,
+      "learning_rate": 1.0900036549707602e-05,
+      "loss": 1.7242,
+      "step": 9959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8933231830596924,
+      "learning_rate": 1.0899122807017544e-05,
+      "loss": 1.8152,
+      "step": 9960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.399965286254883,
+      "learning_rate": 1.0898209064327487e-05,
+      "loss": 1.4029,
+      "step": 9961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.879140853881836,
+      "learning_rate": 1.0897295321637429e-05,
+      "loss": 1.8,
+      "step": 9962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.011038303375244,
+      "learning_rate": 1.0896381578947371e-05,
+      "loss": 1.7417,
+      "step": 9963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5289952754974365,
+      "learning_rate": 1.0895467836257312e-05,
+      "loss": 1.3637,
+      "step": 9964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.657120704650879,
+      "learning_rate": 1.0894554093567254e-05,
+      "loss": 1.6487,
+      "step": 9965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.690764904022217,
+      "learning_rate": 1.0893640350877194e-05,
+      "loss": 1.5053,
+      "step": 9966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.472496747970581,
+      "learning_rate": 1.0892726608187137e-05,
+      "loss": 1.157,
+      "step": 9967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.844998598098755,
+      "learning_rate": 1.0891812865497077e-05,
+      "loss": 1.7642,
+      "step": 9968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.72956919670105,
+      "learning_rate": 1.089089912280702e-05,
+      "loss": 1.8829,
+      "step": 9969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4250872135162354,
+      "learning_rate": 1.0889985380116962e-05,
+      "loss": 1.7712,
+      "step": 9970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.612295627593994,
+      "learning_rate": 1.0889071637426902e-05,
+      "loss": 1.4564,
+      "step": 9971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.444254159927368,
+      "learning_rate": 1.0888157894736844e-05,
+      "loss": 1.2512,
+      "step": 9972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5347859859466553,
+      "learning_rate": 1.0887244152046785e-05,
+      "loss": 1.7039,
+      "step": 9973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6263983249664307,
+      "learning_rate": 1.0886330409356727e-05,
+      "loss": 1.4023,
+      "step": 9974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.744945764541626,
+      "learning_rate": 1.0885416666666668e-05,
+      "loss": 1.5365,
+      "step": 9975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.509399175643921,
+      "learning_rate": 1.088450292397661e-05,
+      "loss": 1.441,
+      "step": 9976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.116856336593628,
+      "learning_rate": 1.0883589181286552e-05,
+      "loss": 2.1323,
+      "step": 9977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.498579263687134,
+      "learning_rate": 1.0882675438596493e-05,
+      "loss": 1.3199,
+      "step": 9978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9953649044036865,
+      "learning_rate": 1.0881761695906435e-05,
+      "loss": 1.2081,
+      "step": 9979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1906981468200684,
+      "learning_rate": 1.0880847953216375e-05,
+      "loss": 1.4786,
+      "step": 9980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9465107917785645,
+      "learning_rate": 1.0879934210526317e-05,
+      "loss": 1.7425,
+      "step": 9981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.649179458618164,
+      "learning_rate": 1.0879020467836258e-05,
+      "loss": 1.6261,
+      "step": 9982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3044321537017822,
+      "learning_rate": 1.08781067251462e-05,
+      "loss": 1.3159,
+      "step": 9983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7748732566833496,
+      "learning_rate": 1.0877192982456142e-05,
+      "loss": 1.6823,
+      "step": 9984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.429246664047241,
+      "learning_rate": 1.0876279239766083e-05,
+      "loss": 1.5912,
+      "step": 9985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6166436672210693,
+      "learning_rate": 1.0875365497076025e-05,
+      "loss": 1.377,
+      "step": 9986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1478865146636963,
+      "learning_rate": 1.0874451754385966e-05,
+      "loss": 2.5628,
+      "step": 9987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6345267295837402,
+      "learning_rate": 1.0873538011695908e-05,
+      "loss": 1.6835,
+      "step": 9988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.297462224960327,
+      "learning_rate": 1.0872624269005848e-05,
+      "loss": 1.1866,
+      "step": 9989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7709176540374756,
+      "learning_rate": 1.087171052631579e-05,
+      "loss": 1.2932,
+      "step": 9990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.255795955657959,
+      "learning_rate": 1.0870796783625733e-05,
+      "loss": 1.2655,
+      "step": 9991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1777899265289307,
+      "learning_rate": 1.0869883040935673e-05,
+      "loss": 1.3126,
+      "step": 9992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6612637042999268,
+      "learning_rate": 1.0868969298245616e-05,
+      "loss": 1.7176,
+      "step": 9993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8932437896728516,
+      "learning_rate": 1.0868055555555556e-05,
+      "loss": 1.0487,
+      "step": 9994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7251055240631104,
+      "learning_rate": 1.0867141812865498e-05,
+      "loss": 1.4079,
+      "step": 9995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.031526803970337,
+      "learning_rate": 1.0866228070175439e-05,
+      "loss": 1.7702,
+      "step": 9996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.53779935836792,
+      "learning_rate": 1.0865314327485381e-05,
+      "loss": 1.5383,
+      "step": 9997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.647773504257202,
+      "learning_rate": 1.0864400584795323e-05,
+      "loss": 1.6325,
+      "step": 9998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.344674587249756,
+      "learning_rate": 1.0863486842105264e-05,
+      "loss": 1.3533,
+      "step": 9999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.277775526046753,
+      "learning_rate": 1.0862573099415206e-05,
+      "loss": 1.3611,
+      "step": 10000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.783418655395508,
+      "learning_rate": 1.0861659356725147e-05,
+      "loss": 1.3774,
+      "step": 10001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8623294830322266,
+      "learning_rate": 1.0860745614035089e-05,
+      "loss": 1.4009,
+      "step": 10002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1330158710479736,
+      "learning_rate": 1.085983187134503e-05,
+      "loss": 1.4883,
+      "step": 10003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6566503047943115,
+      "learning_rate": 1.0858918128654972e-05,
+      "loss": 1.5381,
+      "step": 10004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.972858428955078,
+      "learning_rate": 1.0858004385964914e-05,
+      "loss": 1.6784,
+      "step": 10005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6833269596099854,
+      "learning_rate": 1.0857090643274854e-05,
+      "loss": 1.7878,
+      "step": 10006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6047096252441406,
+      "learning_rate": 1.0856176900584797e-05,
+      "loss": 1.4494,
+      "step": 10007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.472952365875244,
+      "learning_rate": 1.0855263157894737e-05,
+      "loss": 1.1455,
+      "step": 10008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2756001949310303,
+      "learning_rate": 1.085434941520468e-05,
+      "loss": 1.4303,
+      "step": 10009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2141542434692383,
+      "learning_rate": 1.085343567251462e-05,
+      "loss": 1.1947,
+      "step": 10010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3259291648864746,
+      "learning_rate": 1.0852521929824562e-05,
+      "loss": 1.461,
+      "step": 10011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7334301471710205,
+      "learning_rate": 1.0851608187134504e-05,
+      "loss": 1.5546,
+      "step": 10012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4070308208465576,
+      "learning_rate": 1.0850694444444445e-05,
+      "loss": 1.4874,
+      "step": 10013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.726919174194336,
+      "learning_rate": 1.0849780701754387e-05,
+      "loss": 1.6892,
+      "step": 10014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5115740299224854,
+      "learning_rate": 1.0848866959064328e-05,
+      "loss": 1.5186,
+      "step": 10015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.727222442626953,
+      "learning_rate": 1.084795321637427e-05,
+      "loss": 1.6347,
+      "step": 10016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.747563362121582,
+      "learning_rate": 1.084703947368421e-05,
+      "loss": 1.3498,
+      "step": 10017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0399463176727295,
+      "learning_rate": 1.0846125730994153e-05,
+      "loss": 1.2197,
+      "step": 10018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2212321758270264,
+      "learning_rate": 1.0845211988304095e-05,
+      "loss": 1.3547,
+      "step": 10019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.040321111679077,
+      "learning_rate": 1.0844298245614035e-05,
+      "loss": 1.6807,
+      "step": 10020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.649080514907837,
+      "learning_rate": 1.0843384502923978e-05,
+      "loss": 1.2298,
+      "step": 10021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1589579582214355,
+      "learning_rate": 1.0842470760233918e-05,
+      "loss": 1.781,
+      "step": 10022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.305901527404785,
+      "learning_rate": 1.084155701754386e-05,
+      "loss": 1.3843,
+      "step": 10023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.451519012451172,
+      "learning_rate": 1.08406432748538e-05,
+      "loss": 1.5329,
+      "step": 10024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.915773630142212,
+      "learning_rate": 1.0839729532163743e-05,
+      "loss": 1.7284,
+      "step": 10025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.444133758544922,
+      "learning_rate": 1.0838815789473685e-05,
+      "loss": 1.4181,
+      "step": 10026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6083474159240723,
+      "learning_rate": 1.0837902046783626e-05,
+      "loss": 1.1747,
+      "step": 10027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8816592693328857,
+      "learning_rate": 1.0836988304093568e-05,
+      "loss": 1.7953,
+      "step": 10028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.39399790763855,
+      "learning_rate": 1.0836074561403509e-05,
+      "loss": 1.483,
+      "step": 10029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.527698040008545,
+      "learning_rate": 1.083516081871345e-05,
+      "loss": 1.3591,
+      "step": 10030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4539778232574463,
+      "learning_rate": 1.0834247076023391e-05,
+      "loss": 1.5474,
+      "step": 10031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.007683038711548,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 1.8267,
+      "step": 10032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.974226951599121,
+      "learning_rate": 1.0832419590643276e-05,
+      "loss": 1.9454,
+      "step": 10033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2547221183776855,
+      "learning_rate": 1.0831505847953216e-05,
+      "loss": 1.4638,
+      "step": 10034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.328155040740967,
+      "learning_rate": 1.0830592105263158e-05,
+      "loss": 1.4641,
+      "step": 10035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.506418228149414,
+      "learning_rate": 1.0829678362573099e-05,
+      "loss": 1.3553,
+      "step": 10036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.779236316680908,
+      "learning_rate": 1.0828764619883041e-05,
+      "loss": 1.5214,
+      "step": 10037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.07228684425354,
+      "learning_rate": 1.0827850877192982e-05,
+      "loss": 1.2095,
+      "step": 10038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5976390838623047,
+      "learning_rate": 1.0826937134502924e-05,
+      "loss": 1.4468,
+      "step": 10039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8160078525543213,
+      "learning_rate": 1.0826023391812866e-05,
+      "loss": 1.6519,
+      "step": 10040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.262021780014038,
+      "learning_rate": 1.0825109649122807e-05,
+      "loss": 1.2187,
+      "step": 10041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.542936325073242,
+      "learning_rate": 1.0824195906432749e-05,
+      "loss": 1.4078,
+      "step": 10042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9183497428894043,
+      "learning_rate": 1.082328216374269e-05,
+      "loss": 1.6656,
+      "step": 10043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8492109775543213,
+      "learning_rate": 1.0822368421052632e-05,
+      "loss": 1.569,
+      "step": 10044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.350536346435547,
+      "learning_rate": 1.0821454678362572e-05,
+      "loss": 1.1655,
+      "step": 10045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3006279468536377,
+      "learning_rate": 1.0820540935672514e-05,
+      "loss": 1.3982,
+      "step": 10046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0481297969818115,
+      "learning_rate": 1.0819627192982458e-05,
+      "loss": 1.4646,
+      "step": 10047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7026140689849854,
+      "learning_rate": 1.0818713450292399e-05,
+      "loss": 1.6391,
+      "step": 10048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6953303813934326,
+      "learning_rate": 1.0817799707602341e-05,
+      "loss": 1.5515,
+      "step": 10049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4646122455596924,
+      "learning_rate": 1.0816885964912283e-05,
+      "loss": 1.2889,
+      "step": 10050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.23732590675354,
+      "learning_rate": 1.0815972222222224e-05,
+      "loss": 1.249,
+      "step": 10051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6933484077453613,
+      "learning_rate": 1.0815058479532166e-05,
+      "loss": 2.2146,
+      "step": 10052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0353715419769287,
+      "learning_rate": 1.0814144736842107e-05,
+      "loss": 1.3777,
+      "step": 10053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4234566688537598,
+      "learning_rate": 1.0813230994152049e-05,
+      "loss": 1.4344,
+      "step": 10054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.757073163986206,
+      "learning_rate": 1.081231725146199e-05,
+      "loss": 1.4763,
+      "step": 10055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0185911655426025,
+      "learning_rate": 1.0811403508771932e-05,
+      "loss": 1.7726,
+      "step": 10056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1325252056121826,
+      "learning_rate": 1.0810489766081874e-05,
+      "loss": 1.6675,
+      "step": 10057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.024432897567749,
+      "learning_rate": 1.0809576023391814e-05,
+      "loss": 1.9047,
+      "step": 10058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.720766067504883,
+      "learning_rate": 1.0808662280701757e-05,
+      "loss": 1.7805,
+      "step": 10059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6659555435180664,
+      "learning_rate": 1.0807748538011697e-05,
+      "loss": 1.7229,
+      "step": 10060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5035202503204346,
+      "learning_rate": 1.080683479532164e-05,
+      "loss": 1.4308,
+      "step": 10061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.447458505630493,
+      "learning_rate": 1.080592105263158e-05,
+      "loss": 1.4254,
+      "step": 10062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5855090618133545,
+      "learning_rate": 1.0805007309941522e-05,
+      "loss": 1.4548,
+      "step": 10063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.377254009246826,
+      "learning_rate": 1.0804093567251464e-05,
+      "loss": 1.0492,
+      "step": 10064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.852774143218994,
+      "learning_rate": 1.0803179824561405e-05,
+      "loss": 1.7702,
+      "step": 10065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.850205421447754,
+      "learning_rate": 1.0802266081871347e-05,
+      "loss": 1.3958,
+      "step": 10066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.729340076446533,
+      "learning_rate": 1.0801352339181288e-05,
+      "loss": 1.363,
+      "step": 10067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0246126651763916,
+      "learning_rate": 1.080043859649123e-05,
+      "loss": 1.8385,
+      "step": 10068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4856488704681396,
+      "learning_rate": 1.079952485380117e-05,
+      "loss": 1.4754,
+      "step": 10069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5255861282348633,
+      "learning_rate": 1.0798611111111113e-05,
+      "loss": 1.424,
+      "step": 10070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3997514247894287,
+      "learning_rate": 1.0797697368421055e-05,
+      "loss": 1.1682,
+      "step": 10071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.726513624191284,
+      "learning_rate": 1.0796783625730995e-05,
+      "loss": 1.326,
+      "step": 10072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.519912004470825,
+      "learning_rate": 1.0795869883040937e-05,
+      "loss": 1.6451,
+      "step": 10073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7780933380126953,
+      "learning_rate": 1.0794956140350878e-05,
+      "loss": 1.595,
+      "step": 10074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1982827186584473,
+      "learning_rate": 1.079404239766082e-05,
+      "loss": 1.3237,
+      "step": 10075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.563279867172241,
+      "learning_rate": 1.079312865497076e-05,
+      "loss": 1.7275,
+      "step": 10076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7541282176971436,
+      "learning_rate": 1.0792214912280703e-05,
+      "loss": 1.6984,
+      "step": 10077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.919081926345825,
+      "learning_rate": 1.0791301169590645e-05,
+      "loss": 1.8591,
+      "step": 10078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.395261764526367,
+      "learning_rate": 1.0790387426900586e-05,
+      "loss": 1.2018,
+      "step": 10079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6375129222869873,
+      "learning_rate": 1.0789473684210528e-05,
+      "loss": 1.5237,
+      "step": 10080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.240776777267456,
+      "learning_rate": 1.0788559941520468e-05,
+      "loss": 1.5304,
+      "step": 10081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.76806378364563,
+      "learning_rate": 1.078764619883041e-05,
+      "loss": 1.469,
+      "step": 10082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3536205291748047,
+      "learning_rate": 1.0786732456140351e-05,
+      "loss": 1.3919,
+      "step": 10083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5311098098754883,
+      "learning_rate": 1.0785818713450293e-05,
+      "loss": 1.3283,
+      "step": 10084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4758849143981934,
+      "learning_rate": 1.0784904970760236e-05,
+      "loss": 1.4318,
+      "step": 10085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3056886196136475,
+      "learning_rate": 1.0783991228070176e-05,
+      "loss": 2.1201,
+      "step": 10086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7212932109832764,
+      "learning_rate": 1.0783077485380118e-05,
+      "loss": 1.47,
+      "step": 10087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1942193508148193,
+      "learning_rate": 1.0782163742690059e-05,
+      "loss": 1.1639,
+      "step": 10088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6376280784606934,
+      "learning_rate": 1.0781250000000001e-05,
+      "loss": 1.9825,
+      "step": 10089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.234340190887451,
+      "learning_rate": 1.0780336257309942e-05,
+      "loss": 1.4065,
+      "step": 10090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.473682165145874,
+      "learning_rate": 1.0779422514619884e-05,
+      "loss": 1.5472,
+      "step": 10091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3674063682556152,
+      "learning_rate": 1.0778508771929826e-05,
+      "loss": 1.4997,
+      "step": 10092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6726880073547363,
+      "learning_rate": 1.0777595029239767e-05,
+      "loss": 2.067,
+      "step": 10093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.540954351425171,
+      "learning_rate": 1.0776681286549709e-05,
+      "loss": 1.5407,
+      "step": 10094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.303385019302368,
+      "learning_rate": 1.077576754385965e-05,
+      "loss": 1.5842,
+      "step": 10095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1441426277160645,
+      "learning_rate": 1.0774853801169592e-05,
+      "loss": 1.032,
+      "step": 10096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4445221424102783,
+      "learning_rate": 1.0773940058479532e-05,
+      "loss": 1.3887,
+      "step": 10097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.722951650619507,
+      "learning_rate": 1.0773026315789474e-05,
+      "loss": 1.5416,
+      "step": 10098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.198498010635376,
+      "learning_rate": 1.0772112573099417e-05,
+      "loss": 1.1247,
+      "step": 10099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1497676372528076,
+      "learning_rate": 1.0771198830409357e-05,
+      "loss": 1.7957,
+      "step": 10100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6467232704162598,
+      "learning_rate": 1.07702850877193e-05,
+      "loss": 1.831,
+      "step": 10101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9688498973846436,
+      "learning_rate": 1.076937134502924e-05,
+      "loss": 1.685,
+      "step": 10102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.79890775680542,
+      "learning_rate": 1.0768457602339182e-05,
+      "loss": 1.6564,
+      "step": 10103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6375434398651123,
+      "learning_rate": 1.0767543859649123e-05,
+      "loss": 1.6266,
+      "step": 10104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.296095848083496,
+      "learning_rate": 1.0766630116959065e-05,
+      "loss": 1.4308,
+      "step": 10105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.837282657623291,
+      "learning_rate": 1.0765716374269007e-05,
+      "loss": 1.7921,
+      "step": 10106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.368312358856201,
+      "learning_rate": 1.0764802631578948e-05,
+      "loss": 1.4328,
+      "step": 10107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8652021884918213,
+      "learning_rate": 1.076388888888889e-05,
+      "loss": 1.4414,
+      "step": 10108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6394569873809814,
+      "learning_rate": 1.076297514619883e-05,
+      "loss": 1.6197,
+      "step": 10109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3095757961273193,
+      "learning_rate": 1.0762061403508773e-05,
+      "loss": 1.333,
+      "step": 10110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2183852195739746,
+      "learning_rate": 1.0761147660818713e-05,
+      "loss": 1.0778,
+      "step": 10111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.104113817214966,
+      "learning_rate": 1.0760233918128655e-05,
+      "loss": 1.7536,
+      "step": 10112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4913172721862793,
+      "learning_rate": 1.0759320175438598e-05,
+      "loss": 1.7173,
+      "step": 10113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3199172019958496,
+      "learning_rate": 1.0758406432748538e-05,
+      "loss": 1.9065,
+      "step": 10114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0192008018493652,
+      "learning_rate": 1.075749269005848e-05,
+      "loss": 1.9611,
+      "step": 10115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.604222059249878,
+      "learning_rate": 1.075657894736842e-05,
+      "loss": 1.2838,
+      "step": 10116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7020537853240967,
+      "learning_rate": 1.0755665204678363e-05,
+      "loss": 1.4096,
+      "step": 10117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.768342971801758,
+      "learning_rate": 1.0754751461988304e-05,
+      "loss": 1.4686,
+      "step": 10118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.796732187271118,
+      "learning_rate": 1.0753837719298246e-05,
+      "loss": 1.6705,
+      "step": 10119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3740286827087402,
+      "learning_rate": 1.0752923976608188e-05,
+      "loss": 1.2243,
+      "step": 10120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9479753971099854,
+      "learning_rate": 1.0752010233918129e-05,
+      "loss": 1.9054,
+      "step": 10121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9288675785064697,
+      "learning_rate": 1.075109649122807e-05,
+      "loss": 1.2616,
+      "step": 10122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5936615467071533,
+      "learning_rate": 1.0750182748538011e-05,
+      "loss": 1.5316,
+      "step": 10123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.973201036453247,
+      "learning_rate": 1.0749269005847954e-05,
+      "loss": 1.7409,
+      "step": 10124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1979129314422607,
+      "learning_rate": 1.0748355263157894e-05,
+      "loss": 1.6294,
+      "step": 10125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6796233654022217,
+      "learning_rate": 1.0747441520467836e-05,
+      "loss": 1.3691,
+      "step": 10126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6830334663391113,
+      "learning_rate": 1.0746527777777778e-05,
+      "loss": 1.5847,
+      "step": 10127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.682304859161377,
+      "learning_rate": 1.0745614035087719e-05,
+      "loss": 1.5883,
+      "step": 10128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.768479108810425,
+      "learning_rate": 1.0744700292397661e-05,
+      "loss": 2.0166,
+      "step": 10129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9201769828796387,
+      "learning_rate": 1.0743786549707602e-05,
+      "loss": 1.6096,
+      "step": 10130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3480751514434814,
+      "learning_rate": 1.0742872807017544e-05,
+      "loss": 1.8282,
+      "step": 10131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.665430784225464,
+      "learning_rate": 1.0741959064327488e-05,
+      "loss": 1.3954,
+      "step": 10132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0895113945007324,
+      "learning_rate": 1.0741045321637428e-05,
+      "loss": 1.6483,
+      "step": 10133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0154247283935547,
+      "learning_rate": 1.074013157894737e-05,
+      "loss": 1.6669,
+      "step": 10134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.630188226699829,
+      "learning_rate": 1.0739217836257311e-05,
+      "loss": 1.8501,
+      "step": 10135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.364417552947998,
+      "learning_rate": 1.0738304093567253e-05,
+      "loss": 1.5349,
+      "step": 10136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.700122833251953,
+      "learning_rate": 1.0737390350877196e-05,
+      "loss": 1.7073,
+      "step": 10137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.147740364074707,
+      "learning_rate": 1.0736476608187136e-05,
+      "loss": 1.4309,
+      "step": 10138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8399806022644043,
+      "learning_rate": 1.0735562865497078e-05,
+      "loss": 1.6028,
+      "step": 10139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.868878126144409,
+      "learning_rate": 1.0734649122807019e-05,
+      "loss": 1.8155,
+      "step": 10140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.811234712600708,
+      "learning_rate": 1.0733735380116961e-05,
+      "loss": 1.4713,
+      "step": 10141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2969441413879395,
+      "learning_rate": 1.0732821637426902e-05,
+      "loss": 1.7636,
+      "step": 10142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9922940731048584,
+      "learning_rate": 1.0731907894736844e-05,
+      "loss": 1.9614,
+      "step": 10143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.595014810562134,
+      "learning_rate": 1.0730994152046786e-05,
+      "loss": 1.5492,
+      "step": 10144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.20758318901062,
+      "learning_rate": 1.0730080409356727e-05,
+      "loss": 1.864,
+      "step": 10145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8395674228668213,
+      "learning_rate": 1.0729166666666669e-05,
+      "loss": 1.8112,
+      "step": 10146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9231581687927246,
+      "learning_rate": 1.072825292397661e-05,
+      "loss": 1.6633,
+      "step": 10147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7147905826568604,
+      "learning_rate": 1.0727339181286552e-05,
+      "loss": 1.4152,
+      "step": 10148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5404441356658936,
+      "learning_rate": 1.0726425438596492e-05,
+      "loss": 1.5797,
+      "step": 10149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8404948711395264,
+      "learning_rate": 1.0725511695906434e-05,
+      "loss": 2.013,
+      "step": 10150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5931167602539062,
+      "learning_rate": 1.0724597953216377e-05,
+      "loss": 1.5744,
+      "step": 10151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2063848972320557,
+      "learning_rate": 1.0723684210526317e-05,
+      "loss": 1.6245,
+      "step": 10152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1669721603393555,
+      "learning_rate": 1.072277046783626e-05,
+      "loss": 1.2863,
+      "step": 10153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4545278549194336,
+      "learning_rate": 1.07218567251462e-05,
+      "loss": 1.3346,
+      "step": 10154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.81925368309021,
+      "learning_rate": 1.0720942982456142e-05,
+      "loss": 1.8178,
+      "step": 10155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.946624755859375,
+      "learning_rate": 1.0720029239766083e-05,
+      "loss": 1.4904,
+      "step": 10156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5575814247131348,
+      "learning_rate": 1.0719115497076025e-05,
+      "loss": 1.3709,
+      "step": 10157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.896972417831421,
+      "learning_rate": 1.0718201754385967e-05,
+      "loss": 1.6555,
+      "step": 10158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6877927780151367,
+      "learning_rate": 1.0717288011695908e-05,
+      "loss": 1.5027,
+      "step": 10159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7796568870544434,
+      "learning_rate": 1.071637426900585e-05,
+      "loss": 1.7063,
+      "step": 10160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7380218505859375,
+      "learning_rate": 1.071546052631579e-05,
+      "loss": 1.7343,
+      "step": 10161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3783583641052246,
+      "learning_rate": 1.0714546783625733e-05,
+      "loss": 1.5343,
+      "step": 10162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.455000400543213,
+      "learning_rate": 1.0713633040935673e-05,
+      "loss": 1.4457,
+      "step": 10163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.505979537963867,
+      "learning_rate": 1.0712719298245615e-05,
+      "loss": 1.6803,
+      "step": 10164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8517017364501953,
+      "learning_rate": 1.0711805555555557e-05,
+      "loss": 1.3484,
+      "step": 10165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.687098503112793,
+      "learning_rate": 1.0710891812865498e-05,
+      "loss": 1.5838,
+      "step": 10166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8379833698272705,
+      "learning_rate": 1.070997807017544e-05,
+      "loss": 1.1556,
+      "step": 10167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.703181266784668,
+      "learning_rate": 1.070906432748538e-05,
+      "loss": 1.6191,
+      "step": 10168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2659690380096436,
+      "learning_rate": 1.0708150584795323e-05,
+      "loss": 1.3591,
+      "step": 10169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.196148157119751,
+      "learning_rate": 1.0707236842105264e-05,
+      "loss": 1.1948,
+      "step": 10170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0096547603607178,
+      "learning_rate": 1.0706323099415206e-05,
+      "loss": 1.5541,
+      "step": 10171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.95538330078125,
+      "learning_rate": 1.0705409356725148e-05,
+      "loss": 1.6169,
+      "step": 10172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4105706214904785,
+      "learning_rate": 1.0704495614035088e-05,
+      "loss": 1.4317,
+      "step": 10173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5878143310546875,
+      "learning_rate": 1.070358187134503e-05,
+      "loss": 1.6148,
+      "step": 10174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.583611011505127,
+      "learning_rate": 1.0702668128654971e-05,
+      "loss": 1.388,
+      "step": 10175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6080217361450195,
+      "learning_rate": 1.0701754385964913e-05,
+      "loss": 1.7981,
+      "step": 10176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.45131516456604,
+      "learning_rate": 1.0700840643274854e-05,
+      "loss": 1.3401,
+      "step": 10177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8887765407562256,
+      "learning_rate": 1.0699926900584796e-05,
+      "loss": 1.3692,
+      "step": 10178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9206197261810303,
+      "learning_rate": 1.0699013157894738e-05,
+      "loss": 1.6501,
+      "step": 10179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.176806688308716,
+      "learning_rate": 1.0698099415204679e-05,
+      "loss": 1.2288,
+      "step": 10180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3626668453216553,
+      "learning_rate": 1.0697185672514621e-05,
+      "loss": 1.2767,
+      "step": 10181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9425671100616455,
+      "learning_rate": 1.0696271929824562e-05,
+      "loss": 1.7079,
+      "step": 10182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0652048587799072,
+      "learning_rate": 1.0695358187134504e-05,
+      "loss": 1.6577,
+      "step": 10183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.723869562149048,
+      "learning_rate": 1.0694444444444444e-05,
+      "loss": 1.3163,
+      "step": 10184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3768513202667236,
+      "learning_rate": 1.0693530701754387e-05,
+      "loss": 1.2865,
+      "step": 10185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4129254817962646,
+      "learning_rate": 1.0692616959064329e-05,
+      "loss": 1.285,
+      "step": 10186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.780907154083252,
+      "learning_rate": 1.069170321637427e-05,
+      "loss": 1.6477,
+      "step": 10187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4289395809173584,
+      "learning_rate": 1.0690789473684212e-05,
+      "loss": 1.4049,
+      "step": 10188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.968845844268799,
+      "learning_rate": 1.0689875730994152e-05,
+      "loss": 1.5715,
+      "step": 10189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5559027194976807,
+      "learning_rate": 1.0688961988304094e-05,
+      "loss": 1.8134,
+      "step": 10190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0342652797698975,
+      "learning_rate": 1.0688048245614035e-05,
+      "loss": 1.807,
+      "step": 10191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3155112266540527,
+      "learning_rate": 1.0687134502923977e-05,
+      "loss": 1.329,
+      "step": 10192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.797358751296997,
+      "learning_rate": 1.068622076023392e-05,
+      "loss": 1.3612,
+      "step": 10193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8236594200134277,
+      "learning_rate": 1.068530701754386e-05,
+      "loss": 1.8308,
+      "step": 10194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.723693609237671,
+      "learning_rate": 1.0684393274853802e-05,
+      "loss": 1.3968,
+      "step": 10195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0321426391601562,
+      "learning_rate": 1.0683479532163743e-05,
+      "loss": 1.747,
+      "step": 10196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6968002319335938,
+      "learning_rate": 1.0682565789473685e-05,
+      "loss": 1.5247,
+      "step": 10197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.567272901535034,
+      "learning_rate": 1.0681652046783625e-05,
+      "loss": 1.2086,
+      "step": 10198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4077913761138916,
+      "learning_rate": 1.0680738304093568e-05,
+      "loss": 1.4614,
+      "step": 10199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5798048973083496,
+      "learning_rate": 1.067982456140351e-05,
+      "loss": 1.4607,
+      "step": 10200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.82039737701416,
+      "learning_rate": 1.067891081871345e-05,
+      "loss": 1.5435,
+      "step": 10201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7002217769622803,
+      "learning_rate": 1.0677997076023393e-05,
+      "loss": 1.5871,
+      "step": 10202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.638329029083252,
+      "learning_rate": 1.0677083333333333e-05,
+      "loss": 1.5642,
+      "step": 10203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3289313316345215,
+      "learning_rate": 1.0676169590643275e-05,
+      "loss": 1.7074,
+      "step": 10204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4219837188720703,
+      "learning_rate": 1.0675255847953216e-05,
+      "loss": 1.2092,
+      "step": 10205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4153945446014404,
+      "learning_rate": 1.0674342105263158e-05,
+      "loss": 1.3185,
+      "step": 10206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4891388416290283,
+      "learning_rate": 1.0673428362573099e-05,
+      "loss": 1.3785,
+      "step": 10207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3092923164367676,
+      "learning_rate": 1.067251461988304e-05,
+      "loss": 1.2057,
+      "step": 10208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0975239276885986,
+      "learning_rate": 1.0671600877192983e-05,
+      "loss": 1.5286,
+      "step": 10209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.574537515640259,
+      "learning_rate": 1.0670687134502924e-05,
+      "loss": 1.5498,
+      "step": 10210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9681141376495361,
+      "learning_rate": 1.0669773391812866e-05,
+      "loss": 1.1762,
+      "step": 10211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5229156017303467,
+      "learning_rate": 1.0668859649122806e-05,
+      "loss": 1.4323,
+      "step": 10212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.922825813293457,
+      "learning_rate": 1.0667945906432749e-05,
+      "loss": 1.2738,
+      "step": 10213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5503251552581787,
+      "learning_rate": 1.0667032163742689e-05,
+      "loss": 1.5332,
+      "step": 10214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.775515556335449,
+      "learning_rate": 1.0666118421052631e-05,
+      "loss": 1.683,
+      "step": 10215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6171298027038574,
+      "learning_rate": 1.0665204678362574e-05,
+      "loss": 1.6604,
+      "step": 10216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6568214893341064,
+      "learning_rate": 1.0664290935672514e-05,
+      "loss": 1.4936,
+      "step": 10217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2519097328186035,
+      "learning_rate": 1.0663377192982458e-05,
+      "loss": 1.3157,
+      "step": 10218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5102508068084717,
+      "learning_rate": 1.06624634502924e-05,
+      "loss": 1.3105,
+      "step": 10219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1296164989471436,
+      "learning_rate": 1.066154970760234e-05,
+      "loss": 1.2805,
+      "step": 10220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1262621879577637,
+      "learning_rate": 1.0660635964912283e-05,
+      "loss": 1.6885,
+      "step": 10221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7422051429748535,
+      "learning_rate": 1.0659722222222223e-05,
+      "loss": 1.7552,
+      "step": 10222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5356435775756836,
+      "learning_rate": 1.0658808479532166e-05,
+      "loss": 1.7064,
+      "step": 10223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7428948879241943,
+      "learning_rate": 1.0657894736842108e-05,
+      "loss": 1.5023,
+      "step": 10224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3167428970336914,
+      "learning_rate": 1.0656980994152048e-05,
+      "loss": 1.3111,
+      "step": 10225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.276918411254883,
+      "learning_rate": 1.065606725146199e-05,
+      "loss": 1.2378,
+      "step": 10226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5398521423339844,
+      "learning_rate": 1.0655153508771931e-05,
+      "loss": 1.5737,
+      "step": 10227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.076207160949707,
+      "learning_rate": 1.0654239766081873e-05,
+      "loss": 1.5706,
+      "step": 10228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.180757761001587,
+      "learning_rate": 1.0653326023391814e-05,
+      "loss": 1.2264,
+      "step": 10229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8219006061553955,
+      "learning_rate": 1.0652412280701756e-05,
+      "loss": 1.4957,
+      "step": 10230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8552520275115967,
+      "learning_rate": 1.0651498538011698e-05,
+      "loss": 1.5997,
+      "step": 10231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3682916164398193,
+      "learning_rate": 1.0650584795321639e-05,
+      "loss": 1.2284,
+      "step": 10232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.063552141189575,
+      "learning_rate": 1.0649671052631581e-05,
+      "loss": 1.5116,
+      "step": 10233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0033318996429443,
+      "learning_rate": 1.0648757309941522e-05,
+      "loss": 1.6568,
+      "step": 10234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.086516857147217,
+      "learning_rate": 1.0647843567251464e-05,
+      "loss": 1.6118,
+      "step": 10235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2907071113586426,
+      "learning_rate": 1.0646929824561404e-05,
+      "loss": 1.9381,
+      "step": 10236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3983774185180664,
+      "learning_rate": 1.0646016081871347e-05,
+      "loss": 1.3709,
+      "step": 10237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.48416805267334,
+      "learning_rate": 1.0645102339181289e-05,
+      "loss": 1.2841,
+      "step": 10238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.725142240524292,
+      "learning_rate": 1.064418859649123e-05,
+      "loss": 1.6023,
+      "step": 10239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5772054195404053,
+      "learning_rate": 1.0643274853801172e-05,
+      "loss": 1.6313,
+      "step": 10240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5807764530181885,
+      "learning_rate": 1.0642361111111112e-05,
+      "loss": 1.5867,
+      "step": 10241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9556453227996826,
+      "learning_rate": 1.0641447368421054e-05,
+      "loss": 1.8559,
+      "step": 10242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1362228393554688,
+      "learning_rate": 1.0640533625730995e-05,
+      "loss": 1.6758,
+      "step": 10243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.214769124984741,
+      "learning_rate": 1.0639619883040937e-05,
+      "loss": 1.212,
+      "step": 10244
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6674082279205322,
+      "learning_rate": 1.0638706140350878e-05,
+      "loss": 1.6378,
+      "step": 10245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9743492603302,
+      "learning_rate": 1.063779239766082e-05,
+      "loss": 1.6407,
+      "step": 10246
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.424614191055298,
+      "learning_rate": 1.0636878654970762e-05,
+      "loss": 1.2708,
+      "step": 10247
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9885618686676025,
+      "learning_rate": 1.0635964912280703e-05,
+      "loss": 1.4531,
+      "step": 10248
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.940586566925049,
+      "learning_rate": 1.0635051169590645e-05,
+      "loss": 1.5936,
+      "step": 10249
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4171016216278076,
+      "learning_rate": 1.0634137426900585e-05,
+      "loss": 1.1948,
+      "step": 10250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.248405933380127,
+      "learning_rate": 1.0633223684210528e-05,
+      "loss": 1.7746,
+      "step": 10251
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9949826002120972,
+      "learning_rate": 1.0632309941520468e-05,
+      "loss": 1.135,
+      "step": 10252
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7438745498657227,
+      "learning_rate": 1.063139619883041e-05,
+      "loss": 1.6761,
+      "step": 10253
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.536972999572754,
+      "learning_rate": 1.0630482456140353e-05,
+      "loss": 1.6897,
+      "step": 10254
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.225844383239746,
+      "learning_rate": 1.0629568713450293e-05,
+      "loss": 2.0824,
+      "step": 10255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9622693061828613,
+      "learning_rate": 1.0628654970760235e-05,
+      "loss": 1.7204,
+      "step": 10256
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7344160079956055,
+      "learning_rate": 1.0627741228070176e-05,
+      "loss": 1.4363,
+      "step": 10257
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5385067462921143,
+      "learning_rate": 1.0626827485380118e-05,
+      "loss": 1.7178,
+      "step": 10258
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7530486583709717,
+      "learning_rate": 1.0625913742690059e-05,
+      "loss": 1.3908,
+      "step": 10259
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7132110595703125,
+      "learning_rate": 1.0625e-05,
+      "loss": 1.5526,
+      "step": 10260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9023923873901367,
+      "learning_rate": 1.0624086257309943e-05,
+      "loss": 1.7047,
+      "step": 10261
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.46222186088562,
+      "learning_rate": 1.0623172514619884e-05,
+      "loss": 1.493,
+      "step": 10262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4351985454559326,
+      "learning_rate": 1.0622258771929826e-05,
+      "loss": 1.3851,
+      "step": 10263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6057815551757812,
+      "learning_rate": 1.0621345029239766e-05,
+      "loss": 1.4172,
+      "step": 10264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0797064304351807,
+      "learning_rate": 1.0620431286549708e-05,
+      "loss": 1.8615,
+      "step": 10265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.413992166519165,
+      "learning_rate": 1.0619517543859649e-05,
+      "loss": 1.5471,
+      "step": 10266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1648895740509033,
+      "learning_rate": 1.0618603801169591e-05,
+      "loss": 1.5126,
+      "step": 10267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.314656972885132,
+      "learning_rate": 1.0617690058479533e-05,
+      "loss": 1.437,
+      "step": 10268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2552084922790527,
+      "learning_rate": 1.0616776315789474e-05,
+      "loss": 1.5862,
+      "step": 10269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.651075601577759,
+      "learning_rate": 1.0615862573099416e-05,
+      "loss": 1.6389,
+      "step": 10270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2044477462768555,
+      "learning_rate": 1.0614948830409357e-05,
+      "loss": 1.0599,
+      "step": 10271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2428462505340576,
+      "learning_rate": 1.0614035087719299e-05,
+      "loss": 1.1985,
+      "step": 10272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.329155683517456,
+      "learning_rate": 1.061312134502924e-05,
+      "loss": 1.3805,
+      "step": 10273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6521201133728027,
+      "learning_rate": 1.0612207602339182e-05,
+      "loss": 1.5361,
+      "step": 10274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.250166893005371,
+      "learning_rate": 1.0611293859649124e-05,
+      "loss": 1.3325,
+      "step": 10275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5410683155059814,
+      "learning_rate": 1.0610380116959064e-05,
+      "loss": 1.4001,
+      "step": 10276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.565824508666992,
+      "learning_rate": 1.0609466374269007e-05,
+      "loss": 1.3204,
+      "step": 10277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.651179552078247,
+      "learning_rate": 1.0608552631578947e-05,
+      "loss": 1.6843,
+      "step": 10278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5117034912109375,
+      "learning_rate": 1.060763888888889e-05,
+      "loss": 1.5818,
+      "step": 10279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4292149543762207,
+      "learning_rate": 1.060672514619883e-05,
+      "loss": 1.897,
+      "step": 10280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.937344551086426,
+      "learning_rate": 1.0605811403508772e-05,
+      "loss": 1.6269,
+      "step": 10281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.863178014755249,
+      "learning_rate": 1.0604897660818714e-05,
+      "loss": 1.746,
+      "step": 10282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.936270236968994,
+      "learning_rate": 1.0603983918128655e-05,
+      "loss": 1.7299,
+      "step": 10283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9579718112945557,
+      "learning_rate": 1.0603070175438597e-05,
+      "loss": 1.6273,
+      "step": 10284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0043704509735107,
+      "learning_rate": 1.0602156432748538e-05,
+      "loss": 1.601,
+      "step": 10285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7788548469543457,
+      "learning_rate": 1.060124269005848e-05,
+      "loss": 1.5607,
+      "step": 10286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.723848342895508,
+      "learning_rate": 1.060032894736842e-05,
+      "loss": 1.4924,
+      "step": 10287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5565195083618164,
+      "learning_rate": 1.0599415204678363e-05,
+      "loss": 1.4354,
+      "step": 10288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.787081480026245,
+      "learning_rate": 1.0598501461988305e-05,
+      "loss": 1.8268,
+      "step": 10289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.510960102081299,
+      "learning_rate": 1.0597587719298245e-05,
+      "loss": 1.4065,
+      "step": 10290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0202796459198,
+      "learning_rate": 1.0596673976608188e-05,
+      "loss": 1.0173,
+      "step": 10291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9019687175750732,
+      "learning_rate": 1.0595760233918128e-05,
+      "loss": 1.757,
+      "step": 10292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1200151443481445,
+      "learning_rate": 1.059484649122807e-05,
+      "loss": 1.5654,
+      "step": 10293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0972702503204346,
+      "learning_rate": 1.0593932748538011e-05,
+      "loss": 1.6895,
+      "step": 10294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6690101623535156,
+      "learning_rate": 1.0593019005847953e-05,
+      "loss": 1.7036,
+      "step": 10295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5734663009643555,
+      "learning_rate": 1.0592105263157895e-05,
+      "loss": 1.2088,
+      "step": 10296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.076314687728882,
+      "learning_rate": 1.0591191520467836e-05,
+      "loss": 1.8031,
+      "step": 10297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.790383815765381,
+      "learning_rate": 1.0590277777777778e-05,
+      "loss": 1.5121,
+      "step": 10298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5300049781799316,
+      "learning_rate": 1.0589364035087719e-05,
+      "loss": 1.2052,
+      "step": 10299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.601468086242676,
+      "learning_rate": 1.058845029239766e-05,
+      "loss": 1.7681,
+      "step": 10300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.395267963409424,
+      "learning_rate": 1.0587536549707601e-05,
+      "loss": 1.5036,
+      "step": 10301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1340813636779785,
+      "learning_rate": 1.0586622807017544e-05,
+      "loss": 1.2814,
+      "step": 10302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6324918270111084,
+      "learning_rate": 1.0585709064327487e-05,
+      "loss": 1.2772,
+      "step": 10303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.418545961380005,
+      "learning_rate": 1.0584795321637428e-05,
+      "loss": 1.1649,
+      "step": 10304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0391807556152344,
+      "learning_rate": 1.058388157894737e-05,
+      "loss": 1.8404,
+      "step": 10305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.193071126937866,
+      "learning_rate": 1.0582967836257312e-05,
+      "loss": 1.658,
+      "step": 10306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4996654987335205,
+      "learning_rate": 1.0582054093567253e-05,
+      "loss": 1.278,
+      "step": 10307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.801118850708008,
+      "learning_rate": 1.0581140350877195e-05,
+      "loss": 1.6732,
+      "step": 10308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.369262933731079,
+      "learning_rate": 1.0580226608187136e-05,
+      "loss": 1.3085,
+      "step": 10309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2358028888702393,
+      "learning_rate": 1.0579312865497078e-05,
+      "loss": 1.0915,
+      "step": 10310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1808910369873047,
+      "learning_rate": 1.0578399122807018e-05,
+      "loss": 1.4156,
+      "step": 10311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.393073797225952,
+      "learning_rate": 1.057748538011696e-05,
+      "loss": 1.5639,
+      "step": 10312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.397303581237793,
+      "learning_rate": 1.0576571637426903e-05,
+      "loss": 1.6566,
+      "step": 10313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9204652309417725,
+      "learning_rate": 1.0575657894736843e-05,
+      "loss": 1.7333,
+      "step": 10314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5161855220794678,
+      "learning_rate": 1.0574744152046786e-05,
+      "loss": 1.5645,
+      "step": 10315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5821127891540527,
+      "learning_rate": 1.0573830409356726e-05,
+      "loss": 1.5533,
+      "step": 10316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2198486328125,
+      "learning_rate": 1.0572916666666668e-05,
+      "loss": 1.782,
+      "step": 10317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2638661861419678,
+      "learning_rate": 1.0572002923976609e-05,
+      "loss": 1.2937,
+      "step": 10318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0231122970581055,
+      "learning_rate": 1.0571089181286551e-05,
+      "loss": 1.9362,
+      "step": 10319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8825109004974365,
+      "learning_rate": 1.0570175438596493e-05,
+      "loss": 1.3756,
+      "step": 10320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.547267198562622,
+      "learning_rate": 1.0569261695906434e-05,
+      "loss": 1.3193,
+      "step": 10321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.661339521408081,
+      "learning_rate": 1.0568347953216376e-05,
+      "loss": 1.6453,
+      "step": 10322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.701449394226074,
+      "learning_rate": 1.0567434210526317e-05,
+      "loss": 1.6543,
+      "step": 10323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.401703357696533,
+      "learning_rate": 1.0566520467836259e-05,
+      "loss": 1.3452,
+      "step": 10324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.883985757827759,
+      "learning_rate": 1.05656067251462e-05,
+      "loss": 1.8046,
+      "step": 10325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6298320293426514,
+      "learning_rate": 1.0564692982456142e-05,
+      "loss": 1.4557,
+      "step": 10326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5992190837860107,
+      "learning_rate": 1.0563779239766084e-05,
+      "loss": 1.4407,
+      "step": 10327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.465674638748169,
+      "learning_rate": 1.0562865497076024e-05,
+      "loss": 1.5774,
+      "step": 10328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6116085052490234,
+      "learning_rate": 1.0561951754385967e-05,
+      "loss": 1.4871,
+      "step": 10329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6924397945404053,
+      "learning_rate": 1.0561038011695907e-05,
+      "loss": 1.522,
+      "step": 10330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8582592010498047,
+      "learning_rate": 1.056012426900585e-05,
+      "loss": 1.9468,
+      "step": 10331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.156888246536255,
+      "learning_rate": 1.055921052631579e-05,
+      "loss": 1.2735,
+      "step": 10332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8948912620544434,
+      "learning_rate": 1.0558296783625732e-05,
+      "loss": 1.5457,
+      "step": 10333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.23486590385437,
+      "learning_rate": 1.0557383040935674e-05,
+      "loss": 1.2044,
+      "step": 10334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.300158977508545,
+      "learning_rate": 1.0556469298245615e-05,
+      "loss": 1.1326,
+      "step": 10335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.439415454864502,
+      "learning_rate": 1.0555555555555557e-05,
+      "loss": 1.4067,
+      "step": 10336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.6086883544921875,
+      "learning_rate": 1.0554641812865498e-05,
+      "loss": 2.1518,
+      "step": 10337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.347691059112549,
+      "learning_rate": 1.055372807017544e-05,
+      "loss": 1.5858,
+      "step": 10338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5371720790863037,
+      "learning_rate": 1.055281432748538e-05,
+      "loss": 1.5579,
+      "step": 10339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.851198673248291,
+      "learning_rate": 1.0551900584795323e-05,
+      "loss": 1.7818,
+      "step": 10340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.710294246673584,
+      "learning_rate": 1.0550986842105265e-05,
+      "loss": 1.3119,
+      "step": 10341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5567166805267334,
+      "learning_rate": 1.0550073099415205e-05,
+      "loss": 1.42,
+      "step": 10342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3897271156311035,
+      "learning_rate": 1.0549159356725148e-05,
+      "loss": 1.4501,
+      "step": 10343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7067151069641113,
+      "learning_rate": 1.0548245614035088e-05,
+      "loss": 1.4095,
+      "step": 10344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.562727451324463,
+      "learning_rate": 1.054733187134503e-05,
+      "loss": 1.8121,
+      "step": 10345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0253801345825195,
+      "learning_rate": 1.054641812865497e-05,
+      "loss": 1.511,
+      "step": 10346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.376800775527954,
+      "learning_rate": 1.0545504385964913e-05,
+      "loss": 1.3086,
+      "step": 10347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7614388465881348,
+      "learning_rate": 1.0544590643274855e-05,
+      "loss": 1.573,
+      "step": 10348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8297119140625,
+      "learning_rate": 1.0543676900584796e-05,
+      "loss": 1.7523,
+      "step": 10349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.159538507461548,
+      "learning_rate": 1.0542763157894738e-05,
+      "loss": 1.264,
+      "step": 10350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.807015895843506,
+      "learning_rate": 1.0541849415204679e-05,
+      "loss": 1.5474,
+      "step": 10351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1737098693847656,
+      "learning_rate": 1.054093567251462e-05,
+      "loss": 1.3454,
+      "step": 10352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8542892932891846,
+      "learning_rate": 1.0540021929824561e-05,
+      "loss": 2.1189,
+      "step": 10353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1304032802581787,
+      "learning_rate": 1.0539108187134504e-05,
+      "loss": 1.1699,
+      "step": 10354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8920297622680664,
+      "learning_rate": 1.0538194444444446e-05,
+      "loss": 1.6151,
+      "step": 10355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9017174243927,
+      "learning_rate": 1.0537280701754386e-05,
+      "loss": 1.6921,
+      "step": 10356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3409602642059326,
+      "learning_rate": 1.0536366959064328e-05,
+      "loss": 1.0155,
+      "step": 10357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.143871545791626,
+      "learning_rate": 1.0535453216374269e-05,
+      "loss": 1.2466,
+      "step": 10358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4510602951049805,
+      "learning_rate": 1.0534539473684211e-05,
+      "loss": 2.1525,
+      "step": 10359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5908491611480713,
+      "learning_rate": 1.0533625730994152e-05,
+      "loss": 1.4833,
+      "step": 10360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5186057090759277,
+      "learning_rate": 1.0532711988304094e-05,
+      "loss": 1.4685,
+      "step": 10361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.235441207885742,
+      "learning_rate": 1.0531798245614036e-05,
+      "loss": 1.24,
+      "step": 10362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7390904426574707,
+      "learning_rate": 1.0530884502923977e-05,
+      "loss": 1.6757,
+      "step": 10363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8263392448425293,
+      "learning_rate": 1.0529970760233919e-05,
+      "loss": 1.8305,
+      "step": 10364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.699988603591919,
+      "learning_rate": 1.052905701754386e-05,
+      "loss": 1.683,
+      "step": 10365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3114845752716064,
+      "learning_rate": 1.0528143274853802e-05,
+      "loss": 1.3811,
+      "step": 10366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7320001125335693,
+      "learning_rate": 1.0527229532163742e-05,
+      "loss": 1.3655,
+      "step": 10367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4886221885681152,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 1.3653,
+      "step": 10368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2632923126220703,
+      "learning_rate": 1.0525402046783627e-05,
+      "loss": 1.4246,
+      "step": 10369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.504225730895996,
+      "learning_rate": 1.0524488304093567e-05,
+      "loss": 1.5528,
+      "step": 10370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0735459327697754,
+      "learning_rate": 1.052357456140351e-05,
+      "loss": 1.8029,
+      "step": 10371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6826300621032715,
+      "learning_rate": 1.052266081871345e-05,
+      "loss": 1.8055,
+      "step": 10372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.963441848754883,
+      "learning_rate": 1.0521747076023392e-05,
+      "loss": 2.043,
+      "step": 10373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7275936603546143,
+      "learning_rate": 1.0520833333333333e-05,
+      "loss": 1.5257,
+      "step": 10374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.451078414916992,
+      "learning_rate": 1.0519919590643275e-05,
+      "loss": 1.6473,
+      "step": 10375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9780547618865967,
+      "learning_rate": 1.0519005847953217e-05,
+      "loss": 1.7534,
+      "step": 10376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7940917015075684,
+      "learning_rate": 1.0518092105263158e-05,
+      "loss": 1.5038,
+      "step": 10377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4928793907165527,
+      "learning_rate": 1.05171783625731e-05,
+      "loss": 1.3364,
+      "step": 10378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.042435646057129,
+      "learning_rate": 1.051626461988304e-05,
+      "loss": 1.6975,
+      "step": 10379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.549494504928589,
+      "learning_rate": 1.0515350877192983e-05,
+      "loss": 1.4834,
+      "step": 10380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.637944221496582,
+      "learning_rate": 1.0514437134502923e-05,
+      "loss": 1.6977,
+      "step": 10381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5712335109710693,
+      "learning_rate": 1.0513523391812865e-05,
+      "loss": 1.3571,
+      "step": 10382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.371584415435791,
+      "learning_rate": 1.0512609649122808e-05,
+      "loss": 1.3905,
+      "step": 10383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0318737030029297,
+      "learning_rate": 1.0511695906432748e-05,
+      "loss": 1.8212,
+      "step": 10384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5616672039031982,
+      "learning_rate": 1.051078216374269e-05,
+      "loss": 1.5036,
+      "step": 10385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.82065749168396,
+      "learning_rate": 1.0509868421052631e-05,
+      "loss": 1.4708,
+      "step": 10386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4870662689208984,
+      "learning_rate": 1.0508954678362573e-05,
+      "loss": 1.5275,
+      "step": 10387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.918459177017212,
+      "learning_rate": 1.0508040935672514e-05,
+      "loss": 1.5669,
+      "step": 10388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7111151218414307,
+      "learning_rate": 1.0507127192982458e-05,
+      "loss": 1.5627,
+      "step": 10389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9110710620880127,
+      "learning_rate": 1.05062134502924e-05,
+      "loss": 1.5607,
+      "step": 10390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5651443004608154,
+      "learning_rate": 1.050529970760234e-05,
+      "loss": 1.4057,
+      "step": 10391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3472557067871094,
+      "learning_rate": 1.0504385964912283e-05,
+      "loss": 1.4091,
+      "step": 10392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8652641773223877,
+      "learning_rate": 1.0503472222222225e-05,
+      "loss": 1.4177,
+      "step": 10393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.499532461166382,
+      "learning_rate": 1.0502558479532165e-05,
+      "loss": 1.5941,
+      "step": 10394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5236151218414307,
+      "learning_rate": 1.0501644736842107e-05,
+      "loss": 1.1996,
+      "step": 10395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7286839485168457,
+      "learning_rate": 1.0500730994152048e-05,
+      "loss": 1.3997,
+      "step": 10396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.628154993057251,
+      "learning_rate": 1.049981725146199e-05,
+      "loss": 1.4789,
+      "step": 10397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8860507011413574,
+      "learning_rate": 1.049890350877193e-05,
+      "loss": 1.7493,
+      "step": 10398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.026919364929199,
+      "learning_rate": 1.0497989766081873e-05,
+      "loss": 1.7284,
+      "step": 10399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.487351179122925,
+      "learning_rate": 1.0497076023391815e-05,
+      "loss": 1.4046,
+      "step": 10400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7738394737243652,
+      "learning_rate": 1.0496162280701756e-05,
+      "loss": 1.8819,
+      "step": 10401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2586944103240967,
+      "learning_rate": 1.0495248538011698e-05,
+      "loss": 1.3841,
+      "step": 10402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4731862545013428,
+      "learning_rate": 1.0494334795321638e-05,
+      "loss": 1.3729,
+      "step": 10403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5029361248016357,
+      "learning_rate": 1.049342105263158e-05,
+      "loss": 1.5178,
+      "step": 10404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.252671003341675,
+      "learning_rate": 1.0492507309941521e-05,
+      "loss": 1.4796,
+      "step": 10405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5557384490966797,
+      "learning_rate": 1.0491593567251463e-05,
+      "loss": 1.6591,
+      "step": 10406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.130510330200195,
+      "learning_rate": 1.0490679824561406e-05,
+      "loss": 1.6493,
+      "step": 10407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.755817413330078,
+      "learning_rate": 1.0489766081871346e-05,
+      "loss": 1.7754,
+      "step": 10408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5831236839294434,
+      "learning_rate": 1.0488852339181288e-05,
+      "loss": 2.3854,
+      "step": 10409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1624808311462402,
+      "learning_rate": 1.0487938596491229e-05,
+      "loss": 1.6358,
+      "step": 10410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2287774085998535,
+      "learning_rate": 1.0487024853801171e-05,
+      "loss": 1.6834,
+      "step": 10411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9134669303894043,
+      "learning_rate": 1.0486111111111112e-05,
+      "loss": 1.8434,
+      "step": 10412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5728116035461426,
+      "learning_rate": 1.0485197368421054e-05,
+      "loss": 1.3973,
+      "step": 10413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8122949600219727,
+      "learning_rate": 1.0484283625730996e-05,
+      "loss": 1.3548,
+      "step": 10414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1736228466033936,
+      "learning_rate": 1.0483369883040937e-05,
+      "loss": 1.4345,
+      "step": 10415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.686706304550171,
+      "learning_rate": 1.0482456140350879e-05,
+      "loss": 2.1707,
+      "step": 10416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1663670539855957,
+      "learning_rate": 1.048154239766082e-05,
+      "loss": 1.4525,
+      "step": 10417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.737938642501831,
+      "learning_rate": 1.0480628654970762e-05,
+      "loss": 1.3553,
+      "step": 10418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9296393394470215,
+      "learning_rate": 1.0479714912280702e-05,
+      "loss": 1.8301,
+      "step": 10419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4537181854248047,
+      "learning_rate": 1.0478801169590644e-05,
+      "loss": 1.2864,
+      "step": 10420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4963996410369873,
+      "learning_rate": 1.0477887426900587e-05,
+      "loss": 1.5216,
+      "step": 10421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.332458257675171,
+      "learning_rate": 1.0476973684210527e-05,
+      "loss": 1.3563,
+      "step": 10422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.362572193145752,
+      "learning_rate": 1.047605994152047e-05,
+      "loss": 1.2009,
+      "step": 10423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.566633462905884,
+      "learning_rate": 1.047514619883041e-05,
+      "loss": 1.3758,
+      "step": 10424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3757076263427734,
+      "learning_rate": 1.0474232456140352e-05,
+      "loss": 1.9645,
+      "step": 10425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.60992431640625,
+      "learning_rate": 1.0473318713450293e-05,
+      "loss": 1.7481,
+      "step": 10426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.627185344696045,
+      "learning_rate": 1.0472404970760235e-05,
+      "loss": 1.6078,
+      "step": 10427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.75541090965271,
+      "learning_rate": 1.0471491228070177e-05,
+      "loss": 1.4739,
+      "step": 10428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9967164993286133,
+      "learning_rate": 1.0470577485380118e-05,
+      "loss": 1.6988,
+      "step": 10429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4214673042297363,
+      "learning_rate": 1.046966374269006e-05,
+      "loss": 1.6102,
+      "step": 10430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5005412101745605,
+      "learning_rate": 1.046875e-05,
+      "loss": 1.5618,
+      "step": 10431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5876622200012207,
+      "learning_rate": 1.0467836257309943e-05,
+      "loss": 1.6162,
+      "step": 10432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5075488090515137,
+      "learning_rate": 1.0466922514619883e-05,
+      "loss": 1.5308,
+      "step": 10433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3200037479400635,
+      "learning_rate": 1.0466008771929825e-05,
+      "loss": 1.355,
+      "step": 10434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4612786769866943,
+      "learning_rate": 1.0465095029239768e-05,
+      "loss": 1.3972,
+      "step": 10435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0860631465911865,
+      "learning_rate": 1.0464181286549708e-05,
+      "loss": 1.5347,
+      "step": 10436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9200496673583984,
+      "learning_rate": 1.046326754385965e-05,
+      "loss": 1.7838,
+      "step": 10437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1385910511016846,
+      "learning_rate": 1.046235380116959e-05,
+      "loss": 1.3545,
+      "step": 10438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.564622402191162,
+      "learning_rate": 1.0461440058479533e-05,
+      "loss": 1.4051,
+      "step": 10439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0249576568603516,
+      "learning_rate": 1.0460526315789474e-05,
+      "loss": 1.9659,
+      "step": 10440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9736711978912354,
+      "learning_rate": 1.0459612573099416e-05,
+      "loss": 1.4694,
+      "step": 10441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.76222562789917,
+      "learning_rate": 1.0458698830409358e-05,
+      "loss": 1.6705,
+      "step": 10442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.928972005844116,
+      "learning_rate": 1.0457785087719299e-05,
+      "loss": 1.4086,
+      "step": 10443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6883959770202637,
+      "learning_rate": 1.045687134502924e-05,
+      "loss": 1.763,
+      "step": 10444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9597063064575195,
+      "learning_rate": 1.0455957602339181e-05,
+      "loss": 1.4144,
+      "step": 10445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6715595722198486,
+      "learning_rate": 1.0455043859649124e-05,
+      "loss": 1.5898,
+      "step": 10446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2539238929748535,
+      "learning_rate": 1.0454130116959064e-05,
+      "loss": 1.4724,
+      "step": 10447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5343477725982666,
+      "learning_rate": 1.0453216374269006e-05,
+      "loss": 1.4619,
+      "step": 10448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7534995079040527,
+      "learning_rate": 1.0452302631578948e-05,
+      "loss": 1.2813,
+      "step": 10449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9925577640533447,
+      "learning_rate": 1.0451388888888889e-05,
+      "loss": 1.7862,
+      "step": 10450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.295044422149658,
+      "learning_rate": 1.0450475146198831e-05,
+      "loss": 1.3377,
+      "step": 10451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7550742626190186,
+      "learning_rate": 1.0449561403508772e-05,
+      "loss": 1.5002,
+      "step": 10452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3027336597442627,
+      "learning_rate": 1.0448647660818714e-05,
+      "loss": 1.3659,
+      "step": 10453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.685680389404297,
+      "learning_rate": 1.0447733918128655e-05,
+      "loss": 1.8052,
+      "step": 10454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.558892011642456,
+      "learning_rate": 1.0446820175438597e-05,
+      "loss": 1.5444,
+      "step": 10455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.942335844039917,
+      "learning_rate": 1.0445906432748539e-05,
+      "loss": 1.6211,
+      "step": 10456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9195425510406494,
+      "learning_rate": 1.044499269005848e-05,
+      "loss": 1.639,
+      "step": 10457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7612316608428955,
+      "learning_rate": 1.0444078947368422e-05,
+      "loss": 1.6661,
+      "step": 10458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4606125354766846,
+      "learning_rate": 1.0443165204678362e-05,
+      "loss": 1.4183,
+      "step": 10459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.164947986602783,
+      "learning_rate": 1.0442251461988304e-05,
+      "loss": 1.9036,
+      "step": 10460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9560153484344482,
+      "learning_rate": 1.0441337719298245e-05,
+      "loss": 1.665,
+      "step": 10461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7129993438720703,
+      "learning_rate": 1.0440423976608187e-05,
+      "loss": 1.4825,
+      "step": 10462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6975202560424805,
+      "learning_rate": 1.043951023391813e-05,
+      "loss": 1.4885,
+      "step": 10463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3897945880889893,
+      "learning_rate": 1.043859649122807e-05,
+      "loss": 1.502,
+      "step": 10464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.32747220993042,
+      "learning_rate": 1.0437682748538012e-05,
+      "loss": 1.9096,
+      "step": 10465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9567196369171143,
+      "learning_rate": 1.0436769005847953e-05,
+      "loss": 1.5238,
+      "step": 10466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4589622020721436,
+      "learning_rate": 1.0435855263157895e-05,
+      "loss": 1.4601,
+      "step": 10467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8283679485321045,
+      "learning_rate": 1.0434941520467835e-05,
+      "loss": 1.9062,
+      "step": 10468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.463153600692749,
+      "learning_rate": 1.0434027777777778e-05,
+      "loss": 1.3531,
+      "step": 10469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7680065631866455,
+      "learning_rate": 1.043311403508772e-05,
+      "loss": 1.8271,
+      "step": 10470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3541319370269775,
+      "learning_rate": 1.043220029239766e-05,
+      "loss": 1.4067,
+      "step": 10471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4173617362976074,
+      "learning_rate": 1.0431286549707603e-05,
+      "loss": 1.4958,
+      "step": 10472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7422568798065186,
+      "learning_rate": 1.0430372807017543e-05,
+      "loss": 1.3807,
+      "step": 10473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.264557123184204,
+      "learning_rate": 1.0429459064327487e-05,
+      "loss": 1.3026,
+      "step": 10474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5080068111419678,
+      "learning_rate": 1.042854532163743e-05,
+      "loss": 1.4202,
+      "step": 10475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8688011169433594,
+      "learning_rate": 1.042763157894737e-05,
+      "loss": 1.4454,
+      "step": 10476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3638880252838135,
+      "learning_rate": 1.0426717836257312e-05,
+      "loss": 1.4009,
+      "step": 10477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0096426010131836,
+      "learning_rate": 1.0425804093567253e-05,
+      "loss": 1.2752,
+      "step": 10478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.537257671356201,
+      "learning_rate": 1.0424890350877195e-05,
+      "loss": 1.6445,
+      "step": 10479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.762422561645508,
+      "learning_rate": 1.0423976608187137e-05,
+      "loss": 1.6446,
+      "step": 10480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.400115489959717,
+      "learning_rate": 1.0423062865497078e-05,
+      "loss": 1.1497,
+      "step": 10481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.603257179260254,
+      "learning_rate": 1.042214912280702e-05,
+      "loss": 1.5959,
+      "step": 10482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4672489166259766,
+      "learning_rate": 1.042123538011696e-05,
+      "loss": 1.3325,
+      "step": 10483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1817209720611572,
+      "learning_rate": 1.0420321637426903e-05,
+      "loss": 1.1625,
+      "step": 10484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.568927526473999,
+      "learning_rate": 1.0419407894736843e-05,
+      "loss": 1.9392,
+      "step": 10485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.673917293548584,
+      "learning_rate": 1.0418494152046785e-05,
+      "loss": 1.3571,
+      "step": 10486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.448745012283325,
+      "learning_rate": 1.0417580409356727e-05,
+      "loss": 1.5645,
+      "step": 10487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6882669925689697,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 1.8041,
+      "step": 10488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.851785659790039,
+      "learning_rate": 1.041575292397661e-05,
+      "loss": 2.0603,
+      "step": 10489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.609815835952759,
+      "learning_rate": 1.041483918128655e-05,
+      "loss": 1.4397,
+      "step": 10490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6235785484313965,
+      "learning_rate": 1.0413925438596493e-05,
+      "loss": 1.6734,
+      "step": 10491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3720133304595947,
+      "learning_rate": 1.0413011695906434e-05,
+      "loss": 1.3285,
+      "step": 10492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2593629360198975,
+      "learning_rate": 1.0412097953216376e-05,
+      "loss": 1.4263,
+      "step": 10493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.583268165588379,
+      "learning_rate": 1.0411184210526318e-05,
+      "loss": 1.5517,
+      "step": 10494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9677751064300537,
+      "learning_rate": 1.0410270467836258e-05,
+      "loss": 1.2606,
+      "step": 10495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2049567699432373,
+      "learning_rate": 1.04093567251462e-05,
+      "loss": 1.1705,
+      "step": 10496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4676899909973145,
+      "learning_rate": 1.0408442982456141e-05,
+      "loss": 1.4859,
+      "step": 10497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5394484996795654,
+      "learning_rate": 1.0407529239766083e-05,
+      "loss": 1.5989,
+      "step": 10498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.00470232963562,
+      "learning_rate": 1.0406615497076024e-05,
+      "loss": 1.0369,
+      "step": 10499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0349206924438477,
+      "learning_rate": 1.0405701754385966e-05,
+      "loss": 1.8511,
+      "step": 10500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.497551202774048,
+      "learning_rate": 1.0404788011695908e-05,
+      "loss": 1.3622,
+      "step": 10501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.28851318359375,
+      "learning_rate": 1.0403874269005849e-05,
+      "loss": 2.2028,
+      "step": 10502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1661410331726074,
+      "learning_rate": 1.0402960526315791e-05,
+      "loss": 1.9673,
+      "step": 10503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.778843641281128,
+      "learning_rate": 1.0402046783625732e-05,
+      "loss": 1.5438,
+      "step": 10504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.636427402496338,
+      "learning_rate": 1.0401133040935674e-05,
+      "loss": 1.6372,
+      "step": 10505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.621990919113159,
+      "learning_rate": 1.0400219298245614e-05,
+      "loss": 1.6797,
+      "step": 10506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.059713840484619,
+      "learning_rate": 1.0399305555555557e-05,
+      "loss": 1.9634,
+      "step": 10507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7437329292297363,
+      "learning_rate": 1.0398391812865499e-05,
+      "loss": 1.8531,
+      "step": 10508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6171369552612305,
+      "learning_rate": 1.039747807017544e-05,
+      "loss": 1.6779,
+      "step": 10509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2112767696380615,
+      "learning_rate": 1.0396564327485382e-05,
+      "loss": 1.1361,
+      "step": 10510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.773066520690918,
+      "learning_rate": 1.0395650584795322e-05,
+      "loss": 1.5518,
+      "step": 10511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.472989559173584,
+      "learning_rate": 1.0394736842105264e-05,
+      "loss": 1.3328,
+      "step": 10512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.960076332092285,
+      "learning_rate": 1.0393823099415205e-05,
+      "loss": 1.7506,
+      "step": 10513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7134053707122803,
+      "learning_rate": 1.0392909356725147e-05,
+      "loss": 1.6401,
+      "step": 10514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7529098987579346,
+      "learning_rate": 1.039199561403509e-05,
+      "loss": 1.9024,
+      "step": 10515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4951798915863037,
+      "learning_rate": 1.039108187134503e-05,
+      "loss": 1.5662,
+      "step": 10516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5269768238067627,
+      "learning_rate": 1.0390168128654972e-05,
+      "loss": 1.2722,
+      "step": 10517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.354184150695801,
+      "learning_rate": 1.0389254385964913e-05,
+      "loss": 1.3703,
+      "step": 10518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.716263771057129,
+      "learning_rate": 1.0388340643274855e-05,
+      "loss": 1.4088,
+      "step": 10519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.002070903778076,
+      "learning_rate": 1.0387426900584795e-05,
+      "loss": 1.6912,
+      "step": 10520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2190909385681152,
+      "learning_rate": 1.0386513157894738e-05,
+      "loss": 1.8798,
+      "step": 10521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4525351524353027,
+      "learning_rate": 1.038559941520468e-05,
+      "loss": 1.5766,
+      "step": 10522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3334925174713135,
+      "learning_rate": 1.038468567251462e-05,
+      "loss": 1.5663,
+      "step": 10523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.956413507461548,
+      "learning_rate": 1.0383771929824563e-05,
+      "loss": 1.6412,
+      "step": 10524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2217295169830322,
+      "learning_rate": 1.0382858187134503e-05,
+      "loss": 1.1766,
+      "step": 10525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.355633497238159,
+      "learning_rate": 1.0381944444444445e-05,
+      "loss": 1.961,
+      "step": 10526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2967522144317627,
+      "learning_rate": 1.0381030701754386e-05,
+      "loss": 1.1412,
+      "step": 10527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8383960723876953,
+      "learning_rate": 1.0380116959064328e-05,
+      "loss": 1.4249,
+      "step": 10528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.662233829498291,
+      "learning_rate": 1.037920321637427e-05,
+      "loss": 1.3795,
+      "step": 10529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7236006259918213,
+      "learning_rate": 1.037828947368421e-05,
+      "loss": 1.7927,
+      "step": 10530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6232175827026367,
+      "learning_rate": 1.0377375730994153e-05,
+      "loss": 1.1553,
+      "step": 10531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.898808240890503,
+      "learning_rate": 1.0376461988304094e-05,
+      "loss": 1.6746,
+      "step": 10532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.211576223373413,
+      "learning_rate": 1.0375548245614036e-05,
+      "loss": 1.7372,
+      "step": 10533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6584670543670654,
+      "learning_rate": 1.0374634502923976e-05,
+      "loss": 2.0319,
+      "step": 10534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9674699306488037,
+      "learning_rate": 1.0373720760233919e-05,
+      "loss": 1.6585,
+      "step": 10535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.040907144546509,
+      "learning_rate": 1.037280701754386e-05,
+      "loss": 1.6953,
+      "step": 10536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5179996490478516,
+      "learning_rate": 1.0371893274853801e-05,
+      "loss": 1.3752,
+      "step": 10537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0919954776763916,
+      "learning_rate": 1.0370979532163744e-05,
+      "loss": 1.3358,
+      "step": 10538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7230465412139893,
+      "learning_rate": 1.0370065789473684e-05,
+      "loss": 1.4957,
+      "step": 10539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4461491107940674,
+      "learning_rate": 1.0369152046783626e-05,
+      "loss": 1.4253,
+      "step": 10540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.473493814468384,
+      "learning_rate": 1.0368238304093567e-05,
+      "loss": 1.5542,
+      "step": 10541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6164705753326416,
+      "learning_rate": 1.0367324561403509e-05,
+      "loss": 1.4003,
+      "step": 10542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9643120765686035,
+      "learning_rate": 1.0366410818713451e-05,
+      "loss": 1.6106,
+      "step": 10543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.880774974822998,
+      "learning_rate": 1.0365497076023392e-05,
+      "loss": 1.5622,
+      "step": 10544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6534483432769775,
+      "learning_rate": 1.0364583333333334e-05,
+      "loss": 1.5032,
+      "step": 10545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.772057056427002,
+      "learning_rate": 1.0363669590643275e-05,
+      "loss": 1.6866,
+      "step": 10546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1887106895446777,
+      "learning_rate": 1.0362755847953217e-05,
+      "loss": 1.6695,
+      "step": 10547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9593310356140137,
+      "learning_rate": 1.0361842105263157e-05,
+      "loss": 1.759,
+      "step": 10548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6909918785095215,
+      "learning_rate": 1.03609283625731e-05,
+      "loss": 1.5,
+      "step": 10549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.231048107147217,
+      "learning_rate": 1.0360014619883042e-05,
+      "loss": 1.267,
+      "step": 10550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8387327194213867,
+      "learning_rate": 1.0359100877192982e-05,
+      "loss": 1.525,
+      "step": 10551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.412994623184204,
+      "learning_rate": 1.0358187134502924e-05,
+      "loss": 1.3786,
+      "step": 10552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.511293888092041,
+      "learning_rate": 1.0357273391812865e-05,
+      "loss": 1.5961,
+      "step": 10553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2932932376861572,
+      "learning_rate": 1.0356359649122807e-05,
+      "loss": 1.375,
+      "step": 10554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.713688850402832,
+      "learning_rate": 1.0355445906432748e-05,
+      "loss": 1.649,
+      "step": 10555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2030417919158936,
+      "learning_rate": 1.035453216374269e-05,
+      "loss": 1.6534,
+      "step": 10556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.642514228820801,
+      "learning_rate": 1.0353618421052632e-05,
+      "loss": 1.4649,
+      "step": 10557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4946506023406982,
+      "learning_rate": 1.0352704678362573e-05,
+      "loss": 1.2698,
+      "step": 10558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.033408164978027,
+      "learning_rate": 1.0351790935672515e-05,
+      "loss": 1.795,
+      "step": 10559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1520907878875732,
+      "learning_rate": 1.0350877192982459e-05,
+      "loss": 1.8244,
+      "step": 10560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.576150894165039,
+      "learning_rate": 1.03499634502924e-05,
+      "loss": 1.8676,
+      "step": 10561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.426236152648926,
+      "learning_rate": 1.0349049707602342e-05,
+      "loss": 1.6,
+      "step": 10562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7522780895233154,
+      "learning_rate": 1.0348135964912282e-05,
+      "loss": 1.5839,
+      "step": 10563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5819501876831055,
+      "learning_rate": 1.0347222222222224e-05,
+      "loss": 1.4635,
+      "step": 10564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5596909523010254,
+      "learning_rate": 1.0346308479532165e-05,
+      "loss": 1.4257,
+      "step": 10565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.145019054412842,
+      "learning_rate": 1.0345394736842107e-05,
+      "loss": 1.3264,
+      "step": 10566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8389432430267334,
+      "learning_rate": 1.034448099415205e-05,
+      "loss": 1.7792,
+      "step": 10567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.279604911804199,
+      "learning_rate": 1.034356725146199e-05,
+      "loss": 1.408,
+      "step": 10568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5760371685028076,
+      "learning_rate": 1.0342653508771932e-05,
+      "loss": 1.4828,
+      "step": 10569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.123195171356201,
+      "learning_rate": 1.0341739766081873e-05,
+      "loss": 1.0211,
+      "step": 10570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.759169101715088,
+      "learning_rate": 1.0340826023391815e-05,
+      "loss": 1.591,
+      "step": 10571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2832391262054443,
+      "learning_rate": 1.0339912280701755e-05,
+      "loss": 1.6489,
+      "step": 10572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3637590408325195,
+      "learning_rate": 1.0338998538011698e-05,
+      "loss": 1.2944,
+      "step": 10573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.499990940093994,
+      "learning_rate": 1.033808479532164e-05,
+      "loss": 1.3972,
+      "step": 10574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.415781259536743,
+      "learning_rate": 1.033717105263158e-05,
+      "loss": 1.721,
+      "step": 10575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6145577430725098,
+      "learning_rate": 1.0336257309941523e-05,
+      "loss": 1.6088,
+      "step": 10576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.374695062637329,
+      "learning_rate": 1.0335343567251463e-05,
+      "loss": 1.5537,
+      "step": 10577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0825881958007812,
+      "learning_rate": 1.0334429824561405e-05,
+      "loss": 1.478,
+      "step": 10578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7312989234924316,
+      "learning_rate": 1.0333516081871346e-05,
+      "loss": 1.3258,
+      "step": 10579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7337756156921387,
+      "learning_rate": 1.0332602339181288e-05,
+      "loss": 1.4063,
+      "step": 10580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.541769504547119,
+      "learning_rate": 1.033168859649123e-05,
+      "loss": 1.4158,
+      "step": 10581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9968713521957397,
+      "learning_rate": 1.033077485380117e-05,
+      "loss": 1.5205,
+      "step": 10582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6060643196105957,
+      "learning_rate": 1.0329861111111113e-05,
+      "loss": 1.6046,
+      "step": 10583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.023831605911255,
+      "learning_rate": 1.0328947368421054e-05,
+      "loss": 1.8093,
+      "step": 10584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8116884231567383,
+      "learning_rate": 1.0328033625730996e-05,
+      "loss": 1.442,
+      "step": 10585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0486841201782227,
+      "learning_rate": 1.0327119883040936e-05,
+      "loss": 1.9635,
+      "step": 10586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.09814715385437,
+      "learning_rate": 1.0326206140350878e-05,
+      "loss": 1.7949,
+      "step": 10587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5548624992370605,
+      "learning_rate": 1.032529239766082e-05,
+      "loss": 1.2487,
+      "step": 10588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.125204086303711,
+      "learning_rate": 1.0324378654970761e-05,
+      "loss": 1.2951,
+      "step": 10589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.436394214630127,
+      "learning_rate": 1.0323464912280703e-05,
+      "loss": 1.0915,
+      "step": 10590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0087010860443115,
+      "learning_rate": 1.0322551169590644e-05,
+      "loss": 1.798,
+      "step": 10591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5811383724212646,
+      "learning_rate": 1.0321637426900586e-05,
+      "loss": 1.6193,
+      "step": 10592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.482121706008911,
+      "learning_rate": 1.0320723684210527e-05,
+      "loss": 1.3372,
+      "step": 10593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.533761739730835,
+      "learning_rate": 1.0319809941520469e-05,
+      "loss": 1.4223,
+      "step": 10594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.867583751678467,
+      "learning_rate": 1.0318896198830411e-05,
+      "loss": 1.6429,
+      "step": 10595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.014747381210327,
+      "learning_rate": 1.0317982456140352e-05,
+      "loss": 1.6316,
+      "step": 10596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.478015899658203,
+      "learning_rate": 1.0317068713450294e-05,
+      "loss": 1.6238,
+      "step": 10597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5112361907958984,
+      "learning_rate": 1.0316154970760234e-05,
+      "loss": 1.2875,
+      "step": 10598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.539311647415161,
+      "learning_rate": 1.0315241228070177e-05,
+      "loss": 1.3989,
+      "step": 10599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9722108840942383,
+      "learning_rate": 1.0314327485380117e-05,
+      "loss": 1.8768,
+      "step": 10600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6124513149261475,
+      "learning_rate": 1.031341374269006e-05,
+      "loss": 1.4361,
+      "step": 10601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7202916145324707,
+      "learning_rate": 1.0312500000000002e-05,
+      "loss": 1.417,
+      "step": 10602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6036813259124756,
+      "learning_rate": 1.0311586257309942e-05,
+      "loss": 1.5275,
+      "step": 10603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.241347312927246,
+      "learning_rate": 1.0310672514619884e-05,
+      "loss": 1.3877,
+      "step": 10604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4456145763397217,
+      "learning_rate": 1.0309758771929825e-05,
+      "loss": 1.4953,
+      "step": 10605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7429134845733643,
+      "learning_rate": 1.0308845029239767e-05,
+      "loss": 1.2282,
+      "step": 10606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6360881328582764,
+      "learning_rate": 1.0307931286549708e-05,
+      "loss": 1.5274,
+      "step": 10607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.687594413757324,
+      "learning_rate": 1.030701754385965e-05,
+      "loss": 1.5063,
+      "step": 10608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.795964241027832,
+      "learning_rate": 1.0306103801169592e-05,
+      "loss": 1.8641,
+      "step": 10609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.485330104827881,
+      "learning_rate": 1.0305190058479533e-05,
+      "loss": 1.5524,
+      "step": 10610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4400808811187744,
+      "learning_rate": 1.0304276315789475e-05,
+      "loss": 1.4445,
+      "step": 10611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1956706047058105,
+      "learning_rate": 1.0303362573099415e-05,
+      "loss": 1.2637,
+      "step": 10612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.648951292037964,
+      "learning_rate": 1.0302448830409358e-05,
+      "loss": 1.6194,
+      "step": 10613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5529580116271973,
+      "learning_rate": 1.0301535087719298e-05,
+      "loss": 1.5099,
+      "step": 10614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.643975019454956,
+      "learning_rate": 1.030062134502924e-05,
+      "loss": 1.6969,
+      "step": 10615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0070669651031494,
+      "learning_rate": 1.0299707602339183e-05,
+      "loss": 1.664,
+      "step": 10616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.859076976776123,
+      "learning_rate": 1.0298793859649123e-05,
+      "loss": 1.946,
+      "step": 10617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1748342514038086,
+      "learning_rate": 1.0297880116959065e-05,
+      "loss": 1.8503,
+      "step": 10618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.319485664367676,
+      "learning_rate": 1.0296966374269006e-05,
+      "loss": 1.4291,
+      "step": 10619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.610868215560913,
+      "learning_rate": 1.0296052631578948e-05,
+      "loss": 1.3442,
+      "step": 10620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.581456184387207,
+      "learning_rate": 1.0295138888888889e-05,
+      "loss": 1.5325,
+      "step": 10621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4231152534484863,
+      "learning_rate": 1.029422514619883e-05,
+      "loss": 1.3758,
+      "step": 10622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5826282501220703,
+      "learning_rate": 1.0293311403508773e-05,
+      "loss": 1.4415,
+      "step": 10623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1993422508239746,
+      "learning_rate": 1.0292397660818714e-05,
+      "loss": 1.6466,
+      "step": 10624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.972142219543457,
+      "learning_rate": 1.0291483918128656e-05,
+      "loss": 1.6276,
+      "step": 10625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.414904832839966,
+      "learning_rate": 1.0290570175438596e-05,
+      "loss": 1.1449,
+      "step": 10626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.196047306060791,
+      "learning_rate": 1.0289656432748539e-05,
+      "loss": 1.2473,
+      "step": 10627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3131325244903564,
+      "learning_rate": 1.0288742690058479e-05,
+      "loss": 1.5118,
+      "step": 10628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.27571439743042,
+      "learning_rate": 1.0287828947368421e-05,
+      "loss": 1.315,
+      "step": 10629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8243935108184814,
+      "learning_rate": 1.0286915204678364e-05,
+      "loss": 1.4322,
+      "step": 10630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8596854209899902,
+      "learning_rate": 1.0286001461988304e-05,
+      "loss": 1.7508,
+      "step": 10631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4654619693756104,
+      "learning_rate": 1.0285087719298246e-05,
+      "loss": 1.4225,
+      "step": 10632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6681549549102783,
+      "learning_rate": 1.0284173976608187e-05,
+      "loss": 1.3897,
+      "step": 10633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2575056552886963,
+      "learning_rate": 1.0283260233918129e-05,
+      "loss": 1.1373,
+      "step": 10634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4094176292419434,
+      "learning_rate": 1.028234649122807e-05,
+      "loss": 1.4099,
+      "step": 10635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.450827121734619,
+      "learning_rate": 1.0281432748538012e-05,
+      "loss": 1.4782,
+      "step": 10636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6322195529937744,
+      "learning_rate": 1.0280519005847954e-05,
+      "loss": 1.4771,
+      "step": 10637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6216812133789062,
+      "learning_rate": 1.0279605263157895e-05,
+      "loss": 1.3834,
+      "step": 10638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.590496301651001,
+      "learning_rate": 1.0278691520467837e-05,
+      "loss": 1.5027,
+      "step": 10639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.897413492202759,
+      "learning_rate": 1.0277777777777777e-05,
+      "loss": 1.6068,
+      "step": 10640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.947460889816284,
+      "learning_rate": 1.027686403508772e-05,
+      "loss": 1.6221,
+      "step": 10641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.871206045150757,
+      "learning_rate": 1.027595029239766e-05,
+      "loss": 1.2971,
+      "step": 10642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.538646697998047,
+      "learning_rate": 1.0275036549707602e-05,
+      "loss": 1.5535,
+      "step": 10643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.382768154144287,
+      "learning_rate": 1.0274122807017544e-05,
+      "loss": 1.5008,
+      "step": 10644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5307507514953613,
+      "learning_rate": 1.0273209064327487e-05,
+      "loss": 1.3193,
+      "step": 10645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5026237964630127,
+      "learning_rate": 1.0272295321637429e-05,
+      "loss": 1.6349,
+      "step": 10646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.767711639404297,
+      "learning_rate": 1.0271381578947371e-05,
+      "loss": 1.57,
+      "step": 10647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0872042179107666,
+      "learning_rate": 1.0270467836257312e-05,
+      "loss": 1.2244,
+      "step": 10648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.583233118057251,
+      "learning_rate": 1.0269554093567254e-05,
+      "loss": 1.2838,
+      "step": 10649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4956796169281006,
+      "learning_rate": 1.0268640350877194e-05,
+      "loss": 1.6211,
+      "step": 10650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8866569995880127,
+      "learning_rate": 1.0267726608187137e-05,
+      "loss": 1.4355,
+      "step": 10651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1398332118988037,
+      "learning_rate": 1.0266812865497077e-05,
+      "loss": 1.7548,
+      "step": 10652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8446757793426514,
+      "learning_rate": 1.026589912280702e-05,
+      "loss": 1.4543,
+      "step": 10653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2199556827545166,
+      "learning_rate": 1.0264985380116962e-05,
+      "loss": 1.9932,
+      "step": 10654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8080196380615234,
+      "learning_rate": 1.0264071637426902e-05,
+      "loss": 1.5332,
+      "step": 10655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.859015941619873,
+      "learning_rate": 1.0263157894736844e-05,
+      "loss": 1.7895,
+      "step": 10656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.417856454849243,
+      "learning_rate": 1.0262244152046785e-05,
+      "loss": 1.1621,
+      "step": 10657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.5613250732421875,
+      "learning_rate": 1.0261330409356727e-05,
+      "loss": 1.7662,
+      "step": 10658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.385580062866211,
+      "learning_rate": 1.0260416666666668e-05,
+      "loss": 1.5142,
+      "step": 10659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9234471321105957,
+      "learning_rate": 1.025950292397661e-05,
+      "loss": 1.5118,
+      "step": 10660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5817549228668213,
+      "learning_rate": 1.0258589181286552e-05,
+      "loss": 2.0502,
+      "step": 10661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.474158525466919,
+      "learning_rate": 1.0257675438596493e-05,
+      "loss": 1.2299,
+      "step": 10662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9030816555023193,
+      "learning_rate": 1.0256761695906435e-05,
+      "loss": 1.8446,
+      "step": 10663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0552477836608887,
+      "learning_rate": 1.0255847953216375e-05,
+      "loss": 1.8132,
+      "step": 10664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7238712310791016,
+      "learning_rate": 1.0254934210526318e-05,
+      "loss": 1.3624,
+      "step": 10665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9942703247070312,
+      "learning_rate": 1.0254020467836258e-05,
+      "loss": 1.4653,
+      "step": 10666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3998069763183594,
+      "learning_rate": 1.02531067251462e-05,
+      "loss": 1.633,
+      "step": 10667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8484747409820557,
+      "learning_rate": 1.0252192982456143e-05,
+      "loss": 1.5406,
+      "step": 10668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.660768747329712,
+      "learning_rate": 1.0251279239766083e-05,
+      "loss": 1.4108,
+      "step": 10669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9369099140167236,
+      "learning_rate": 1.0250365497076025e-05,
+      "loss": 1.5504,
+      "step": 10670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4997007846832275,
+      "learning_rate": 1.0249451754385966e-05,
+      "loss": 1.3248,
+      "step": 10671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8613128662109375,
+      "learning_rate": 1.0248538011695908e-05,
+      "loss": 1.3981,
+      "step": 10672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7311699390411377,
+      "learning_rate": 1.0247624269005849e-05,
+      "loss": 1.7851,
+      "step": 10673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5073132514953613,
+      "learning_rate": 1.024671052631579e-05,
+      "loss": 1.4613,
+      "step": 10674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.849304437637329,
+      "learning_rate": 1.0245796783625733e-05,
+      "loss": 1.5688,
+      "step": 10675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2273998260498047,
+      "learning_rate": 1.0244883040935674e-05,
+      "loss": 2.1221,
+      "step": 10676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4815168380737305,
+      "learning_rate": 1.0243969298245616e-05,
+      "loss": 1.273,
+      "step": 10677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3799149990081787,
+      "learning_rate": 1.0243055555555556e-05,
+      "loss": 1.4482,
+      "step": 10678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.821640729904175,
+      "learning_rate": 1.0242141812865498e-05,
+      "loss": 1.6958,
+      "step": 10679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.693962812423706,
+      "learning_rate": 1.0241228070175439e-05,
+      "loss": 1.7174,
+      "step": 10680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.961280107498169,
+      "learning_rate": 1.0240314327485381e-05,
+      "loss": 1.6304,
+      "step": 10681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0680196285247803,
+      "learning_rate": 1.0239400584795323e-05,
+      "loss": 1.3707,
+      "step": 10682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4190733432769775,
+      "learning_rate": 1.0238486842105264e-05,
+      "loss": 1.3403,
+      "step": 10683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.807114362716675,
+      "learning_rate": 1.0237573099415206e-05,
+      "loss": 1.7127,
+      "step": 10684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.319704294204712,
+      "learning_rate": 1.0236659356725147e-05,
+      "loss": 1.3711,
+      "step": 10685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1064796447753906,
+      "learning_rate": 1.0235745614035089e-05,
+      "loss": 1.3544,
+      "step": 10686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.381988048553467,
+      "learning_rate": 1.023483187134503e-05,
+      "loss": 2.0902,
+      "step": 10687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9693145751953125,
+      "learning_rate": 1.0233918128654972e-05,
+      "loss": 1.6112,
+      "step": 10688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.061357259750366,
+      "learning_rate": 1.0233004385964914e-05,
+      "loss": 1.8076,
+      "step": 10689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.263179302215576,
+      "learning_rate": 1.0232090643274854e-05,
+      "loss": 1.298,
+      "step": 10690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2814202308654785,
+      "learning_rate": 1.0231176900584797e-05,
+      "loss": 1.4798,
+      "step": 10691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6270439624786377,
+      "learning_rate": 1.0230263157894737e-05,
+      "loss": 1.9174,
+      "step": 10692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0484073162078857,
+      "learning_rate": 1.022934941520468e-05,
+      "loss": 1.7543,
+      "step": 10693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.968663215637207,
+      "learning_rate": 1.022843567251462e-05,
+      "loss": 1.2205,
+      "step": 10694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.21323299407959,
+      "learning_rate": 1.0227521929824562e-05,
+      "loss": 1.517,
+      "step": 10695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.566927671432495,
+      "learning_rate": 1.0226608187134504e-05,
+      "loss": 1.4489,
+      "step": 10696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9123730659484863,
+      "learning_rate": 1.0225694444444445e-05,
+      "loss": 1.0685,
+      "step": 10697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.552865982055664,
+      "learning_rate": 1.0224780701754387e-05,
+      "loss": 1.1973,
+      "step": 10698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3319308757781982,
+      "learning_rate": 1.0223866959064328e-05,
+      "loss": 1.3787,
+      "step": 10699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2306203842163086,
+      "learning_rate": 1.022295321637427e-05,
+      "loss": 1.3847,
+      "step": 10700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0928659439086914,
+      "learning_rate": 1.022203947368421e-05,
+      "loss": 1.2447,
+      "step": 10701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.377872943878174,
+      "learning_rate": 1.0221125730994153e-05,
+      "loss": 1.3866,
+      "step": 10702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0815927982330322,
+      "learning_rate": 1.0220211988304093e-05,
+      "loss": 1.3793,
+      "step": 10703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6475448608398438,
+      "learning_rate": 1.0219298245614035e-05,
+      "loss": 1.7131,
+      "step": 10704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.396312952041626,
+      "learning_rate": 1.0218384502923978e-05,
+      "loss": 1.3264,
+      "step": 10705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7740635871887207,
+      "learning_rate": 1.0217470760233918e-05,
+      "loss": 1.7976,
+      "step": 10706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0015532970428467,
+      "learning_rate": 1.021655701754386e-05,
+      "loss": 1.1998,
+      "step": 10707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.653550148010254,
+      "learning_rate": 1.0215643274853801e-05,
+      "loss": 1.6889,
+      "step": 10708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1755287647247314,
+      "learning_rate": 1.0214729532163743e-05,
+      "loss": 2.2289,
+      "step": 10709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7676196098327637,
+      "learning_rate": 1.0213815789473684e-05,
+      "loss": 1.334,
+      "step": 10710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6138055324554443,
+      "learning_rate": 1.0212902046783626e-05,
+      "loss": 1.3997,
+      "step": 10711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.667423963546753,
+      "learning_rate": 1.0211988304093568e-05,
+      "loss": 1.6496,
+      "step": 10712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.440298318862915,
+      "learning_rate": 1.0211074561403509e-05,
+      "loss": 1.56,
+      "step": 10713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.881380081176758,
+      "learning_rate": 1.021016081871345e-05,
+      "loss": 1.7143,
+      "step": 10714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2665648460388184,
+      "learning_rate": 1.0209247076023391e-05,
+      "loss": 1.1052,
+      "step": 10715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7586281299591064,
+      "learning_rate": 1.0208333333333334e-05,
+      "loss": 1.8014,
+      "step": 10716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.852597951889038,
+      "learning_rate": 1.0207419590643274e-05,
+      "loss": 1.844,
+      "step": 10717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.598212242126465,
+      "learning_rate": 1.0206505847953216e-05,
+      "loss": 1.5594,
+      "step": 10718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3413143157958984,
+      "learning_rate": 1.0205592105263159e-05,
+      "loss": 1.4308,
+      "step": 10719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.412806272506714,
+      "learning_rate": 1.0204678362573099e-05,
+      "loss": 1.3155,
+      "step": 10720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5990114212036133,
+      "learning_rate": 1.0203764619883041e-05,
+      "loss": 1.2878,
+      "step": 10721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5184521675109863,
+      "learning_rate": 1.0202850877192982e-05,
+      "loss": 1.5284,
+      "step": 10722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5891671180725098,
+      "learning_rate": 1.0201937134502924e-05,
+      "loss": 1.4823,
+      "step": 10723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.845775604248047,
+      "learning_rate": 1.0201023391812865e-05,
+      "loss": 1.6885,
+      "step": 10724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.357417106628418,
+      "learning_rate": 1.0200109649122807e-05,
+      "loss": 1.6082,
+      "step": 10725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.475825786590576,
+      "learning_rate": 1.0199195906432749e-05,
+      "loss": 1.3563,
+      "step": 10726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1216931343078613,
+      "learning_rate": 1.019828216374269e-05,
+      "loss": 1.1703,
+      "step": 10727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.077256917953491,
+      "learning_rate": 1.0197368421052632e-05,
+      "loss": 1.6459,
+      "step": 10728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4577479362487793,
+      "learning_rate": 1.0196454678362572e-05,
+      "loss": 1.4443,
+      "step": 10729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7366838455200195,
+      "learning_rate": 1.0195540935672515e-05,
+      "loss": 1.3004,
+      "step": 10730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7411859035491943,
+      "learning_rate": 1.0194627192982458e-05,
+      "loss": 1.5608,
+      "step": 10731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3697621822357178,
+      "learning_rate": 1.0193713450292399e-05,
+      "loss": 1.4589,
+      "step": 10732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2982966899871826,
+      "learning_rate": 1.0192799707602341e-05,
+      "loss": 1.1692,
+      "step": 10733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5544984340667725,
+      "learning_rate": 1.0191885964912283e-05,
+      "loss": 1.4694,
+      "step": 10734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.294710874557495,
+      "learning_rate": 1.0190972222222224e-05,
+      "loss": 1.1734,
+      "step": 10735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.727555274963379,
+      "learning_rate": 1.0190058479532166e-05,
+      "loss": 1.4419,
+      "step": 10736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8085875511169434,
+      "learning_rate": 1.0189144736842107e-05,
+      "loss": 1.5093,
+      "step": 10737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.400475263595581,
+      "learning_rate": 1.0188230994152049e-05,
+      "loss": 1.4965,
+      "step": 10738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.296308755874634,
+      "learning_rate": 1.018731725146199e-05,
+      "loss": 1.8323,
+      "step": 10739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.157564878463745,
+      "learning_rate": 1.0186403508771932e-05,
+      "loss": 1.748,
+      "step": 10740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.088371515274048,
+      "learning_rate": 1.0185489766081872e-05,
+      "loss": 1.8641,
+      "step": 10741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.288400173187256,
+      "learning_rate": 1.0184576023391814e-05,
+      "loss": 1.234,
+      "step": 10742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.295750379562378,
+      "learning_rate": 1.0183662280701757e-05,
+      "loss": 1.4087,
+      "step": 10743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.59818696975708,
+      "learning_rate": 1.0182748538011697e-05,
+      "loss": 1.398,
+      "step": 10744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3870980739593506,
+      "learning_rate": 1.018183479532164e-05,
+      "loss": 1.2772,
+      "step": 10745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5273520946502686,
+      "learning_rate": 1.018092105263158e-05,
+      "loss": 1.3718,
+      "step": 10746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.409776210784912,
+      "learning_rate": 1.0180007309941522e-05,
+      "loss": 1.528,
+      "step": 10747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8650505542755127,
+      "learning_rate": 1.0179093567251463e-05,
+      "loss": 1.751,
+      "step": 10748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8809828758239746,
+      "learning_rate": 1.0178179824561405e-05,
+      "loss": 2.1004,
+      "step": 10749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6451408863067627,
+      "learning_rate": 1.0177266081871347e-05,
+      "loss": 1.3777,
+      "step": 10750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.500903606414795,
+      "learning_rate": 1.0176352339181288e-05,
+      "loss": 1.3511,
+      "step": 10751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.753645896911621,
+      "learning_rate": 1.017543859649123e-05,
+      "loss": 1.6331,
+      "step": 10752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7879350185394287,
+      "learning_rate": 1.017452485380117e-05,
+      "loss": 1.3632,
+      "step": 10753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.715212821960449,
+      "learning_rate": 1.0173611111111113e-05,
+      "loss": 1.8257,
+      "step": 10754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0272340774536133,
+      "learning_rate": 1.0172697368421053e-05,
+      "loss": 1.5748,
+      "step": 10755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.86910080909729,
+      "learning_rate": 1.0171783625730995e-05,
+      "loss": 1.5566,
+      "step": 10756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.562009811401367,
+      "learning_rate": 1.0170869883040938e-05,
+      "loss": 1.3868,
+      "step": 10757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6508233547210693,
+      "learning_rate": 1.0169956140350878e-05,
+      "loss": 1.4959,
+      "step": 10758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.423780918121338,
+      "learning_rate": 1.016904239766082e-05,
+      "loss": 1.2412,
+      "step": 10759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1749584674835205,
+      "learning_rate": 1.016812865497076e-05,
+      "loss": 1.4516,
+      "step": 10760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7078745365142822,
+      "learning_rate": 1.0167214912280703e-05,
+      "loss": 1.8433,
+      "step": 10761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6013338565826416,
+      "learning_rate": 1.0166301169590644e-05,
+      "loss": 1.4351,
+      "step": 10762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.716399908065796,
+      "learning_rate": 1.0165387426900586e-05,
+      "loss": 1.4402,
+      "step": 10763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.265815019607544,
+      "learning_rate": 1.0164473684210528e-05,
+      "loss": 1.2203,
+      "step": 10764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2779746055603027,
+      "learning_rate": 1.0163559941520469e-05,
+      "loss": 1.7503,
+      "step": 10765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3397374153137207,
+      "learning_rate": 1.016264619883041e-05,
+      "loss": 1.3932,
+      "step": 10766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9315006732940674,
+      "learning_rate": 1.0161732456140351e-05,
+      "loss": 1.8444,
+      "step": 10767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9690377712249756,
+      "learning_rate": 1.0160818713450294e-05,
+      "loss": 1.5403,
+      "step": 10768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2777979373931885,
+      "learning_rate": 1.0159904970760234e-05,
+      "loss": 1.564,
+      "step": 10769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7124533653259277,
+      "learning_rate": 1.0158991228070176e-05,
+      "loss": 1.2198,
+      "step": 10770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.823244571685791,
+      "learning_rate": 1.0158077485380118e-05,
+      "loss": 1.3661,
+      "step": 10771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.857950448989868,
+      "learning_rate": 1.0157163742690059e-05,
+      "loss": 1.3528,
+      "step": 10772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2816507816314697,
+      "learning_rate": 1.0156250000000001e-05,
+      "loss": 1.552,
+      "step": 10773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.382518768310547,
+      "learning_rate": 1.0155336257309942e-05,
+      "loss": 1.3912,
+      "step": 10774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4502370357513428,
+      "learning_rate": 1.0154422514619884e-05,
+      "loss": 1.613,
+      "step": 10775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5878095626831055,
+      "learning_rate": 1.0153508771929825e-05,
+      "loss": 1.4673,
+      "step": 10776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.136468887329102,
+      "learning_rate": 1.0152595029239767e-05,
+      "loss": 1.5234,
+      "step": 10777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4151031970977783,
+      "learning_rate": 1.0151681286549709e-05,
+      "loss": 1.5613,
+      "step": 10778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5627317428588867,
+      "learning_rate": 1.015076754385965e-05,
+      "loss": 1.6586,
+      "step": 10779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4614715576171875,
+      "learning_rate": 1.0149853801169592e-05,
+      "loss": 1.7322,
+      "step": 10780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7724056243896484,
+      "learning_rate": 1.0148940058479532e-05,
+      "loss": 1.5042,
+      "step": 10781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0150914192199707,
+      "learning_rate": 1.0148026315789474e-05,
+      "loss": 1.7282,
+      "step": 10782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9523253440856934,
+      "learning_rate": 1.0147112573099415e-05,
+      "loss": 1.3811,
+      "step": 10783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9261791706085205,
+      "learning_rate": 1.0146198830409357e-05,
+      "loss": 1.9848,
+      "step": 10784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7296385765075684,
+      "learning_rate": 1.01452850877193e-05,
+      "loss": 1.5676,
+      "step": 10785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.281402587890625,
+      "learning_rate": 1.014437134502924e-05,
+      "loss": 1.2591,
+      "step": 10786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9401543140411377,
+      "learning_rate": 1.0143457602339182e-05,
+      "loss": 1.5524,
+      "step": 10787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.896836042404175,
+      "learning_rate": 1.0142543859649123e-05,
+      "loss": 1.7008,
+      "step": 10788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.488201856613159,
+      "learning_rate": 1.0141630116959065e-05,
+      "loss": 1.5981,
+      "step": 10789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7349467277526855,
+      "learning_rate": 1.0140716374269005e-05,
+      "loss": 1.4306,
+      "step": 10790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4765970706939697,
+      "learning_rate": 1.0139802631578948e-05,
+      "loss": 1.2183,
+      "step": 10791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.728827953338623,
+      "learning_rate": 1.013888888888889e-05,
+      "loss": 1.531,
+      "step": 10792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6621627807617188,
+      "learning_rate": 1.013797514619883e-05,
+      "loss": 1.3941,
+      "step": 10793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.776642084121704,
+      "learning_rate": 1.0137061403508773e-05,
+      "loss": 1.3949,
+      "step": 10794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.026982545852661,
+      "learning_rate": 1.0136147660818713e-05,
+      "loss": 2.0112,
+      "step": 10795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4612996578216553,
+      "learning_rate": 1.0135233918128655e-05,
+      "loss": 1.5504,
+      "step": 10796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6742851734161377,
+      "learning_rate": 1.0134320175438596e-05,
+      "loss": 1.5456,
+      "step": 10797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3663089275360107,
+      "learning_rate": 1.0133406432748538e-05,
+      "loss": 1.3649,
+      "step": 10798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4013373851776123,
+      "learning_rate": 1.013249269005848e-05,
+      "loss": 1.3712,
+      "step": 10799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5365242958068848,
+      "learning_rate": 1.0131578947368421e-05,
+      "loss": 1.2603,
+      "step": 10800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.741793632507324,
+      "learning_rate": 1.0130665204678363e-05,
+      "loss": 1.5977,
+      "step": 10801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5218780040740967,
+      "learning_rate": 1.0129751461988304e-05,
+      "loss": 1.4152,
+      "step": 10802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3241376876831055,
+      "learning_rate": 1.0128837719298246e-05,
+      "loss": 1.4273,
+      "step": 10803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9765630960464478,
+      "learning_rate": 1.0127923976608186e-05,
+      "loss": 1.3494,
+      "step": 10804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.672942876815796,
+      "learning_rate": 1.0127010233918129e-05,
+      "loss": 1.6439,
+      "step": 10805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.570922374725342,
+      "learning_rate": 1.012609649122807e-05,
+      "loss": 1.8316,
+      "step": 10806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4030139446258545,
+      "learning_rate": 1.0125182748538011e-05,
+      "loss": 1.7732,
+      "step": 10807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0289969444274902,
+      "learning_rate": 1.0124269005847954e-05,
+      "loss": 1.8937,
+      "step": 10808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.469782590866089,
+      "learning_rate": 1.0123355263157894e-05,
+      "loss": 1.7567,
+      "step": 10809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4440629482269287,
+      "learning_rate": 1.0122441520467836e-05,
+      "loss": 1.6333,
+      "step": 10810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2169837951660156,
+      "learning_rate": 1.0121527777777777e-05,
+      "loss": 1.3742,
+      "step": 10811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2581324577331543,
+      "learning_rate": 1.0120614035087719e-05,
+      "loss": 1.2166,
+      "step": 10812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.418588399887085,
+      "learning_rate": 1.0119700292397661e-05,
+      "loss": 1.3501,
+      "step": 10813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8795063495635986,
+      "learning_rate": 1.0118786549707602e-05,
+      "loss": 1.4423,
+      "step": 10814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.682236433029175,
+      "learning_rate": 1.0117872807017544e-05,
+      "loss": 1.7869,
+      "step": 10815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6760377883911133,
+      "learning_rate": 1.0116959064327488e-05,
+      "loss": 1.7214,
+      "step": 10816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.136479616165161,
+      "learning_rate": 1.0116045321637428e-05,
+      "loss": 1.264,
+      "step": 10817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.599740743637085,
+      "learning_rate": 1.011513157894737e-05,
+      "loss": 1.6825,
+      "step": 10818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3641674518585205,
+      "learning_rate": 1.0114217836257311e-05,
+      "loss": 1.5948,
+      "step": 10819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0580198764801025,
+      "learning_rate": 1.0113304093567253e-05,
+      "loss": 1.8934,
+      "step": 10820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0538151264190674,
+      "learning_rate": 1.0112390350877194e-05,
+      "loss": 2.0102,
+      "step": 10821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0388927459716797,
+      "learning_rate": 1.0111476608187136e-05,
+      "loss": 1.82,
+      "step": 10822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.909109115600586,
+      "learning_rate": 1.0110562865497078e-05,
+      "loss": 1.6034,
+      "step": 10823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0882298946380615,
+      "learning_rate": 1.0109649122807019e-05,
+      "loss": 1.265,
+      "step": 10824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.221625566482544,
+      "learning_rate": 1.0108735380116961e-05,
+      "loss": 1.3101,
+      "step": 10825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9915363788604736,
+      "learning_rate": 1.0107821637426902e-05,
+      "loss": 1.5124,
+      "step": 10826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5280003547668457,
+      "learning_rate": 1.0106907894736844e-05,
+      "loss": 1.6465,
+      "step": 10827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9503815174102783,
+      "learning_rate": 1.0105994152046784e-05,
+      "loss": 1.0075,
+      "step": 10828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6871488094329834,
+      "learning_rate": 1.0105080409356727e-05,
+      "loss": 1.7357,
+      "step": 10829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4418721199035645,
+      "learning_rate": 1.0104166666666669e-05,
+      "loss": 1.5385,
+      "step": 10830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6426568031311035,
+      "learning_rate": 1.010325292397661e-05,
+      "loss": 1.584,
+      "step": 10831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.138400077819824,
+      "learning_rate": 1.0102339181286552e-05,
+      "loss": 1.1691,
+      "step": 10832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8195035457611084,
+      "learning_rate": 1.0101425438596492e-05,
+      "loss": 1.6354,
+      "step": 10833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.980703592300415,
+      "learning_rate": 1.0100511695906434e-05,
+      "loss": 1.9321,
+      "step": 10834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.310925245285034,
+      "learning_rate": 1.0099597953216375e-05,
+      "loss": 1.3351,
+      "step": 10835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.340684652328491,
+      "learning_rate": 1.0098684210526317e-05,
+      "loss": 1.5363,
+      "step": 10836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5576469898223877,
+      "learning_rate": 1.009777046783626e-05,
+      "loss": 1.481,
+      "step": 10837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1961872577667236,
+      "learning_rate": 1.00968567251462e-05,
+      "loss": 1.3521,
+      "step": 10838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.464731216430664,
+      "learning_rate": 1.0095942982456142e-05,
+      "loss": 1.4552,
+      "step": 10839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7044005393981934,
+      "learning_rate": 1.0095029239766083e-05,
+      "loss": 1.5689,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8445093631744385,
+      "learning_rate": 1.0094115497076025e-05,
+      "loss": 1.3624,
+      "step": 10841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.361462354660034,
+      "learning_rate": 1.0093201754385965e-05,
+      "loss": 1.3735,
+      "step": 10842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8504936695098877,
+      "learning_rate": 1.0092288011695908e-05,
+      "loss": 1.0905,
+      "step": 10843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9512701034545898,
+      "learning_rate": 1.009137426900585e-05,
+      "loss": 1.1998,
+      "step": 10844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6895458698272705,
+      "learning_rate": 1.009046052631579e-05,
+      "loss": 1.3406,
+      "step": 10845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7474722862243652,
+      "learning_rate": 1.0089546783625733e-05,
+      "loss": 1.455,
+      "step": 10846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8722589015960693,
+      "learning_rate": 1.0088633040935673e-05,
+      "loss": 1.763,
+      "step": 10847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9818830490112305,
+      "learning_rate": 1.0087719298245615e-05,
+      "loss": 2.0213,
+      "step": 10848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8679258823394775,
+      "learning_rate": 1.0086805555555556e-05,
+      "loss": 1.3764,
+      "step": 10849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4979584217071533,
+      "learning_rate": 1.0085891812865498e-05,
+      "loss": 1.5693,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.869565963745117,
+      "learning_rate": 1.008497807017544e-05,
+      "loss": 1.5934,
+      "step": 10851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.648930311203003,
+      "learning_rate": 1.008406432748538e-05,
+      "loss": 1.504,
+      "step": 10852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7108829021453857,
+      "learning_rate": 1.0083150584795323e-05,
+      "loss": 1.3556,
+      "step": 10853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.725947618484497,
+      "learning_rate": 1.0082236842105264e-05,
+      "loss": 1.6454,
+      "step": 10854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2958617210388184,
+      "learning_rate": 1.0081323099415206e-05,
+      "loss": 1.1881,
+      "step": 10855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8034143447875977,
+      "learning_rate": 1.0080409356725146e-05,
+      "loss": 1.5014,
+      "step": 10856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0718255043029785,
+      "learning_rate": 1.0079495614035089e-05,
+      "loss": 1.5981,
+      "step": 10857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.607469320297241,
+      "learning_rate": 1.007858187134503e-05,
+      "loss": 1.4591,
+      "step": 10858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.862760543823242,
+      "learning_rate": 1.0077668128654971e-05,
+      "loss": 1.4159,
+      "step": 10859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3622894287109375,
+      "learning_rate": 1.0076754385964914e-05,
+      "loss": 1.3507,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6307597160339355,
+      "learning_rate": 1.0075840643274854e-05,
+      "loss": 1.4416,
+      "step": 10861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5088865756988525,
+      "learning_rate": 1.0074926900584796e-05,
+      "loss": 1.465,
+      "step": 10862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0884358882904053,
+      "learning_rate": 1.0074013157894737e-05,
+      "loss": 1.6861,
+      "step": 10863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0728509426116943,
+      "learning_rate": 1.0073099415204679e-05,
+      "loss": 1.6908,
+      "step": 10864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2712743282318115,
+      "learning_rate": 1.0072185672514621e-05,
+      "loss": 1.3879,
+      "step": 10865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3056745529174805,
+      "learning_rate": 1.0071271929824562e-05,
+      "loss": 1.349,
+      "step": 10866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.387251138687134,
+      "learning_rate": 1.0070358187134504e-05,
+      "loss": 1.4931,
+      "step": 10867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6945695877075195,
+      "learning_rate": 1.0069444444444445e-05,
+      "loss": 1.3754,
+      "step": 10868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0976390838623047,
+      "learning_rate": 1.0068530701754387e-05,
+      "loss": 1.6571,
+      "step": 10869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4384210109710693,
+      "learning_rate": 1.0067616959064327e-05,
+      "loss": 1.5453,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4011974334716797,
+      "learning_rate": 1.006670321637427e-05,
+      "loss": 1.6195,
+      "step": 10871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9240589141845703,
+      "learning_rate": 1.0065789473684212e-05,
+      "loss": 1.4374,
+      "step": 10872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2411022186279297,
+      "learning_rate": 1.0064875730994152e-05,
+      "loss": 1.1631,
+      "step": 10873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0262508392333984,
+      "learning_rate": 1.0063961988304094e-05,
+      "loss": 1.6518,
+      "step": 10874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5809381008148193,
+      "learning_rate": 1.0063048245614035e-05,
+      "loss": 1.1775,
+      "step": 10875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9426422119140625,
+      "learning_rate": 1.0062134502923977e-05,
+      "loss": 1.6223,
+      "step": 10876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4182446002960205,
+      "learning_rate": 1.0061220760233918e-05,
+      "loss": 1.4464,
+      "step": 10877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9138941764831543,
+      "learning_rate": 1.006030701754386e-05,
+      "loss": 1.7105,
+      "step": 10878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.530644655227661,
+      "learning_rate": 1.0059393274853802e-05,
+      "loss": 1.4569,
+      "step": 10879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0585238933563232,
+      "learning_rate": 1.0058479532163743e-05,
+      "loss": 1.5133,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5863234996795654,
+      "learning_rate": 1.0057565789473685e-05,
+      "loss": 1.7667,
+      "step": 10881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4572913646698,
+      "learning_rate": 1.0056652046783625e-05,
+      "loss": 1.5046,
+      "step": 10882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1864030361175537,
+      "learning_rate": 1.0055738304093568e-05,
+      "loss": 1.4297,
+      "step": 10883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.300680160522461,
+      "learning_rate": 1.0054824561403508e-05,
+      "loss": 1.2312,
+      "step": 10884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.926211357116699,
+      "learning_rate": 1.005391081871345e-05,
+      "loss": 1.6744,
+      "step": 10885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.748401165008545,
+      "learning_rate": 1.0052997076023393e-05,
+      "loss": 1.6296,
+      "step": 10886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.648275375366211,
+      "learning_rate": 1.0052083333333333e-05,
+      "loss": 1.5662,
+      "step": 10887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.454634189605713,
+      "learning_rate": 1.0051169590643275e-05,
+      "loss": 1.6567,
+      "step": 10888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.832324266433716,
+      "learning_rate": 1.0050255847953216e-05,
+      "loss": 1.6737,
+      "step": 10889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8603503704071045,
+      "learning_rate": 1.0049342105263158e-05,
+      "loss": 1.5598,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0269832611083984,
+      "learning_rate": 1.0048428362573099e-05,
+      "loss": 2.1768,
+      "step": 10891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4870312213897705,
+      "learning_rate": 1.0047514619883041e-05,
+      "loss": 1.4047,
+      "step": 10892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4832818508148193,
+      "learning_rate": 1.0046600877192983e-05,
+      "loss": 1.3428,
+      "step": 10893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5337564945220947,
+      "learning_rate": 1.0045687134502924e-05,
+      "loss": 1.2253,
+      "step": 10894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.311459541320801,
+      "learning_rate": 1.0044773391812866e-05,
+      "loss": 1.3592,
+      "step": 10895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.575918197631836,
+      "learning_rate": 1.0043859649122806e-05,
+      "loss": 1.4212,
+      "step": 10896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.291391611099243,
+      "learning_rate": 1.0042945906432749e-05,
+      "loss": 1.5264,
+      "step": 10897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.828467845916748,
+      "learning_rate": 1.0042032163742689e-05,
+      "loss": 1.7304,
+      "step": 10898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.47454833984375,
+      "learning_rate": 1.0041118421052631e-05,
+      "loss": 1.3165,
+      "step": 10899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8720269203186035,
+      "learning_rate": 1.0040204678362574e-05,
+      "loss": 1.2701,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8735101222991943,
+      "learning_rate": 1.0039290935672514e-05,
+      "loss": 1.3206,
+      "step": 10901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4872210025787354,
+      "learning_rate": 1.0038377192982458e-05,
+      "loss": 1.3773,
+      "step": 10902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4150569438934326,
+      "learning_rate": 1.00374634502924e-05,
+      "loss": 1.4007,
+      "step": 10903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6323533058166504,
+      "learning_rate": 1.003654970760234e-05,
+      "loss": 1.3719,
+      "step": 10904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5583958625793457,
+      "learning_rate": 1.0035635964912283e-05,
+      "loss": 1.535,
+      "step": 10905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3205935955047607,
+      "learning_rate": 1.0034722222222224e-05,
+      "loss": 1.482,
+      "step": 10906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.141493797302246,
+      "learning_rate": 1.0033808479532166e-05,
+      "loss": 1.344,
+      "step": 10907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4084718227386475,
+      "learning_rate": 1.0032894736842106e-05,
+      "loss": 1.2739,
+      "step": 10908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9144840240478516,
+      "learning_rate": 1.0031980994152048e-05,
+      "loss": 1.5269,
+      "step": 10909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8494231700897217,
+      "learning_rate": 1.003106725146199e-05,
+      "loss": 1.7417,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3244473934173584,
+      "learning_rate": 1.0030153508771931e-05,
+      "loss": 1.3441,
+      "step": 10911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2726614475250244,
+      "learning_rate": 1.0029239766081873e-05,
+      "loss": 1.3625,
+      "step": 10912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3663690090179443,
+      "learning_rate": 1.0028326023391814e-05,
+      "loss": 1.4743,
+      "step": 10913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.753505229949951,
+      "learning_rate": 1.0027412280701756e-05,
+      "loss": 1.6375,
+      "step": 10914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.853595495223999,
+      "learning_rate": 1.0026498538011697e-05,
+      "loss": 1.9285,
+      "step": 10915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.423316955566406,
+      "learning_rate": 1.0025584795321639e-05,
+      "loss": 1.5352,
+      "step": 10916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0191650390625,
+      "learning_rate": 1.0024671052631581e-05,
+      "loss": 1.4216,
+      "step": 10917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7125418186187744,
+      "learning_rate": 1.0023757309941522e-05,
+      "loss": 1.6666,
+      "step": 10918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.406773328781128,
+      "learning_rate": 1.0022843567251464e-05,
+      "loss": 1.3547,
+      "step": 10919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.628647565841675,
+      "learning_rate": 1.0021929824561404e-05,
+      "loss": 1.3343,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8897476196289062,
+      "learning_rate": 1.0021016081871347e-05,
+      "loss": 1.5304,
+      "step": 10921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.36112904548645,
+      "learning_rate": 1.0020102339181287e-05,
+      "loss": 1.212,
+      "step": 10922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2814528942108154,
+      "learning_rate": 1.001918859649123e-05,
+      "loss": 1.1183,
+      "step": 10923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.577449321746826,
+      "learning_rate": 1.0018274853801172e-05,
+      "loss": 1.3594,
+      "step": 10924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.578990936279297,
+      "learning_rate": 1.0017361111111112e-05,
+      "loss": 1.5225,
+      "step": 10925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.419494867324829,
+      "learning_rate": 1.0016447368421054e-05,
+      "loss": 1.3558,
+      "step": 10926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7176172733306885,
+      "learning_rate": 1.0015533625730995e-05,
+      "loss": 1.0583,
+      "step": 10927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.575261354446411,
+      "learning_rate": 1.0014619883040937e-05,
+      "loss": 1.7054,
+      "step": 10928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4443745613098145,
+      "learning_rate": 1.0013706140350878e-05,
+      "loss": 1.3552,
+      "step": 10929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6803348064422607,
+      "learning_rate": 1.001279239766082e-05,
+      "loss": 1.6073,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.9040350914001465,
+      "learning_rate": 1.0011878654970762e-05,
+      "loss": 1.8844,
+      "step": 10931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.674764633178711,
+      "learning_rate": 1.0010964912280703e-05,
+      "loss": 1.3962,
+      "step": 10932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7027246952056885,
+      "learning_rate": 1.0010051169590645e-05,
+      "loss": 1.47,
+      "step": 10933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.636099100112915,
+      "learning_rate": 1.0009137426900585e-05,
+      "loss": 1.5336,
+      "step": 10934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.450664758682251,
+      "learning_rate": 1.0008223684210528e-05,
+      "loss": 1.3915,
+      "step": 10935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8736164569854736,
+      "learning_rate": 1.0007309941520468e-05,
+      "loss": 1.8089,
+      "step": 10936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.648749589920044,
+      "learning_rate": 1.000639619883041e-05,
+      "loss": 1.3683,
+      "step": 10937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.764993190765381,
+      "learning_rate": 1.0005482456140353e-05,
+      "loss": 1.4622,
+      "step": 10938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5810108184814453,
+      "learning_rate": 1.0004568713450293e-05,
+      "loss": 1.5526,
+      "step": 10939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5114381313323975,
+      "learning_rate": 1.0003654970760235e-05,
+      "loss": 1.7496,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5010147094726562,
+      "learning_rate": 1.0002741228070176e-05,
+      "loss": 1.2961,
+      "step": 10941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6064205169677734,
+      "learning_rate": 1.0001827485380118e-05,
+      "loss": 1.4091,
+      "step": 10942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.406364679336548,
+      "learning_rate": 1.0000913742690059e-05,
+      "loss": 1.4511,
+      "step": 10943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.72839093208313,
+      "learning_rate": 1e-05,
+      "loss": 1.6078,
+      "step": 10944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6121695041656494,
+      "learning_rate": 9.999086257309943e-06,
+      "loss": 1.5749,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5,
+      "eval_train_bleu": 0.23156597696882233,
+      "eval_train_loss": 2.1183290481567383,
+      "eval_train_runtime": 24.17,
+      "eval_train_samples_per_second": 41.415,
+      "eval_train_steps_per_second": 1.324,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3375723361968994,
+      "learning_rate": 9.998172514619884e-06,
+      "loss": 1.5042,
+      "step": 10946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.021474838256836,
+      "learning_rate": 9.997258771929826e-06,
+      "loss": 1.5047,
+      "step": 10947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3990070819854736,
+      "learning_rate": 9.996345029239766e-06,
+      "loss": 1.9952,
+      "step": 10948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.723991632461548,
+      "learning_rate": 9.995431286549709e-06,
+      "loss": 1.9361,
+      "step": 10949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.269439697265625,
+      "learning_rate": 9.994517543859649e-06,
+      "loss": 1.2494,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.410555839538574,
+      "learning_rate": 9.993603801169591e-06,
+      "loss": 1.3452,
+      "step": 10951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.311365842819214,
+      "learning_rate": 9.992690058479534e-06,
+      "loss": 1.494,
+      "step": 10952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5362391471862793,
+      "learning_rate": 9.991776315789474e-06,
+      "loss": 1.4196,
+      "step": 10953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2023110389709473,
+      "learning_rate": 9.990862573099416e-06,
+      "loss": 1.2293,
+      "step": 10954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.25726580619812,
+      "learning_rate": 9.989948830409357e-06,
+      "loss": 1.2853,
+      "step": 10955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.156320333480835,
+      "learning_rate": 9.989035087719299e-06,
+      "loss": 1.1872,
+      "step": 10956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6989262104034424,
+      "learning_rate": 9.98812134502924e-06,
+      "loss": 1.7385,
+      "step": 10957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.655233144760132,
+      "learning_rate": 9.987207602339182e-06,
+      "loss": 1.6612,
+      "step": 10958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2815473079681396,
+      "learning_rate": 9.986293859649124e-06,
+      "loss": 1.5784,
+      "step": 10959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7270348072052,
+      "learning_rate": 9.985380116959065e-06,
+      "loss": 1.4507,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.816555976867676,
+      "learning_rate": 9.984466374269007e-06,
+      "loss": 1.6418,
+      "step": 10961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6487326622009277,
+      "learning_rate": 9.983552631578947e-06,
+      "loss": 1.3188,
+      "step": 10962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7178242206573486,
+      "learning_rate": 9.98263888888889e-06,
+      "loss": 1.4663,
+      "step": 10963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3262925148010254,
+      "learning_rate": 9.98172514619883e-06,
+      "loss": 1.546,
+      "step": 10964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8471951484680176,
+      "learning_rate": 9.980811403508772e-06,
+      "loss": 1.5924,
+      "step": 10965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4171743392944336,
+      "learning_rate": 9.979897660818714e-06,
+      "loss": 1.8115,
+      "step": 10966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0071990489959717,
+      "learning_rate": 9.978983918128657e-06,
+      "loss": 1.0185,
+      "step": 10967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2874393463134766,
+      "learning_rate": 9.978070175438597e-06,
+      "loss": 1.2994,
+      "step": 10968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3525214195251465,
+      "learning_rate": 9.97715643274854e-06,
+      "loss": 1.2467,
+      "step": 10969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.34763240814209,
+      "learning_rate": 9.97624269005848e-06,
+      "loss": 1.8367,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.398235321044922,
+      "learning_rate": 9.975328947368422e-06,
+      "loss": 1.4102,
+      "step": 10971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7776365280151367,
+      "learning_rate": 9.974415204678364e-06,
+      "loss": 1.4027,
+      "step": 10972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.320322275161743,
+      "learning_rate": 9.973501461988305e-06,
+      "loss": 1.4526,
+      "step": 10973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.886779308319092,
+      "learning_rate": 9.972587719298247e-06,
+      "loss": 1.9149,
+      "step": 10974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6224453449249268,
+      "learning_rate": 9.971673976608188e-06,
+      "loss": 1.6294,
+      "step": 10975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.320352077484131,
+      "learning_rate": 9.97076023391813e-06,
+      "loss": 1.3044,
+      "step": 10976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4868664741516113,
+      "learning_rate": 9.96984649122807e-06,
+      "loss": 1.0926,
+      "step": 10977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6847660541534424,
+      "learning_rate": 9.968932748538013e-06,
+      "loss": 1.1716,
+      "step": 10978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8523075580596924,
+      "learning_rate": 9.968019005847955e-06,
+      "loss": 1.8594,
+      "step": 10979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.809959650039673,
+      "learning_rate": 9.967105263157895e-06,
+      "loss": 1.5129,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2722697257995605,
+      "learning_rate": 9.966191520467838e-06,
+      "loss": 1.9196,
+      "step": 10981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7113490104675293,
+      "learning_rate": 9.965277777777778e-06,
+      "loss": 1.5607,
+      "step": 10982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.475499153137207,
+      "learning_rate": 9.96436403508772e-06,
+      "loss": 1.2073,
+      "step": 10983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8417134284973145,
+      "learning_rate": 9.963450292397661e-06,
+      "loss": 1.4095,
+      "step": 10984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.510714530944824,
+      "learning_rate": 9.962536549707603e-06,
+      "loss": 1.4493,
+      "step": 10985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5518252849578857,
+      "learning_rate": 9.961622807017545e-06,
+      "loss": 1.6698,
+      "step": 10986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.758763074874878,
+      "learning_rate": 9.960709064327486e-06,
+      "loss": 1.5358,
+      "step": 10987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7527849674224854,
+      "learning_rate": 9.959795321637428e-06,
+      "loss": 1.2986,
+      "step": 10988
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3172619342803955,
+      "learning_rate": 9.958881578947369e-06,
+      "loss": 1.4493,
+      "step": 10989
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.636207342147827,
+      "learning_rate": 9.95796783625731e-06,
+      "loss": 1.438,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5291664600372314,
+      "learning_rate": 9.957054093567251e-06,
+      "loss": 1.4593,
+      "step": 10991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4541049003601074,
+      "learning_rate": 9.956140350877194e-06,
+      "loss": 1.2693,
+      "step": 10992
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3802804946899414,
+      "learning_rate": 9.955226608187136e-06,
+      "loss": 1.3954,
+      "step": 10993
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1707797050476074,
+      "learning_rate": 9.954312865497076e-06,
+      "loss": 1.0183,
+      "step": 10994
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9520466327667236,
+      "learning_rate": 9.953399122807019e-06,
+      "loss": 1.7632,
+      "step": 10995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6652793884277344,
+      "learning_rate": 9.952485380116959e-06,
+      "loss": 1.5663,
+      "step": 10996
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5437510013580322,
+      "learning_rate": 9.951571637426901e-06,
+      "loss": 1.6577,
+      "step": 10997
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.082970142364502,
+      "learning_rate": 9.950657894736842e-06,
+      "loss": 1.9277,
+      "step": 10998
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5860061645507812,
+      "learning_rate": 9.949744152046784e-06,
+      "loss": 1.6404,
+      "step": 10999
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8394408226013184,
+      "learning_rate": 9.948830409356726e-06,
+      "loss": 1.7216,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6052322387695312,
+      "learning_rate": 9.947916666666667e-06,
+      "loss": 1.8636,
+      "step": 11001
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7982141971588135,
+      "learning_rate": 9.947002923976609e-06,
+      "loss": 1.7667,
+      "step": 11002
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.197585344314575,
+      "learning_rate": 9.94608918128655e-06,
+      "loss": 1.3878,
+      "step": 11003
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.602652072906494,
+      "learning_rate": 9.945175438596492e-06,
+      "loss": 1.8475,
+      "step": 11004
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4811415672302246,
+      "learning_rate": 9.944261695906432e-06,
+      "loss": 1.6635,
+      "step": 11005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3493471145629883,
+      "learning_rate": 9.943347953216375e-06,
+      "loss": 1.314,
+      "step": 11006
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5023813247680664,
+      "learning_rate": 9.942434210526317e-06,
+      "loss": 1.279,
+      "step": 11007
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5086262226104736,
+      "learning_rate": 9.941520467836257e-06,
+      "loss": 1.5017,
+      "step": 11008
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.673765182495117,
+      "learning_rate": 9.9406067251462e-06,
+      "loss": 1.3864,
+      "step": 11009
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9719114303588867,
+      "learning_rate": 9.939692982456142e-06,
+      "loss": 1.3895,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2460641860961914,
+      "learning_rate": 9.938779239766084e-06,
+      "loss": 1.8709,
+      "step": 11011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3988535404205322,
+      "learning_rate": 9.937865497076024e-06,
+      "loss": 1.2067,
+      "step": 11012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.113844633102417,
+      "learning_rate": 9.936951754385967e-06,
+      "loss": 1.6183,
+      "step": 11013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7481350898742676,
+      "learning_rate": 9.936038011695907e-06,
+      "loss": 1.6189,
+      "step": 11014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.275416135787964,
+      "learning_rate": 9.93512426900585e-06,
+      "loss": 1.3807,
+      "step": 11015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3784902095794678,
+      "learning_rate": 9.93421052631579e-06,
+      "loss": 1.4208,
+      "step": 11016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.740760087966919,
+      "learning_rate": 9.933296783625732e-06,
+      "loss": 1.7832,
+      "step": 11017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6785943508148193,
+      "learning_rate": 9.932383040935674e-06,
+      "loss": 1.4467,
+      "step": 11018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0582597255706787,
+      "learning_rate": 9.931469298245615e-06,
+      "loss": 1.757,
+      "step": 11019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.44982647895813,
+      "learning_rate": 9.930555555555557e-06,
+      "loss": 1.3392,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3195481300354004,
+      "learning_rate": 9.929641812865498e-06,
+      "loss": 1.7698,
+      "step": 11021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4881746768951416,
+      "learning_rate": 9.92872807017544e-06,
+      "loss": 1.5079,
+      "step": 11022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2299561500549316,
+      "learning_rate": 9.92781432748538e-06,
+      "loss": 1.0638,
+      "step": 11023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3644654750823975,
+      "learning_rate": 9.926900584795323e-06,
+      "loss": 1.2691,
+      "step": 11024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1858859062194824,
+      "learning_rate": 9.925986842105265e-06,
+      "loss": 1.2477,
+      "step": 11025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.252446413040161,
+      "learning_rate": 9.925073099415205e-06,
+      "loss": 1.1836,
+      "step": 11026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.178835153579712,
+      "learning_rate": 9.924159356725148e-06,
+      "loss": 1.3019,
+      "step": 11027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2584712505340576,
+      "learning_rate": 9.923245614035088e-06,
+      "loss": 1.4984,
+      "step": 11028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3611230850219727,
+      "learning_rate": 9.92233187134503e-06,
+      "loss": 1.4002,
+      "step": 11029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3223681449890137,
+      "learning_rate": 9.921418128654971e-06,
+      "loss": 1.264,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.460803508758545,
+      "learning_rate": 9.920504385964913e-06,
+      "loss": 1.2516,
+      "step": 11031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.830230951309204,
+      "learning_rate": 9.919590643274855e-06,
+      "loss": 1.6069,
+      "step": 11032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.006349563598633,
+      "learning_rate": 9.918676900584796e-06,
+      "loss": 1.7892,
+      "step": 11033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7881908416748047,
+      "learning_rate": 9.917763157894738e-06,
+      "loss": 1.7543,
+      "step": 11034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.407104730606079,
+      "learning_rate": 9.916849415204679e-06,
+      "loss": 1.4347,
+      "step": 11035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8206818103790283,
+      "learning_rate": 9.91593567251462e-06,
+      "loss": 1.3898,
+      "step": 11036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2760326862335205,
+      "learning_rate": 9.915021929824561e-06,
+      "loss": 1.4059,
+      "step": 11037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.064077138900757,
+      "learning_rate": 9.914108187134504e-06,
+      "loss": 1.5418,
+      "step": 11038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.842862606048584,
+      "learning_rate": 9.913194444444446e-06,
+      "loss": 1.4118,
+      "step": 11039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9157824516296387,
+      "learning_rate": 9.912280701754386e-06,
+      "loss": 1.8555,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7960033416748047,
+      "learning_rate": 9.911366959064329e-06,
+      "loss": 1.4895,
+      "step": 11041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6939361095428467,
+      "learning_rate": 9.910453216374269e-06,
+      "loss": 1.4615,
+      "step": 11042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7273688316345215,
+      "learning_rate": 9.909539473684211e-06,
+      "loss": 1.6395,
+      "step": 11043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1254959106445312,
+      "learning_rate": 9.908625730994152e-06,
+      "loss": 1.9406,
+      "step": 11044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4586057662963867,
+      "learning_rate": 9.907711988304094e-06,
+      "loss": 1.5684,
+      "step": 11045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2471394538879395,
+      "learning_rate": 9.906798245614036e-06,
+      "loss": 1.2441,
+      "step": 11046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.765718460083008,
+      "learning_rate": 9.905884502923977e-06,
+      "loss": 1.636,
+      "step": 11047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4614927768707275,
+      "learning_rate": 9.904970760233919e-06,
+      "loss": 1.092,
+      "step": 11048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3760738372802734,
+      "learning_rate": 9.90405701754386e-06,
+      "loss": 1.4993,
+      "step": 11049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9953455924987793,
+      "learning_rate": 9.903143274853802e-06,
+      "loss": 1.1288,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5664055347442627,
+      "learning_rate": 9.902229532163744e-06,
+      "loss": 1.8942,
+      "step": 11051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7789089679718018,
+      "learning_rate": 9.901315789473686e-06,
+      "loss": 1.4536,
+      "step": 11052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.3955912590026855,
+      "learning_rate": 9.900402046783627e-06,
+      "loss": 1.3799,
+      "step": 11053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3184115886688232,
+      "learning_rate": 9.899488304093569e-06,
+      "loss": 1.2331,
+      "step": 11054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.425961971282959,
+      "learning_rate": 9.89857456140351e-06,
+      "loss": 1.2893,
+      "step": 11055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.230025291442871,
+      "learning_rate": 9.897660818713452e-06,
+      "loss": 1.2438,
+      "step": 11056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.52268648147583,
+      "learning_rate": 9.896747076023392e-06,
+      "loss": 1.5147,
+      "step": 11057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2447874546051025,
+      "learning_rate": 9.895833333333334e-06,
+      "loss": 1.2445,
+      "step": 11058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5323667526245117,
+      "learning_rate": 9.894919590643277e-06,
+      "loss": 1.5964,
+      "step": 11059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9891514778137207,
+      "learning_rate": 9.894005847953217e-06,
+      "loss": 1.3711,
+      "step": 11060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.612774133682251,
+      "learning_rate": 9.89309210526316e-06,
+      "loss": 1.4604,
+      "step": 11061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4326798915863037,
+      "learning_rate": 9.8921783625731e-06,
+      "loss": 1.5572,
+      "step": 11062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.279714822769165,
+      "learning_rate": 9.891264619883042e-06,
+      "loss": 1.3449,
+      "step": 11063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6594793796539307,
+      "learning_rate": 9.890350877192983e-06,
+      "loss": 1.3639,
+      "step": 11064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.946251392364502,
+      "learning_rate": 9.889437134502925e-06,
+      "loss": 1.6858,
+      "step": 11065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.285019874572754,
+      "learning_rate": 9.888523391812867e-06,
+      "loss": 1.4288,
+      "step": 11066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4964327812194824,
+      "learning_rate": 9.887609649122808e-06,
+      "loss": 1.3314,
+      "step": 11067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6634998321533203,
+      "learning_rate": 9.88669590643275e-06,
+      "loss": 1.6744,
+      "step": 11068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5119707584381104,
+      "learning_rate": 9.88578216374269e-06,
+      "loss": 1.288,
+      "step": 11069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7581136226654053,
+      "learning_rate": 9.884868421052633e-06,
+      "loss": 1.401,
+      "step": 11070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4341907501220703,
+      "learning_rate": 9.883954678362573e-06,
+      "loss": 1.598,
+      "step": 11071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.160808563232422,
+      "learning_rate": 9.883040935672515e-06,
+      "loss": 1.1803,
+      "step": 11072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3620150089263916,
+      "learning_rate": 9.882127192982458e-06,
+      "loss": 1.4902,
+      "step": 11073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.830901861190796,
+      "learning_rate": 9.881213450292398e-06,
+      "loss": 1.6552,
+      "step": 11074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.402625322341919,
+      "learning_rate": 9.88029970760234e-06,
+      "loss": 1.8676,
+      "step": 11075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.226336717605591,
+      "learning_rate": 9.879385964912281e-06,
+      "loss": 1.4386,
+      "step": 11076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1076483726501465,
+      "learning_rate": 9.878472222222223e-06,
+      "loss": 1.9295,
+      "step": 11077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.355164051055908,
+      "learning_rate": 9.877558479532164e-06,
+      "loss": 1.2962,
+      "step": 11078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.754554271697998,
+      "learning_rate": 9.876644736842106e-06,
+      "loss": 1.4585,
+      "step": 11079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.606537342071533,
+      "learning_rate": 9.875730994152048e-06,
+      "loss": 1.6392,
+      "step": 11080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.384974479675293,
+      "learning_rate": 9.874817251461989e-06,
+      "loss": 1.7166,
+      "step": 11081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.755925178527832,
+      "learning_rate": 9.87390350877193e-06,
+      "loss": 1.4312,
+      "step": 11082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0296554565429688,
+      "learning_rate": 9.872989766081871e-06,
+      "loss": 1.5036,
+      "step": 11083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.31366229057312,
+      "learning_rate": 9.872076023391814e-06,
+      "loss": 1.4971,
+      "step": 11084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4898250102996826,
+      "learning_rate": 9.871162280701754e-06,
+      "loss": 1.623,
+      "step": 11085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.840562343597412,
+      "learning_rate": 9.870248538011696e-06,
+      "loss": 1.8141,
+      "step": 11086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5553882122039795,
+      "learning_rate": 9.869334795321639e-06,
+      "loss": 1.2914,
+      "step": 11087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6110000610351562,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 1.592,
+      "step": 11088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2584826946258545,
+      "learning_rate": 9.867507309941521e-06,
+      "loss": 1.3061,
+      "step": 11089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.689908504486084,
+      "learning_rate": 9.866593567251462e-06,
+      "loss": 2.0001,
+      "step": 11090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4208691120147705,
+      "learning_rate": 9.865679824561404e-06,
+      "loss": 1.5243,
+      "step": 11091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.250774621963501,
+      "learning_rate": 9.864766081871345e-06,
+      "loss": 1.4462,
+      "step": 11092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4983878135681152,
+      "learning_rate": 9.863852339181287e-06,
+      "loss": 1.5184,
+      "step": 11093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.572784423828125,
+      "learning_rate": 9.862938596491229e-06,
+      "loss": 1.4726,
+      "step": 11094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.290468454360962,
+      "learning_rate": 9.862024853801171e-06,
+      "loss": 1.2701,
+      "step": 11095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.61505389213562,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 1.4288,
+      "step": 11096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.612595319747925,
+      "learning_rate": 9.860197368421054e-06,
+      "loss": 1.5425,
+      "step": 11097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8129029273986816,
+      "learning_rate": 9.859283625730996e-06,
+      "loss": 1.7516,
+      "step": 11098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7697622776031494,
+      "learning_rate": 9.858369883040937e-06,
+      "loss": 1.5186,
+      "step": 11099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2666664123535156,
+      "learning_rate": 9.857456140350879e-06,
+      "loss": 1.2908,
+      "step": 11100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4666879177093506,
+      "learning_rate": 9.85654239766082e-06,
+      "loss": 1.2208,
+      "step": 11101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.920041084289551,
+      "learning_rate": 9.855628654970762e-06,
+      "loss": 1.7565,
+      "step": 11102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4712722301483154,
+      "learning_rate": 9.854714912280702e-06,
+      "loss": 1.8601,
+      "step": 11103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7679991722106934,
+      "learning_rate": 9.853801169590644e-06,
+      "loss": 1.7291,
+      "step": 11104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.533782720565796,
+      "learning_rate": 9.852887426900587e-06,
+      "loss": 1.5208,
+      "step": 11105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.674186944961548,
+      "learning_rate": 9.851973684210527e-06,
+      "loss": 1.8764,
+      "step": 11106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7998735904693604,
+      "learning_rate": 9.85105994152047e-06,
+      "loss": 1.6294,
+      "step": 11107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0696935653686523,
+      "learning_rate": 9.85014619883041e-06,
+      "loss": 1.3679,
+      "step": 11108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.901989459991455,
+      "learning_rate": 9.849232456140352e-06,
+      "loss": 1.4548,
+      "step": 11109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7154910564422607,
+      "learning_rate": 9.848318713450293e-06,
+      "loss": 1.5535,
+      "step": 11110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5265398025512695,
+      "learning_rate": 9.847404970760235e-06,
+      "loss": 1.5632,
+      "step": 11111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.082237958908081,
+      "learning_rate": 9.846491228070177e-06,
+      "loss": 1.1965,
+      "step": 11112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2919259071350098,
+      "learning_rate": 9.845577485380118e-06,
+      "loss": 1.4061,
+      "step": 11113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.747373342514038,
+      "learning_rate": 9.84466374269006e-06,
+      "loss": 1.592,
+      "step": 11114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0839250087738037,
+      "learning_rate": 9.84375e-06,
+      "loss": 1.2999,
+      "step": 11115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.649200439453125,
+      "learning_rate": 9.842836257309943e-06,
+      "loss": 1.387,
+      "step": 11116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6451010704040527,
+      "learning_rate": 9.841922514619883e-06,
+      "loss": 1.7879,
+      "step": 11117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.929834246635437,
+      "learning_rate": 9.841008771929825e-06,
+      "loss": 1.1365,
+      "step": 11118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.058509349822998,
+      "learning_rate": 9.840095029239768e-06,
+      "loss": 1.1297,
+      "step": 11119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.680366039276123,
+      "learning_rate": 9.839181286549708e-06,
+      "loss": 1.6139,
+      "step": 11120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4107797145843506,
+      "learning_rate": 9.83826754385965e-06,
+      "loss": 1.2696,
+      "step": 11121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.641897678375244,
+      "learning_rate": 9.837353801169591e-06,
+      "loss": 1.4163,
+      "step": 11122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.695734977722168,
+      "learning_rate": 9.836440058479533e-06,
+      "loss": 1.3922,
+      "step": 11123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3676555156707764,
+      "learning_rate": 9.835526315789474e-06,
+      "loss": 1.4257,
+      "step": 11124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5718159675598145,
+      "learning_rate": 9.834612573099416e-06,
+      "loss": 1.6648,
+      "step": 11125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1981582641601562,
+      "learning_rate": 9.833698830409358e-06,
+      "loss": 1.1434,
+      "step": 11126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.747328042984009,
+      "learning_rate": 9.832785087719299e-06,
+      "loss": 1.8,
+      "step": 11127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.93519926071167,
+      "learning_rate": 9.83187134502924e-06,
+      "loss": 1.8711,
+      "step": 11128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6109156608581543,
+      "learning_rate": 9.830957602339181e-06,
+      "loss": 1.395,
+      "step": 11129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2938828468322754,
+      "learning_rate": 9.830043859649124e-06,
+      "loss": 1.291,
+      "step": 11130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.732703924179077,
+      "learning_rate": 9.829130116959064e-06,
+      "loss": 1.3974,
+      "step": 11131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.27274751663208,
+      "learning_rate": 9.828216374269006e-06,
+      "loss": 1.9528,
+      "step": 11132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5417444705963135,
+      "learning_rate": 9.827302631578949e-06,
+      "loss": 1.5013,
+      "step": 11133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8093016147613525,
+      "learning_rate": 9.826388888888889e-06,
+      "loss": 1.7068,
+      "step": 11134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.919367790222168,
+      "learning_rate": 9.825475146198831e-06,
+      "loss": 1.5889,
+      "step": 11135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.765162467956543,
+      "learning_rate": 9.824561403508772e-06,
+      "loss": 1.6305,
+      "step": 11136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6221392154693604,
+      "learning_rate": 9.823647660818714e-06,
+      "loss": 1.2744,
+      "step": 11137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4899067878723145,
+      "learning_rate": 9.822733918128656e-06,
+      "loss": 1.1254,
+      "step": 11138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.597041606903076,
+      "learning_rate": 9.821820175438598e-06,
+      "loss": 1.5298,
+      "step": 11139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.707172393798828,
+      "learning_rate": 9.820906432748539e-06,
+      "loss": 1.6183,
+      "step": 11140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.706455945968628,
+      "learning_rate": 9.819992690058481e-06,
+      "loss": 1.8417,
+      "step": 11141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.748915910720825,
+      "learning_rate": 9.819078947368422e-06,
+      "loss": 1.6236,
+      "step": 11142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.545222520828247,
+      "learning_rate": 9.818165204678364e-06,
+      "loss": 1.5718,
+      "step": 11143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9756674766540527,
+      "learning_rate": 9.817251461988305e-06,
+      "loss": 1.5935,
+      "step": 11144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8553593158721924,
+      "learning_rate": 9.816337719298247e-06,
+      "loss": 1.6497,
+      "step": 11145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.044281005859375,
+      "learning_rate": 9.815423976608189e-06,
+      "loss": 1.7776,
+      "step": 11146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6086833477020264,
+      "learning_rate": 9.81451023391813e-06,
+      "loss": 1.3263,
+      "step": 11147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7477853298187256,
+      "learning_rate": 9.813596491228072e-06,
+      "loss": 1.7714,
+      "step": 11148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3064324855804443,
+      "learning_rate": 9.812682748538012e-06,
+      "loss": 1.2323,
+      "step": 11149
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.449878692626953,
+      "learning_rate": 9.811769005847954e-06,
+      "loss": 1.5187,
+      "step": 11150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.70434832572937,
+      "learning_rate": 9.810855263157895e-06,
+      "loss": 1.5703,
+      "step": 11151
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.99876868724823,
+      "learning_rate": 9.809941520467837e-06,
+      "loss": 1.135,
+      "step": 11152
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4913980960845947,
+      "learning_rate": 9.80902777777778e-06,
+      "loss": 1.625,
+      "step": 11153
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.784186840057373,
+      "learning_rate": 9.80811403508772e-06,
+      "loss": 1.5996,
+      "step": 11154
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.52898907661438,
+      "learning_rate": 9.807200292397662e-06,
+      "loss": 1.2568,
+      "step": 11155
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8174118995666504,
+      "learning_rate": 9.806286549707603e-06,
+      "loss": 1.5289,
+      "step": 11156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5405123233795166,
+      "learning_rate": 9.805372807017545e-06,
+      "loss": 1.4924,
+      "step": 11157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0840237140655518,
+      "learning_rate": 9.804459064327485e-06,
+      "loss": 1.3826,
+      "step": 11158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.55741024017334,
+      "learning_rate": 9.803545321637428e-06,
+      "loss": 1.4755,
+      "step": 11159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.998729944229126,
+      "learning_rate": 9.80263157894737e-06,
+      "loss": 1.0944,
+      "step": 11160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.382517099380493,
+      "learning_rate": 9.80171783625731e-06,
+      "loss": 1.3733,
+      "step": 11161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.483569622039795,
+      "learning_rate": 9.800804093567253e-06,
+      "loss": 1.4187,
+      "step": 11162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9070513248443604,
+      "learning_rate": 9.799890350877193e-06,
+      "loss": 2.0313,
+      "step": 11163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6475048065185547,
+      "learning_rate": 9.798976608187135e-06,
+      "loss": 1.4338,
+      "step": 11164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3268415927886963,
+      "learning_rate": 9.798062865497076e-06,
+      "loss": 1.3563,
+      "step": 11165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.526718854904175,
+      "learning_rate": 9.797149122807018e-06,
+      "loss": 1.7206,
+      "step": 11166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0308001041412354,
+      "learning_rate": 9.79623538011696e-06,
+      "loss": 1.8896,
+      "step": 11167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3755452632904053,
+      "learning_rate": 9.795321637426901e-06,
+      "loss": 1.3848,
+      "step": 11168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.191661834716797,
+      "learning_rate": 9.794407894736843e-06,
+      "loss": 1.3504,
+      "step": 11169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.863131046295166,
+      "learning_rate": 9.793494152046784e-06,
+      "loss": 1.8509,
+      "step": 11170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2713592052459717,
+      "learning_rate": 9.792580409356726e-06,
+      "loss": 1.7224,
+      "step": 11171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.329387664794922,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 1.2097,
+      "step": 11172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7551229000091553,
+      "learning_rate": 9.790752923976609e-06,
+      "loss": 1.0292,
+      "step": 11173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.141111135482788,
+      "learning_rate": 9.78983918128655e-06,
+      "loss": 1.2176,
+      "step": 11174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5001771450042725,
+      "learning_rate": 9.788925438596491e-06,
+      "loss": 1.4191,
+      "step": 11175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7172186374664307,
+      "learning_rate": 9.788011695906434e-06,
+      "loss": 1.8139,
+      "step": 11176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.595672607421875,
+      "learning_rate": 9.787097953216374e-06,
+      "loss": 1.7269,
+      "step": 11177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4781672954559326,
+      "learning_rate": 9.786184210526316e-06,
+      "loss": 1.3927,
+      "step": 11178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2997546195983887,
+      "learning_rate": 9.785270467836257e-06,
+      "loss": 1.5008,
+      "step": 11179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.458543539047241,
+      "learning_rate": 9.7843567251462e-06,
+      "loss": 1.6608,
+      "step": 11180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4218380451202393,
+      "learning_rate": 9.783442982456141e-06,
+      "loss": 1.4348,
+      "step": 11181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.253934621810913,
+      "learning_rate": 9.782529239766084e-06,
+      "loss": 1.0711,
+      "step": 11182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2414095401763916,
+      "learning_rate": 9.781615497076024e-06,
+      "loss": 1.2418,
+      "step": 11183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7913594245910645,
+      "learning_rate": 9.780701754385966e-06,
+      "loss": 1.9699,
+      "step": 11184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4833202362060547,
+      "learning_rate": 9.779788011695908e-06,
+      "loss": 1.402,
+      "step": 11185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0805885791778564,
+      "learning_rate": 9.778874269005849e-06,
+      "loss": 1.679,
+      "step": 11186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9431393146514893,
+      "learning_rate": 9.777960526315791e-06,
+      "loss": 1.9531,
+      "step": 11187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3496904373168945,
+      "learning_rate": 9.777046783625732e-06,
+      "loss": 1.3047,
+      "step": 11188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.689568042755127,
+      "learning_rate": 9.776133040935674e-06,
+      "loss": 1.447,
+      "step": 11189
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.451564311981201,
+      "learning_rate": 9.775219298245615e-06,
+      "loss": 1.3833,
+      "step": 11190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.776219129562378,
+      "learning_rate": 9.774305555555557e-06,
+      "loss": 1.4314,
+      "step": 11191
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7798867225646973,
+      "learning_rate": 9.773391812865497e-06,
+      "loss": 1.8811,
+      "step": 11192
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1661131381988525,
+      "learning_rate": 9.77247807017544e-06,
+      "loss": 1.7999,
+      "step": 11193
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.045259714126587,
+      "learning_rate": 9.771564327485382e-06,
+      "loss": 1.2508,
+      "step": 11194
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9862003326416016,
+      "learning_rate": 9.770650584795322e-06,
+      "loss": 1.6373,
+      "step": 11195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7947945594787598,
+      "learning_rate": 9.769736842105264e-06,
+      "loss": 1.5096,
+      "step": 11196
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5782673358917236,
+      "learning_rate": 9.768823099415205e-06,
+      "loss": 1.528,
+      "step": 11197
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9949910640716553,
+      "learning_rate": 9.767909356725147e-06,
+      "loss": 1.7405,
+      "step": 11198
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.836287498474121,
+      "learning_rate": 9.766995614035088e-06,
+      "loss": 1.4414,
+      "step": 11199
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6927154064178467,
+      "learning_rate": 9.76608187134503e-06,
+      "loss": 1.4675,
+      "step": 11200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5370936393737793,
+      "learning_rate": 9.765168128654972e-06,
+      "loss": 1.4863,
+      "step": 11201
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2003424167633057,
+      "learning_rate": 9.764254385964913e-06,
+      "loss": 1.9258,
+      "step": 11202
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3147547245025635,
+      "learning_rate": 9.763340643274855e-06,
+      "loss": 1.302,
+      "step": 11203
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2510273456573486,
+      "learning_rate": 9.762426900584795e-06,
+      "loss": 1.3101,
+      "step": 11204
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.878493547439575,
+      "learning_rate": 9.761513157894738e-06,
+      "loss": 1.5925,
+      "step": 11205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.64401912689209,
+      "learning_rate": 9.760599415204678e-06,
+      "loss": 1.4594,
+      "step": 11206
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.122631072998047,
+      "learning_rate": 9.75968567251462e-06,
+      "loss": 1.7769,
+      "step": 11207
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3216917514801025,
+      "learning_rate": 9.758771929824563e-06,
+      "loss": 1.5004,
+      "step": 11208
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.410745143890381,
+      "learning_rate": 9.757858187134503e-06,
+      "loss": 1.5472,
+      "step": 11209
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1318981647491455,
+      "learning_rate": 9.756944444444445e-06,
+      "loss": 1.0323,
+      "step": 11210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.342785358428955,
+      "learning_rate": 9.756030701754386e-06,
+      "loss": 1.4096,
+      "step": 11211
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.142498016357422,
+      "learning_rate": 9.755116959064328e-06,
+      "loss": 1.4483,
+      "step": 11212
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5757036209106445,
+      "learning_rate": 9.754203216374269e-06,
+      "loss": 1.4819,
+      "step": 11213
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7584285736083984,
+      "learning_rate": 9.753289473684211e-06,
+      "loss": 1.3299,
+      "step": 11214
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.555208444595337,
+      "learning_rate": 9.752375730994153e-06,
+      "loss": 1.3428,
+      "step": 11215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.509713888168335,
+      "learning_rate": 9.751461988304094e-06,
+      "loss": 1.2358,
+      "step": 11216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6634714603424072,
+      "learning_rate": 9.750548245614036e-06,
+      "loss": 1.6924,
+      "step": 11217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0222156047821045,
+      "learning_rate": 9.749634502923976e-06,
+      "loss": 1.263,
+      "step": 11218
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9295637607574463,
+      "learning_rate": 9.748720760233919e-06,
+      "loss": 1.8142,
+      "step": 11219
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9020092487335205,
+      "learning_rate": 9.747807017543859e-06,
+      "loss": 1.545,
+      "step": 11220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.275303840637207,
+      "learning_rate": 9.746893274853801e-06,
+      "loss": 1.102,
+      "step": 11221
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8439249992370605,
+      "learning_rate": 9.745979532163744e-06,
+      "loss": 1.4333,
+      "step": 11222
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.580559730529785,
+      "learning_rate": 9.745065789473686e-06,
+      "loss": 1.7007,
+      "step": 11223
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8371548652648926,
+      "learning_rate": 9.744152046783626e-06,
+      "loss": 1.4716,
+      "step": 11224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7008910179138184,
+      "learning_rate": 9.743238304093569e-06,
+      "loss": 1.4963,
+      "step": 11225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3254382610321045,
+      "learning_rate": 9.74232456140351e-06,
+      "loss": 1.3889,
+      "step": 11226
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.599501609802246,
+      "learning_rate": 9.741410818713451e-06,
+      "loss": 1.2749,
+      "step": 11227
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.615316152572632,
+      "learning_rate": 9.740497076023394e-06,
+      "loss": 1.8856,
+      "step": 11228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4904885292053223,
+      "learning_rate": 9.739583333333334e-06,
+      "loss": 1.4252,
+      "step": 11229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.371931314468384,
+      "learning_rate": 9.738669590643276e-06,
+      "loss": 1.4375,
+      "step": 11230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.76693058013916,
+      "learning_rate": 9.737755847953217e-06,
+      "loss": 1.4657,
+      "step": 11231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3794331550598145,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 1.4784,
+      "step": 11232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.627026081085205,
+      "learning_rate": 9.735928362573101e-06,
+      "loss": 1.5821,
+      "step": 11233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.878023386001587,
+      "learning_rate": 9.735014619883042e-06,
+      "loss": 1.6823,
+      "step": 11234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6419386863708496,
+      "learning_rate": 9.734100877192984e-06,
+      "loss": 1.7158,
+      "step": 11235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3292059898376465,
+      "learning_rate": 9.733187134502925e-06,
+      "loss": 1.3305,
+      "step": 11236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.234919309616089,
+      "learning_rate": 9.732273391812867e-06,
+      "loss": 1.0717,
+      "step": 11237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0357859134674072,
+      "learning_rate": 9.731359649122807e-06,
+      "loss": 1.2407,
+      "step": 11238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2240939140319824,
+      "learning_rate": 9.73044590643275e-06,
+      "loss": 1.2356,
+      "step": 11239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5764622688293457,
+      "learning_rate": 9.729532163742692e-06,
+      "loss": 1.3745,
+      "step": 11240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7097103595733643,
+      "learning_rate": 9.728618421052632e-06,
+      "loss": 1.5774,
+      "step": 11241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.741924285888672,
+      "learning_rate": 9.727704678362574e-06,
+      "loss": 1.697,
+      "step": 11242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.005967855453491,
+      "learning_rate": 9.726790935672515e-06,
+      "loss": 1.4086,
+      "step": 11243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5802078247070312,
+      "learning_rate": 9.725877192982457e-06,
+      "loss": 1.4076,
+      "step": 11244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.882269859313965,
+      "learning_rate": 9.724963450292398e-06,
+      "loss": 1.5739,
+      "step": 11245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.405327081680298,
+      "learning_rate": 9.72404970760234e-06,
+      "loss": 1.5002,
+      "step": 11246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.794809103012085,
+      "learning_rate": 9.723135964912282e-06,
+      "loss": 1.3835,
+      "step": 11247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.832587718963623,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 1.3262,
+      "step": 11248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2940549850463867,
+      "learning_rate": 9.721308479532165e-06,
+      "loss": 1.306,
+      "step": 11249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.719806432723999,
+      "learning_rate": 9.720394736842105e-06,
+      "loss": 1.508,
+      "step": 11250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5130691528320312,
+      "learning_rate": 9.719480994152048e-06,
+      "loss": 1.5062,
+      "step": 11251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.793213129043579,
+      "learning_rate": 9.718567251461988e-06,
+      "loss": 1.8058,
+      "step": 11252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8212172985076904,
+      "learning_rate": 9.71765350877193e-06,
+      "loss": 1.7692,
+      "step": 11253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5304243564605713,
+      "learning_rate": 9.716739766081873e-06,
+      "loss": 1.3276,
+      "step": 11254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.678046941757202,
+      "learning_rate": 9.715826023391813e-06,
+      "loss": 1.3316,
+      "step": 11255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.893630266189575,
+      "learning_rate": 9.714912280701755e-06,
+      "loss": 1.576,
+      "step": 11256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.651790142059326,
+      "learning_rate": 9.713998538011696e-06,
+      "loss": 1.4362,
+      "step": 11257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4884915351867676,
+      "learning_rate": 9.713084795321638e-06,
+      "loss": 1.4173,
+      "step": 11258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.387237548828125,
+      "learning_rate": 9.712171052631579e-06,
+      "loss": 1.3176,
+      "step": 11259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7639424800872803,
+      "learning_rate": 9.711257309941521e-06,
+      "loss": 1.3894,
+      "step": 11260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1595969200134277,
+      "learning_rate": 9.710343567251463e-06,
+      "loss": 1.1114,
+      "step": 11261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.404039144515991,
+      "learning_rate": 9.709429824561404e-06,
+      "loss": 1.0447,
+      "step": 11262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1355578899383545,
+      "learning_rate": 9.708516081871346e-06,
+      "loss": 1.9069,
+      "step": 11263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5452704429626465,
+      "learning_rate": 9.707602339181286e-06,
+      "loss": 1.5343,
+      "step": 11264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3483686447143555,
+      "learning_rate": 9.706688596491229e-06,
+      "loss": 1.879,
+      "step": 11265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.390066623687744,
+      "learning_rate": 9.70577485380117e-06,
+      "loss": 1.5404,
+      "step": 11266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9725520610809326,
+      "learning_rate": 9.704861111111113e-06,
+      "loss": 1.5701,
+      "step": 11267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4029769897460938,
+      "learning_rate": 9.703947368421054e-06,
+      "loss": 1.2816,
+      "step": 11268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.573756694793701,
+      "learning_rate": 9.703033625730996e-06,
+      "loss": 1.4015,
+      "step": 11269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6718432903289795,
+      "learning_rate": 9.702119883040936e-06,
+      "loss": 1.4371,
+      "step": 11270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9045255184173584,
+      "learning_rate": 9.701206140350879e-06,
+      "loss": 1.186,
+      "step": 11271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.467757225036621,
+      "learning_rate": 9.700292397660819e-06,
+      "loss": 1.4876,
+      "step": 11272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8705315589904785,
+      "learning_rate": 9.699378654970761e-06,
+      "loss": 1.0729,
+      "step": 11273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.743238925933838,
+      "learning_rate": 9.698464912280704e-06,
+      "loss": 1.4116,
+      "step": 11274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4329652786254883,
+      "learning_rate": 9.697551169590644e-06,
+      "loss": 1.6268,
+      "step": 11275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6229403018951416,
+      "learning_rate": 9.696637426900586e-06,
+      "loss": 1.4109,
+      "step": 11276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.421302080154419,
+      "learning_rate": 9.695723684210527e-06,
+      "loss": 1.5406,
+      "step": 11277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6105499267578125,
+      "learning_rate": 9.694809941520469e-06,
+      "loss": 1.7251,
+      "step": 11278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2153713703155518,
+      "learning_rate": 9.69389619883041e-06,
+      "loss": 2.1438,
+      "step": 11279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4492437839508057,
+      "learning_rate": 9.692982456140352e-06,
+      "loss": 1.5922,
+      "step": 11280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.273364782333374,
+      "learning_rate": 9.692068713450294e-06,
+      "loss": 1.5872,
+      "step": 11281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6292450428009033,
+      "learning_rate": 9.691154970760235e-06,
+      "loss": 1.3238,
+      "step": 11282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5042660236358643,
+      "learning_rate": 9.690241228070177e-06,
+      "loss": 1.3396,
+      "step": 11283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.490842342376709,
+      "learning_rate": 9.689327485380117e-06,
+      "loss": 1.7733,
+      "step": 11284
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.827833890914917,
+      "learning_rate": 9.68841374269006e-06,
+      "loss": 1.5511,
+      "step": 11285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7276649475097656,
+      "learning_rate": 9.6875e-06,
+      "loss": 1.2036,
+      "step": 11286
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.787299871444702,
+      "learning_rate": 9.686586257309942e-06,
+      "loss": 1.648,
+      "step": 11287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.153902769088745,
+      "learning_rate": 9.685672514619884e-06,
+      "loss": 1.4516,
+      "step": 11288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.759549856185913,
+      "learning_rate": 9.684758771929825e-06,
+      "loss": 1.9372,
+      "step": 11289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.728407382965088,
+      "learning_rate": 9.683845029239767e-06,
+      "loss": 1.4392,
+      "step": 11290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2873940467834473,
+      "learning_rate": 9.682931286549708e-06,
+      "loss": 1.3877,
+      "step": 11291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.336345911026001,
+      "learning_rate": 9.68201754385965e-06,
+      "loss": 1.5493,
+      "step": 11292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.784440279006958,
+      "learning_rate": 9.68110380116959e-06,
+      "loss": 1.836,
+      "step": 11293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6794521808624268,
+      "learning_rate": 9.680190058479533e-06,
+      "loss": 1.3793,
+      "step": 11294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5333971977233887,
+      "learning_rate": 9.679276315789475e-06,
+      "loss": 1.4049,
+      "step": 11295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.023639440536499,
+      "learning_rate": 9.678362573099415e-06,
+      "loss": 1.996,
+      "step": 11296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0840938091278076,
+      "learning_rate": 9.677448830409358e-06,
+      "loss": 1.2663,
+      "step": 11297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.817237615585327,
+      "learning_rate": 9.676535087719298e-06,
+      "loss": 1.3949,
+      "step": 11298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6553189754486084,
+      "learning_rate": 9.67562134502924e-06,
+      "loss": 1.4033,
+      "step": 11299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8745696544647217,
+      "learning_rate": 9.674707602339181e-06,
+      "loss": 1.428,
+      "step": 11300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9042056798934937,
+      "learning_rate": 9.673793859649123e-06,
+      "loss": 1.0825,
+      "step": 11301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1049656867980957,
+      "learning_rate": 9.672880116959065e-06,
+      "loss": 1.9171,
+      "step": 11302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7607598304748535,
+      "learning_rate": 9.671966374269006e-06,
+      "loss": 1.5123,
+      "step": 11303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7473361492156982,
+      "learning_rate": 9.671052631578948e-06,
+      "loss": 1.3119,
+      "step": 11304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.805947780609131,
+      "learning_rate": 9.670138888888889e-06,
+      "loss": 1.6699,
+      "step": 11305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4509482383728027,
+      "learning_rate": 9.669225146198831e-06,
+      "loss": 1.3023,
+      "step": 11306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.506134271621704,
+      "learning_rate": 9.668311403508771e-06,
+      "loss": 1.3779,
+      "step": 11307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.386150598526001,
+      "learning_rate": 9.667397660818715e-06,
+      "loss": 1.5486,
+      "step": 11308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.373067855834961,
+      "learning_rate": 9.666483918128656e-06,
+      "loss": 2.1375,
+      "step": 11309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6231024265289307,
+      "learning_rate": 9.665570175438598e-06,
+      "loss": 1.4064,
+      "step": 11310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0406997203826904,
+      "learning_rate": 9.664656432748539e-06,
+      "loss": 1.8461,
+      "step": 11311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6890718936920166,
+      "learning_rate": 9.66374269005848e-06,
+      "loss": 1.4802,
+      "step": 11312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.051825523376465,
+      "learning_rate": 9.662828947368423e-06,
+      "loss": 1.6149,
+      "step": 11313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9626545906066895,
+      "learning_rate": 9.661915204678364e-06,
+      "loss": 1.777,
+      "step": 11314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3236236572265625,
+      "learning_rate": 9.661001461988306e-06,
+      "loss": 1.3984,
+      "step": 11315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.056617021560669,
+      "learning_rate": 9.660087719298246e-06,
+      "loss": 1.4734,
+      "step": 11316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7046453952789307,
+      "learning_rate": 9.659173976608189e-06,
+      "loss": 1.7166,
+      "step": 11317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4139902591705322,
+      "learning_rate": 9.658260233918129e-06,
+      "loss": 1.477,
+      "step": 11318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9088237285614014,
+      "learning_rate": 9.657346491228071e-06,
+      "loss": 1.6898,
+      "step": 11319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.455141544342041,
+      "learning_rate": 9.656432748538014e-06,
+      "loss": 1.4026,
+      "step": 11320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3124520778656006,
+      "learning_rate": 9.655519005847954e-06,
+      "loss": 1.4028,
+      "step": 11321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.50107741355896,
+      "learning_rate": 9.654605263157896e-06,
+      "loss": 1.4052,
+      "step": 11322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.67866587638855,
+      "learning_rate": 9.653691520467837e-06,
+      "loss": 1.1919,
+      "step": 11323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.301713466644287,
+      "learning_rate": 9.652777777777779e-06,
+      "loss": 1.5209,
+      "step": 11324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2763478755950928,
+      "learning_rate": 9.65186403508772e-06,
+      "loss": 2.0231,
+      "step": 11325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5492184162139893,
+      "learning_rate": 9.650950292397662e-06,
+      "loss": 1.648,
+      "step": 11326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.177684783935547,
+      "learning_rate": 9.650036549707604e-06,
+      "loss": 1.2884,
+      "step": 11327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9197261333465576,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 1.5761,
+      "step": 11328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.483031988143921,
+      "learning_rate": 9.648209064327487e-06,
+      "loss": 1.2937,
+      "step": 11329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0336036682128906,
+      "learning_rate": 9.647295321637427e-06,
+      "loss": 1.6223,
+      "step": 11330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.710543155670166,
+      "learning_rate": 9.64638157894737e-06,
+      "loss": 1.4401,
+      "step": 11331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.119197130203247,
+      "learning_rate": 9.64546783625731e-06,
+      "loss": 1.3894,
+      "step": 11332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1773805618286133,
+      "learning_rate": 9.644554093567252e-06,
+      "loss": 1.5533,
+      "step": 11333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0875611305236816,
+      "learning_rate": 9.643640350877194e-06,
+      "loss": 1.3618,
+      "step": 11334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5267322063446045,
+      "learning_rate": 9.642726608187135e-06,
+      "loss": 1.4293,
+      "step": 11335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.569758415222168,
+      "learning_rate": 9.641812865497077e-06,
+      "loss": 1.2798,
+      "step": 11336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.69389271736145,
+      "learning_rate": 9.640899122807018e-06,
+      "loss": 1.4918,
+      "step": 11337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.780628204345703,
+      "learning_rate": 9.63998538011696e-06,
+      "loss": 1.4888,
+      "step": 11338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0700793266296387,
+      "learning_rate": 9.6390716374269e-06,
+      "loss": 1.0667,
+      "step": 11339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.850130558013916,
+      "learning_rate": 9.638157894736843e-06,
+      "loss": 1.2061,
+      "step": 11340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3105976581573486,
+      "learning_rate": 9.637244152046785e-06,
+      "loss": 1.2994,
+      "step": 11341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.355313539505005,
+      "learning_rate": 9.636330409356725e-06,
+      "loss": 2.0473,
+      "step": 11342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5496010780334473,
+      "learning_rate": 9.635416666666668e-06,
+      "loss": 1.4518,
+      "step": 11343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9861143827438354,
+      "learning_rate": 9.634502923976608e-06,
+      "loss": 1.0321,
+      "step": 11344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4620394706726074,
+      "learning_rate": 9.63358918128655e-06,
+      "loss": 1.7335,
+      "step": 11345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2213666439056396,
+      "learning_rate": 9.632675438596491e-06,
+      "loss": 1.2714,
+      "step": 11346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.311274766921997,
+      "learning_rate": 9.631761695906433e-06,
+      "loss": 1.3811,
+      "step": 11347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7590603828430176,
+      "learning_rate": 9.630847953216375e-06,
+      "loss": 1.4328,
+      "step": 11348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5870308876037598,
+      "learning_rate": 9.629934210526316e-06,
+      "loss": 1.5567,
+      "step": 11349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.69323468208313,
+      "learning_rate": 9.629020467836258e-06,
+      "loss": 1.5651,
+      "step": 11350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4442617893218994,
+      "learning_rate": 9.6281067251462e-06,
+      "loss": 1.4725,
+      "step": 11351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.526275157928467,
+      "learning_rate": 9.627192982456141e-06,
+      "loss": 1.6335,
+      "step": 11352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.460568428039551,
+      "learning_rate": 9.626279239766083e-06,
+      "loss": 1.4681,
+      "step": 11353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7082021236419678,
+      "learning_rate": 9.625365497076025e-06,
+      "loss": 1.4866,
+      "step": 11354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.207448720932007,
+      "learning_rate": 9.624451754385966e-06,
+      "loss": 1.2145,
+      "step": 11355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.78875994682312,
+      "learning_rate": 9.623538011695908e-06,
+      "loss": 1.5017,
+      "step": 11356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7531161308288574,
+      "learning_rate": 9.622624269005849e-06,
+      "loss": 1.382,
+      "step": 11357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.261714458465576,
+      "learning_rate": 9.62171052631579e-06,
+      "loss": 1.3797,
+      "step": 11358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8696022033691406,
+      "learning_rate": 9.620796783625731e-06,
+      "loss": 1.6121,
+      "step": 11359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5054759979248047,
+      "learning_rate": 9.619883040935674e-06,
+      "loss": 1.3458,
+      "step": 11360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.502446413040161,
+      "learning_rate": 9.618969298245616e-06,
+      "loss": 1.6196,
+      "step": 11361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4146294593811035,
+      "learning_rate": 9.618055555555556e-06,
+      "loss": 1.6555,
+      "step": 11362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0084779262542725,
+      "learning_rate": 9.617141812865499e-06,
+      "loss": 1.5655,
+      "step": 11363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6398298740386963,
+      "learning_rate": 9.616228070175439e-06,
+      "loss": 1.5479,
+      "step": 11364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5924715995788574,
+      "learning_rate": 9.615314327485381e-06,
+      "loss": 1.6254,
+      "step": 11365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.390500545501709,
+      "learning_rate": 9.614400584795322e-06,
+      "loss": 1.3482,
+      "step": 11366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.626124382019043,
+      "learning_rate": 9.613486842105264e-06,
+      "loss": 1.4839,
+      "step": 11367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9765256643295288,
+      "learning_rate": 9.612573099415206e-06,
+      "loss": 1.1985,
+      "step": 11368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.603867769241333,
+      "learning_rate": 9.611659356725147e-06,
+      "loss": 1.5816,
+      "step": 11369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6933579444885254,
+      "learning_rate": 9.610745614035089e-06,
+      "loss": 1.2172,
+      "step": 11370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.175246238708496,
+      "learning_rate": 9.60983187134503e-06,
+      "loss": 1.3954,
+      "step": 11371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0685908794403076,
+      "learning_rate": 9.608918128654972e-06,
+      "loss": 1.8062,
+      "step": 11372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7835075855255127,
+      "learning_rate": 9.608004385964912e-06,
+      "loss": 1.8422,
+      "step": 11373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2362771034240723,
+      "learning_rate": 9.607090643274855e-06,
+      "loss": 1.3855,
+      "step": 11374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.362558603286743,
+      "learning_rate": 9.606176900584797e-06,
+      "loss": 1.3982,
+      "step": 11375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.579624652862549,
+      "learning_rate": 9.605263157894737e-06,
+      "loss": 1.5347,
+      "step": 11376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1611289978027344,
+      "learning_rate": 9.60434941520468e-06,
+      "loss": 1.3198,
+      "step": 11377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6026487350463867,
+      "learning_rate": 9.60343567251462e-06,
+      "loss": 1.4322,
+      "step": 11378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.483462333679199,
+      "learning_rate": 9.602521929824562e-06,
+      "loss": 1.4441,
+      "step": 11379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.758582353591919,
+      "learning_rate": 9.601608187134503e-06,
+      "loss": 1.6479,
+      "step": 11380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.594226360321045,
+      "learning_rate": 9.600694444444445e-06,
+      "loss": 1.6336,
+      "step": 11381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.752774953842163,
+      "learning_rate": 9.599780701754387e-06,
+      "loss": 1.8428,
+      "step": 11382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.192128896713257,
+      "learning_rate": 9.598866959064328e-06,
+      "loss": 1.2079,
+      "step": 11383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5372018814086914,
+      "learning_rate": 9.59795321637427e-06,
+      "loss": 1.2533,
+      "step": 11384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6113085746765137,
+      "learning_rate": 9.59703947368421e-06,
+      "loss": 1.5144,
+      "step": 11385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3562326431274414,
+      "learning_rate": 9.596125730994153e-06,
+      "loss": 1.4729,
+      "step": 11386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4675581455230713,
+      "learning_rate": 9.595211988304093e-06,
+      "loss": 2.2354,
+      "step": 11387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5509066581726074,
+      "learning_rate": 9.594298245614035e-06,
+      "loss": 1.6002,
+      "step": 11388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.408424139022827,
+      "learning_rate": 9.593384502923978e-06,
+      "loss": 1.137,
+      "step": 11389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6641347408294678,
+      "learning_rate": 9.592470760233918e-06,
+      "loss": 1.5158,
+      "step": 11390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3291330337524414,
+      "learning_rate": 9.59155701754386e-06,
+      "loss": 1.8389,
+      "step": 11391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4499096870422363,
+      "learning_rate": 9.590643274853801e-06,
+      "loss": 1.3065,
+      "step": 11392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.44948148727417,
+      "learning_rate": 9.589729532163745e-06,
+      "loss": 1.419,
+      "step": 11393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1004414558410645,
+      "learning_rate": 9.588815789473685e-06,
+      "loss": 1.1863,
+      "step": 11394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.05682110786438,
+      "learning_rate": 9.587902046783628e-06,
+      "loss": 1.7285,
+      "step": 11395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0896856784820557,
+      "learning_rate": 9.586988304093568e-06,
+      "loss": 1.8202,
+      "step": 11396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5199625492095947,
+      "learning_rate": 9.58607456140351e-06,
+      "loss": 1.5321,
+      "step": 11397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.240527629852295,
+      "learning_rate": 9.585160818713451e-06,
+      "loss": 1.1097,
+      "step": 11398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7041263580322266,
+      "learning_rate": 9.584247076023393e-06,
+      "loss": 1.6196,
+      "step": 11399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2079954147338867,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 1.4283,
+      "step": 11400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0671515464782715,
+      "learning_rate": 9.582419590643276e-06,
+      "loss": 1.12,
+      "step": 11401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9082770347595215,
+      "learning_rate": 9.581505847953218e-06,
+      "loss": 1.7108,
+      "step": 11402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5095126628875732,
+      "learning_rate": 9.580592105263159e-06,
+      "loss": 1.438,
+      "step": 11403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6130361557006836,
+      "learning_rate": 9.5796783625731e-06,
+      "loss": 1.442,
+      "step": 11404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.632561683654785,
+      "learning_rate": 9.578764619883041e-06,
+      "loss": 1.5043,
+      "step": 11405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7598185539245605,
+      "learning_rate": 9.577850877192984e-06,
+      "loss": 2.1826,
+      "step": 11406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0121572017669678,
+      "learning_rate": 9.576937134502926e-06,
+      "loss": 1.7067,
+      "step": 11407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.643794059753418,
+      "learning_rate": 9.576023391812866e-06,
+      "loss": 1.4544,
+      "step": 11408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.802023410797119,
+      "learning_rate": 9.575109649122809e-06,
+      "loss": 1.6308,
+      "step": 11409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.871880531311035,
+      "learning_rate": 9.574195906432749e-06,
+      "loss": 1.8239,
+      "step": 11410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5025546550750732,
+      "learning_rate": 9.573282163742691e-06,
+      "loss": 1.4622,
+      "step": 11411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.750943422317505,
+      "learning_rate": 9.572368421052632e-06,
+      "loss": 2.0165,
+      "step": 11412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.198329210281372,
+      "learning_rate": 9.571454678362574e-06,
+      "loss": 1.8002,
+      "step": 11413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8020143508911133,
+      "learning_rate": 9.570540935672516e-06,
+      "loss": 1.3821,
+      "step": 11414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.957961082458496,
+      "learning_rate": 9.569627192982457e-06,
+      "loss": 1.9939,
+      "step": 11415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6487503051757812,
+      "learning_rate": 9.568713450292399e-06,
+      "loss": 1.2139,
+      "step": 11416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6975324153900146,
+      "learning_rate": 9.56779970760234e-06,
+      "loss": 1.2729,
+      "step": 11417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4082372188568115,
+      "learning_rate": 9.566885964912282e-06,
+      "loss": 1.4465,
+      "step": 11418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5785865783691406,
+      "learning_rate": 9.565972222222222e-06,
+      "loss": 1.9308,
+      "step": 11419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1864452362060547,
+      "learning_rate": 9.565058479532164e-06,
+      "loss": 1.4439,
+      "step": 11420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6471972465515137,
+      "learning_rate": 9.564144736842105e-06,
+      "loss": 1.6603,
+      "step": 11421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.654052495956421,
+      "learning_rate": 9.563230994152047e-06,
+      "loss": 1.9847,
+      "step": 11422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.425567626953125,
+      "learning_rate": 9.56231725146199e-06,
+      "loss": 1.2431,
+      "step": 11423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.582951307296753,
+      "learning_rate": 9.56140350877193e-06,
+      "loss": 1.1767,
+      "step": 11424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6393797397613525,
+      "learning_rate": 9.560489766081872e-06,
+      "loss": 1.486,
+      "step": 11425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8891239166259766,
+      "learning_rate": 9.559576023391813e-06,
+      "loss": 1.5647,
+      "step": 11426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.487154483795166,
+      "learning_rate": 9.558662280701755e-06,
+      "loss": 1.1751,
+      "step": 11427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.383928060531616,
+      "learning_rate": 9.557748538011695e-06,
+      "loss": 1.5025,
+      "step": 11428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3498599529266357,
+      "learning_rate": 9.556834795321638e-06,
+      "loss": 1.1112,
+      "step": 11429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5144708156585693,
+      "learning_rate": 9.55592105263158e-06,
+      "loss": 1.2673,
+      "step": 11430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8394346237182617,
+      "learning_rate": 9.55500730994152e-06,
+      "loss": 1.4611,
+      "step": 11431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7755181789398193,
+      "learning_rate": 9.554093567251463e-06,
+      "loss": 1.4589,
+      "step": 11432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0195157527923584,
+      "learning_rate": 9.553179824561403e-06,
+      "loss": 1.8368,
+      "step": 11433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.229330539703369,
+      "learning_rate": 9.552266081871345e-06,
+      "loss": 1.5152,
+      "step": 11434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.501276731491089,
+      "learning_rate": 9.551352339181286e-06,
+      "loss": 1.3224,
+      "step": 11435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5509986877441406,
+      "learning_rate": 9.55043859649123e-06,
+      "loss": 1.9351,
+      "step": 11436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6236629486083984,
+      "learning_rate": 9.54952485380117e-06,
+      "loss": 1.4915,
+      "step": 11437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5899460315704346,
+      "learning_rate": 9.548611111111113e-06,
+      "loss": 1.3933,
+      "step": 11438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.17494535446167,
+      "learning_rate": 9.547697368421053e-06,
+      "loss": 1.8061,
+      "step": 11439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8293380737304688,
+      "learning_rate": 9.546783625730995e-06,
+      "loss": 1.4312,
+      "step": 11440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4738879203796387,
+      "learning_rate": 9.545869883040938e-06,
+      "loss": 1.5653,
+      "step": 11441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7505621910095215,
+      "learning_rate": 9.544956140350878e-06,
+      "loss": 1.6902,
+      "step": 11442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1421055793762207,
+      "learning_rate": 9.54404239766082e-06,
+      "loss": 1.3155,
+      "step": 11443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4118480682373047,
+      "learning_rate": 9.543128654970761e-06,
+      "loss": 1.4969,
+      "step": 11444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2424912452697754,
+      "learning_rate": 9.542214912280703e-06,
+      "loss": 1.3617,
+      "step": 11445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0322375297546387,
+      "learning_rate": 9.541301169590644e-06,
+      "loss": 1.1203,
+      "step": 11446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1145524978637695,
+      "learning_rate": 9.540387426900586e-06,
+      "loss": 1.7746,
+      "step": 11447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3408501148223877,
+      "learning_rate": 9.539473684210528e-06,
+      "loss": 1.5037,
+      "step": 11448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.103818416595459,
+      "learning_rate": 9.538559941520469e-06,
+      "loss": 1.338,
+      "step": 11449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.442863941192627,
+      "learning_rate": 9.53764619883041e-06,
+      "loss": 1.3067,
+      "step": 11450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.442962169647217,
+      "learning_rate": 9.536732456140351e-06,
+      "loss": 1.3289,
+      "step": 11451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.852052927017212,
+      "learning_rate": 9.535818713450294e-06,
+      "loss": 1.7269,
+      "step": 11452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9459309577941895,
+      "learning_rate": 9.534904970760234e-06,
+      "loss": 1.6341,
+      "step": 11453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9041497707366943,
+      "learning_rate": 9.533991228070176e-06,
+      "loss": 1.5427,
+      "step": 11454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9699859619140625,
+      "learning_rate": 9.533077485380119e-06,
+      "loss": 1.1782,
+      "step": 11455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.560986042022705,
+      "learning_rate": 9.532163742690059e-06,
+      "loss": 1.3374,
+      "step": 11456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.640242338180542,
+      "learning_rate": 9.531250000000001e-06,
+      "loss": 1.4644,
+      "step": 11457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4708492755889893,
+      "learning_rate": 9.530336257309942e-06,
+      "loss": 1.3264,
+      "step": 11458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3683345317840576,
+      "learning_rate": 9.529422514619884e-06,
+      "loss": 1.3561,
+      "step": 11459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0099411010742188,
+      "learning_rate": 9.528508771929825e-06,
+      "loss": 1.2473,
+      "step": 11460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5610737800598145,
+      "learning_rate": 9.527595029239767e-06,
+      "loss": 1.4757,
+      "step": 11461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4391977787017822,
+      "learning_rate": 9.526681286549709e-06,
+      "loss": 1.4527,
+      "step": 11462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1932125091552734,
+      "learning_rate": 9.52576754385965e-06,
+      "loss": 1.3608,
+      "step": 11463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.134422779083252,
+      "learning_rate": 9.524853801169592e-06,
+      "loss": 1.2742,
+      "step": 11464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.773090124130249,
+      "learning_rate": 9.523940058479532e-06,
+      "loss": 1.4343,
+      "step": 11465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7719945907592773,
+      "learning_rate": 9.523026315789474e-06,
+      "loss": 1.4926,
+      "step": 11466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9648222923278809,
+      "learning_rate": 9.522112573099415e-06,
+      "loss": 1.354,
+      "step": 11467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5711002349853516,
+      "learning_rate": 9.521198830409357e-06,
+      "loss": 1.4221,
+      "step": 11468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5756988525390625,
+      "learning_rate": 9.5202850877193e-06,
+      "loss": 1.4568,
+      "step": 11469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.010523557662964,
+      "learning_rate": 9.51937134502924e-06,
+      "loss": 1.7321,
+      "step": 11470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1113057136535645,
+      "learning_rate": 9.518457602339182e-06,
+      "loss": 1.436,
+      "step": 11471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.696895122528076,
+      "learning_rate": 9.517543859649123e-06,
+      "loss": 1.5894,
+      "step": 11472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6033666133880615,
+      "learning_rate": 9.516630116959065e-06,
+      "loss": 1.4445,
+      "step": 11473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3699522018432617,
+      "learning_rate": 9.515716374269005e-06,
+      "loss": 1.384,
+      "step": 11474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4728658199310303,
+      "learning_rate": 9.514802631578948e-06,
+      "loss": 1.7965,
+      "step": 11475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7262115478515625,
+      "learning_rate": 9.51388888888889e-06,
+      "loss": 1.483,
+      "step": 11476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8430702686309814,
+      "learning_rate": 9.51297514619883e-06,
+      "loss": 1.4102,
+      "step": 11477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2180774211883545,
+      "learning_rate": 9.512061403508773e-06,
+      "loss": 1.0865,
+      "step": 11478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5458428859710693,
+      "learning_rate": 9.511147660818715e-06,
+      "loss": 1.3264,
+      "step": 11479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.792233467102051,
+      "learning_rate": 9.510233918128655e-06,
+      "loss": 1.6197,
+      "step": 11480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.678013324737549,
+      "learning_rate": 9.509320175438598e-06,
+      "loss": 1.4132,
+      "step": 11481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.566412925720215,
+      "learning_rate": 9.50840643274854e-06,
+      "loss": 1.4632,
+      "step": 11482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.002624034881592,
+      "learning_rate": 9.50749269005848e-06,
+      "loss": 1.548,
+      "step": 11483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8011186122894287,
+      "learning_rate": 9.506578947368423e-06,
+      "loss": 1.5181,
+      "step": 11484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.898599624633789,
+      "learning_rate": 9.505665204678363e-06,
+      "loss": 1.7127,
+      "step": 11485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7444379329681396,
+      "learning_rate": 9.504751461988305e-06,
+      "loss": 1.5248,
+      "step": 11486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8730106353759766,
+      "learning_rate": 9.503837719298246e-06,
+      "loss": 1.4739,
+      "step": 11487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6416237354278564,
+      "learning_rate": 9.502923976608188e-06,
+      "loss": 1.4356,
+      "step": 11488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4958386421203613,
+      "learning_rate": 9.50201023391813e-06,
+      "loss": 1.3418,
+      "step": 11489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4506070613861084,
+      "learning_rate": 9.501096491228071e-06,
+      "loss": 1.5025,
+      "step": 11490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5396242141723633,
+      "learning_rate": 9.500182748538013e-06,
+      "loss": 1.4979,
+      "step": 11491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9331068992614746,
+      "learning_rate": 9.499269005847954e-06,
+      "loss": 1.699,
+      "step": 11492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5080230236053467,
+      "learning_rate": 9.498355263157896e-06,
+      "loss": 1.5126,
+      "step": 11493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8837192058563232,
+      "learning_rate": 9.497441520467836e-06,
+      "loss": 1.5336,
+      "step": 11494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6339058876037598,
+      "learning_rate": 9.496527777777779e-06,
+      "loss": 1.4126,
+      "step": 11495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1513657569885254,
+      "learning_rate": 9.49561403508772e-06,
+      "loss": 1.4019,
+      "step": 11496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.750800132751465,
+      "learning_rate": 9.494700292397661e-06,
+      "loss": 2.0087,
+      "step": 11497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.452503204345703,
+      "learning_rate": 9.493786549707604e-06,
+      "loss": 1.4448,
+      "step": 11498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2228991985321045,
+      "learning_rate": 9.492872807017544e-06,
+      "loss": 1.4089,
+      "step": 11499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.127479553222656,
+      "learning_rate": 9.491959064327486e-06,
+      "loss": 1.3331,
+      "step": 11500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8800463676452637,
+      "learning_rate": 9.491045321637427e-06,
+      "loss": 1.5674,
+      "step": 11501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6352427005767822,
+      "learning_rate": 9.490131578947369e-06,
+      "loss": 1.4458,
+      "step": 11502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4597184658050537,
+      "learning_rate": 9.489217836257311e-06,
+      "loss": 1.6347,
+      "step": 11503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.706900119781494,
+      "learning_rate": 9.488304093567252e-06,
+      "loss": 1.6973,
+      "step": 11504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.138138294219971,
+      "learning_rate": 9.487390350877194e-06,
+      "loss": 1.7586,
+      "step": 11505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.441453218460083,
+      "learning_rate": 9.486476608187135e-06,
+      "loss": 1.5834,
+      "step": 11506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.509979486465454,
+      "learning_rate": 9.485562865497077e-06,
+      "loss": 1.5511,
+      "step": 11507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1908037662506104,
+      "learning_rate": 9.484649122807017e-06,
+      "loss": 1.4356,
+      "step": 11508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5853822231292725,
+      "learning_rate": 9.48373538011696e-06,
+      "loss": 1.6734,
+      "step": 11509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2709102630615234,
+      "learning_rate": 9.482821637426902e-06,
+      "loss": 1.2763,
+      "step": 11510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.339428424835205,
+      "learning_rate": 9.481907894736842e-06,
+      "loss": 1.1936,
+      "step": 11511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.404978036880493,
+      "learning_rate": 9.480994152046784e-06,
+      "loss": 1.4444,
+      "step": 11512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.536374807357788,
+      "learning_rate": 9.480080409356725e-06,
+      "loss": 1.6247,
+      "step": 11513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.643986463546753,
+      "learning_rate": 9.479166666666667e-06,
+      "loss": 1.7737,
+      "step": 11514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.570136785507202,
+      "learning_rate": 9.478252923976608e-06,
+      "loss": 1.497,
+      "step": 11515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3792972564697266,
+      "learning_rate": 9.47733918128655e-06,
+      "loss": 1.5192,
+      "step": 11516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5110459327697754,
+      "learning_rate": 9.476425438596492e-06,
+      "loss": 1.3908,
+      "step": 11517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3582277297973633,
+      "learning_rate": 9.475511695906433e-06,
+      "loss": 1.2787,
+      "step": 11518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6052544116973877,
+      "learning_rate": 9.474597953216375e-06,
+      "loss": 1.6842,
+      "step": 11519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9595534801483154,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 1.1347,
+      "step": 11520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.891531229019165,
+      "learning_rate": 9.472770467836258e-06,
+      "loss": 1.6384,
+      "step": 11521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0005455017089844,
+      "learning_rate": 9.4718567251462e-06,
+      "loss": 1.8467,
+      "step": 11522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6507580280303955,
+      "learning_rate": 9.470942982456142e-06,
+      "loss": 1.7363,
+      "step": 11523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.501232624053955,
+      "learning_rate": 9.470029239766083e-06,
+      "loss": 1.5644,
+      "step": 11524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6349785327911377,
+      "learning_rate": 9.469115497076025e-06,
+      "loss": 1.316,
+      "step": 11525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2376291751861572,
+      "learning_rate": 9.468201754385965e-06,
+      "loss": 1.9122,
+      "step": 11526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.730424642562866,
+      "learning_rate": 9.467288011695908e-06,
+      "loss": 1.7791,
+      "step": 11527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4556970596313477,
+      "learning_rate": 9.46637426900585e-06,
+      "loss": 1.4014,
+      "step": 11528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5575196743011475,
+      "learning_rate": 9.46546052631579e-06,
+      "loss": 1.3716,
+      "step": 11529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6524977684020996,
+      "learning_rate": 9.464546783625733e-06,
+      "loss": 1.3352,
+      "step": 11530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.977388858795166,
+      "learning_rate": 9.463633040935673e-06,
+      "loss": 1.8058,
+      "step": 11531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6138908863067627,
+      "learning_rate": 9.462719298245615e-06,
+      "loss": 1.5997,
+      "step": 11532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.277451515197754,
+      "learning_rate": 9.461805555555556e-06,
+      "loss": 1.8901,
+      "step": 11533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.379181146621704,
+      "learning_rate": 9.460891812865498e-06,
+      "loss": 1.5369,
+      "step": 11534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.361208200454712,
+      "learning_rate": 9.45997807017544e-06,
+      "loss": 1.3362,
+      "step": 11535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.082859516143799,
+      "learning_rate": 9.459064327485381e-06,
+      "loss": 1.5896,
+      "step": 11536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1576595306396484,
+      "learning_rate": 9.458150584795323e-06,
+      "loss": 1.2151,
+      "step": 11537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6118292808532715,
+      "learning_rate": 9.457236842105264e-06,
+      "loss": 1.4662,
+      "step": 11538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9400365352630615,
+      "learning_rate": 9.456323099415206e-06,
+      "loss": 1.6198,
+      "step": 11539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.793041229248047,
+      "learning_rate": 9.455409356725146e-06,
+      "loss": 1.4754,
+      "step": 11540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6752634048461914,
+      "learning_rate": 9.454495614035089e-06,
+      "loss": 1.8762,
+      "step": 11541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3352530002593994,
+      "learning_rate": 9.45358187134503e-06,
+      "loss": 1.777,
+      "step": 11542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9620375633239746,
+      "learning_rate": 9.452668128654971e-06,
+      "loss": 1.6178,
+      "step": 11543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7894644737243652,
+      "learning_rate": 9.451754385964914e-06,
+      "loss": 1.5811,
+      "step": 11544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.930227518081665,
+      "learning_rate": 9.450840643274854e-06,
+      "loss": 1.6244,
+      "step": 11545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6854686737060547,
+      "learning_rate": 9.449926900584796e-06,
+      "loss": 1.6304,
+      "step": 11546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9866726398468018,
+      "learning_rate": 9.449013157894737e-06,
+      "loss": 1.5193,
+      "step": 11547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0401296615600586,
+      "learning_rate": 9.448099415204679e-06,
+      "loss": 1.4741,
+      "step": 11548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5952086448669434,
+      "learning_rate": 9.447185672514621e-06,
+      "loss": 1.3209,
+      "step": 11549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9330601692199707,
+      "learning_rate": 9.446271929824562e-06,
+      "loss": 0.9451,
+      "step": 11550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1197988986968994,
+      "learning_rate": 9.445358187134504e-06,
+      "loss": 1.7455,
+      "step": 11551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.557201623916626,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.3554,
+      "step": 11552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9599685668945312,
+      "learning_rate": 9.443530701754387e-06,
+      "loss": 1.597,
+      "step": 11553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.310471534729004,
+      "learning_rate": 9.442616959064327e-06,
+      "loss": 1.2137,
+      "step": 11554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4082424640655518,
+      "learning_rate": 9.44170321637427e-06,
+      "loss": 1.3848,
+      "step": 11555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4787144660949707,
+      "learning_rate": 9.440789473684212e-06,
+      "loss": 1.6426,
+      "step": 11556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.138444423675537,
+      "learning_rate": 9.439875730994152e-06,
+      "loss": 1.7754,
+      "step": 11557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6710317134857178,
+      "learning_rate": 9.438961988304094e-06,
+      "loss": 1.6374,
+      "step": 11558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3724961280822754,
+      "learning_rate": 9.438048245614035e-06,
+      "loss": 1.9812,
+      "step": 11559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.208301305770874,
+      "learning_rate": 9.437134502923977e-06,
+      "loss": 1.845,
+      "step": 11560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5191702842712402,
+      "learning_rate": 9.436220760233918e-06,
+      "loss": 1.6323,
+      "step": 11561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4609169960021973,
+      "learning_rate": 9.43530701754386e-06,
+      "loss": 1.6154,
+      "step": 11562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6168124675750732,
+      "learning_rate": 9.434393274853802e-06,
+      "loss": 1.5709,
+      "step": 11563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6205263137817383,
+      "learning_rate": 9.433479532163744e-06,
+      "loss": 1.6928,
+      "step": 11564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.749253988265991,
+      "learning_rate": 9.432565789473685e-06,
+      "loss": 1.5038,
+      "step": 11565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8546335697174072,
+      "learning_rate": 9.431652046783627e-06,
+      "loss": 1.7751,
+      "step": 11566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.9855451583862305,
+      "learning_rate": 9.430738304093568e-06,
+      "loss": 1.5522,
+      "step": 11567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7044615745544434,
+      "learning_rate": 9.42982456140351e-06,
+      "loss": 1.5264,
+      "step": 11568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.624769926071167,
+      "learning_rate": 9.428910818713452e-06,
+      "loss": 1.4678,
+      "step": 11569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2076704502105713,
+      "learning_rate": 9.427997076023393e-06,
+      "loss": 1.3194,
+      "step": 11570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6229021549224854,
+      "learning_rate": 9.427083333333335e-06,
+      "loss": 1.5768,
+      "step": 11571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.182713031768799,
+      "learning_rate": 9.426169590643275e-06,
+      "loss": 1.4076,
+      "step": 11572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2779393196105957,
+      "learning_rate": 9.425255847953218e-06,
+      "loss": 1.4186,
+      "step": 11573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.428171396255493,
+      "learning_rate": 9.424342105263158e-06,
+      "loss": 1.4334,
+      "step": 11574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.382089853286743,
+      "learning_rate": 9.4234283625731e-06,
+      "loss": 1.4446,
+      "step": 11575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.416550636291504,
+      "learning_rate": 9.422514619883043e-06,
+      "loss": 1.4166,
+      "step": 11576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9625203609466553,
+      "learning_rate": 9.421600877192983e-06,
+      "loss": 1.7661,
+      "step": 11577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.349076271057129,
+      "learning_rate": 9.420687134502925e-06,
+      "loss": 1.1886,
+      "step": 11578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7667927742004395,
+      "learning_rate": 9.419773391812866e-06,
+      "loss": 1.4173,
+      "step": 11579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.651991128921509,
+      "learning_rate": 9.418859649122808e-06,
+      "loss": 1.5406,
+      "step": 11580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5055253505706787,
+      "learning_rate": 9.417945906432749e-06,
+      "loss": 1.4051,
+      "step": 11581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1624035835266113,
+      "learning_rate": 9.417032163742691e-06,
+      "loss": 1.1149,
+      "step": 11582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.766085386276245,
+      "learning_rate": 9.416118421052633e-06,
+      "loss": 1.5105,
+      "step": 11583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.695531129837036,
+      "learning_rate": 9.415204678362574e-06,
+      "loss": 1.7519,
+      "step": 11584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5280895233154297,
+      "learning_rate": 9.414290935672516e-06,
+      "loss": 1.3653,
+      "step": 11585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8945271968841553,
+      "learning_rate": 9.413377192982456e-06,
+      "loss": 1.5604,
+      "step": 11586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.172914743423462,
+      "learning_rate": 9.412463450292399e-06,
+      "loss": 1.0789,
+      "step": 11587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.189293146133423,
+      "learning_rate": 9.411549707602339e-06,
+      "loss": 1.5226,
+      "step": 11588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6099865436553955,
+      "learning_rate": 9.410635964912281e-06,
+      "loss": 1.5621,
+      "step": 11589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0841009616851807,
+      "learning_rate": 9.409722222222224e-06,
+      "loss": 1.7068,
+      "step": 11590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.589848518371582,
+      "learning_rate": 9.408808479532164e-06,
+      "loss": 1.5049,
+      "step": 11591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9984827041625977,
+      "learning_rate": 9.407894736842106e-06,
+      "loss": 1.6512,
+      "step": 11592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5457661151885986,
+      "learning_rate": 9.406980994152047e-06,
+      "loss": 1.4819,
+      "step": 11593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.778000593185425,
+      "learning_rate": 9.406067251461989e-06,
+      "loss": 1.5891,
+      "step": 11594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1405117511749268,
+      "learning_rate": 9.40515350877193e-06,
+      "loss": 1.7184,
+      "step": 11595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.865985155105591,
+      "learning_rate": 9.404239766081872e-06,
+      "loss": 1.7734,
+      "step": 11596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.477785587310791,
+      "learning_rate": 9.403326023391814e-06,
+      "loss": 1.5499,
+      "step": 11597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4733853340148926,
+      "learning_rate": 9.402412280701755e-06,
+      "loss": 1.5089,
+      "step": 11598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.518723964691162,
+      "learning_rate": 9.401498538011697e-06,
+      "loss": 1.2585,
+      "step": 11599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5476465225219727,
+      "learning_rate": 9.400584795321637e-06,
+      "loss": 1.532,
+      "step": 11600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.720019817352295,
+      "learning_rate": 9.39967105263158e-06,
+      "loss": 1.7364,
+      "step": 11601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4903571605682373,
+      "learning_rate": 9.39875730994152e-06,
+      "loss": 1.4399,
+      "step": 11602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.411219358444214,
+      "learning_rate": 9.397843567251462e-06,
+      "loss": 1.466,
+      "step": 11603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.390822172164917,
+      "learning_rate": 9.396929824561404e-06,
+      "loss": 1.6587,
+      "step": 11604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9411087036132812,
+      "learning_rate": 9.396016081871345e-06,
+      "loss": 1.454,
+      "step": 11605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.506544589996338,
+      "learning_rate": 9.395102339181287e-06,
+      "loss": 1.4756,
+      "step": 11606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7654216289520264,
+      "learning_rate": 9.39418859649123e-06,
+      "loss": 1.8275,
+      "step": 11607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.739598512649536,
+      "learning_rate": 9.393274853801172e-06,
+      "loss": 1.6413,
+      "step": 11608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4323277473449707,
+      "learning_rate": 9.392361111111112e-06,
+      "loss": 1.5291,
+      "step": 11609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6992533206939697,
+      "learning_rate": 9.391447368421054e-06,
+      "loss": 1.3235,
+      "step": 11610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2395708560943604,
+      "learning_rate": 9.390533625730995e-06,
+      "loss": 1.321,
+      "step": 11611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.804616928100586,
+      "learning_rate": 9.389619883040937e-06,
+      "loss": 1.6741,
+      "step": 11612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.517254590988159,
+      "learning_rate": 9.388706140350878e-06,
+      "loss": 1.5624,
+      "step": 11613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.019562244415283,
+      "learning_rate": 9.38779239766082e-06,
+      "loss": 1.8136,
+      "step": 11614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.448155164718628,
+      "learning_rate": 9.386878654970762e-06,
+      "loss": 1.4547,
+      "step": 11615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4156277179718018,
+      "learning_rate": 9.385964912280703e-06,
+      "loss": 1.5418,
+      "step": 11616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2459449768066406,
+      "learning_rate": 9.385051169590645e-06,
+      "loss": 1.1543,
+      "step": 11617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6213696002960205,
+      "learning_rate": 9.384137426900585e-06,
+      "loss": 1.3867,
+      "step": 11618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3991527557373047,
+      "learning_rate": 9.383223684210528e-06,
+      "loss": 1.3827,
+      "step": 11619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6592371463775635,
+      "learning_rate": 9.382309941520468e-06,
+      "loss": 1.5378,
+      "step": 11620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.801987886428833,
+      "learning_rate": 9.38139619883041e-06,
+      "loss": 1.4485,
+      "step": 11621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7404961585998535,
+      "learning_rate": 9.380482456140353e-06,
+      "loss": 1.4534,
+      "step": 11622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.370941638946533,
+      "learning_rate": 9.379568713450293e-06,
+      "loss": 1.5146,
+      "step": 11623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9437222480773926,
+      "learning_rate": 9.378654970760235e-06,
+      "loss": 1.3232,
+      "step": 11624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4826059341430664,
+      "learning_rate": 9.377741228070176e-06,
+      "loss": 1.4394,
+      "step": 11625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.216716766357422,
+      "learning_rate": 9.376827485380118e-06,
+      "loss": 1.2696,
+      "step": 11626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3878002166748047,
+      "learning_rate": 9.375913742690059e-06,
+      "loss": 1.4326,
+      "step": 11627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5278375148773193,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.4644,
+      "step": 11628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.760928153991699,
+      "learning_rate": 9.374086257309943e-06,
+      "loss": 1.4401,
+      "step": 11629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.59060001373291,
+      "learning_rate": 9.373172514619884e-06,
+      "loss": 1.5567,
+      "step": 11630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5512936115264893,
+      "learning_rate": 9.372258771929826e-06,
+      "loss": 1.5344,
+      "step": 11631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9666106700897217,
+      "learning_rate": 9.371345029239766e-06,
+      "loss": 1.6372,
+      "step": 11632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1248412132263184,
+      "learning_rate": 9.370431286549709e-06,
+      "loss": 1.5377,
+      "step": 11633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.577254056930542,
+      "learning_rate": 9.369517543859649e-06,
+      "loss": 1.5583,
+      "step": 11634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5354421138763428,
+      "learning_rate": 9.368603801169591e-06,
+      "loss": 1.5073,
+      "step": 11635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1406471729278564,
+      "learning_rate": 9.367690058479534e-06,
+      "loss": 1.7355,
+      "step": 11636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.225742816925049,
+      "learning_rate": 9.366776315789474e-06,
+      "loss": 1.154,
+      "step": 11637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7592203617095947,
+      "learning_rate": 9.365862573099416e-06,
+      "loss": 1.2113,
+      "step": 11638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.798495292663574,
+      "learning_rate": 9.364948830409357e-06,
+      "loss": 1.773,
+      "step": 11639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8220622539520264,
+      "learning_rate": 9.364035087719299e-06,
+      "loss": 1.4769,
+      "step": 11640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6605756282806396,
+      "learning_rate": 9.36312134502924e-06,
+      "loss": 1.2323,
+      "step": 11641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3143231868743896,
+      "learning_rate": 9.362207602339182e-06,
+      "loss": 1.1736,
+      "step": 11642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.492964506149292,
+      "learning_rate": 9.361293859649124e-06,
+      "loss": 1.6175,
+      "step": 11643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1850101947784424,
+      "learning_rate": 9.360380116959065e-06,
+      "loss": 1.1985,
+      "step": 11644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9253735542297363,
+      "learning_rate": 9.359466374269007e-06,
+      "loss": 1.5113,
+      "step": 11645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.151972770690918,
+      "learning_rate": 9.358552631578947e-06,
+      "loss": 1.168,
+      "step": 11646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.621558904647827,
+      "learning_rate": 9.35763888888889e-06,
+      "loss": 1.6412,
+      "step": 11647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2670719623565674,
+      "learning_rate": 9.35672514619883e-06,
+      "loss": 1.2941,
+      "step": 11648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6208605766296387,
+      "learning_rate": 9.355811403508772e-06,
+      "loss": 1.4035,
+      "step": 11649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.806313991546631,
+      "learning_rate": 9.354897660818714e-06,
+      "loss": 1.4505,
+      "step": 11650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2609689235687256,
+      "learning_rate": 9.353983918128657e-06,
+      "loss": 1.2195,
+      "step": 11651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3820836544036865,
+      "learning_rate": 9.353070175438597e-06,
+      "loss": 1.3689,
+      "step": 11652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.612901449203491,
+      "learning_rate": 9.35215643274854e-06,
+      "loss": 1.539,
+      "step": 11653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.678649425506592,
+      "learning_rate": 9.35124269005848e-06,
+      "loss": 1.2624,
+      "step": 11654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5595552921295166,
+      "learning_rate": 9.350328947368422e-06,
+      "loss": 1.5205,
+      "step": 11655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.418203830718994,
+      "learning_rate": 9.349415204678364e-06,
+      "loss": 1.3223,
+      "step": 11656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5992019176483154,
+      "learning_rate": 9.348501461988305e-06,
+      "loss": 1.6144,
+      "step": 11657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4303159713745117,
+      "learning_rate": 9.347587719298247e-06,
+      "loss": 1.3898,
+      "step": 11658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6353352069854736,
+      "learning_rate": 9.346673976608188e-06,
+      "loss": 1.7011,
+      "step": 11659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7429981231689453,
+      "learning_rate": 9.34576023391813e-06,
+      "loss": 1.3382,
+      "step": 11660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.656670570373535,
+      "learning_rate": 9.34484649122807e-06,
+      "loss": 1.5717,
+      "step": 11661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.183286666870117,
+      "learning_rate": 9.343932748538013e-06,
+      "loss": 1.5361,
+      "step": 11662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9534294605255127,
+      "learning_rate": 9.343019005847955e-06,
+      "loss": 1.4787,
+      "step": 11663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.453273296356201,
+      "learning_rate": 9.342105263157895e-06,
+      "loss": 1.3901,
+      "step": 11664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2025046348571777,
+      "learning_rate": 9.341191520467838e-06,
+      "loss": 1.2856,
+      "step": 11665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5771830081939697,
+      "learning_rate": 9.340277777777778e-06,
+      "loss": 1.4497,
+      "step": 11666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0105881690979004,
+      "learning_rate": 9.33936403508772e-06,
+      "loss": 1.5277,
+      "step": 11667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5125269889831543,
+      "learning_rate": 9.338450292397661e-06,
+      "loss": 1.4363,
+      "step": 11668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.228107452392578,
+      "learning_rate": 9.337536549707603e-06,
+      "loss": 1.3678,
+      "step": 11669
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7178759574890137,
+      "learning_rate": 9.336622807017545e-06,
+      "loss": 1.9781,
+      "step": 11670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.612980604171753,
+      "learning_rate": 9.335709064327486e-06,
+      "loss": 1.5388,
+      "step": 11671
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5883290767669678,
+      "learning_rate": 9.334795321637428e-06,
+      "loss": 1.2603,
+      "step": 11672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4904730319976807,
+      "learning_rate": 9.333881578947369e-06,
+      "loss": 1.3133,
+      "step": 11673
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3266050815582275,
+      "learning_rate": 9.332967836257311e-06,
+      "loss": 1.4775,
+      "step": 11674
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.255162239074707,
+      "learning_rate": 9.332054093567251e-06,
+      "loss": 1.3696,
+      "step": 11675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.844595193862915,
+      "learning_rate": 9.331140350877194e-06,
+      "loss": 1.5877,
+      "step": 11676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8809924125671387,
+      "learning_rate": 9.330226608187136e-06,
+      "loss": 1.4523,
+      "step": 11677
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.701469659805298,
+      "learning_rate": 9.329312865497076e-06,
+      "loss": 1.5821,
+      "step": 11678
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0508925914764404,
+      "learning_rate": 9.328399122807019e-06,
+      "loss": 1.614,
+      "step": 11679
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5423669815063477,
+      "learning_rate": 9.327485380116959e-06,
+      "loss": 1.4316,
+      "step": 11680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.928332567214966,
+      "learning_rate": 9.326571637426901e-06,
+      "loss": 1.6798,
+      "step": 11681
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.244922637939453,
+      "learning_rate": 9.325657894736842e-06,
+      "loss": 1.3144,
+      "step": 11682
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0699942111968994,
+      "learning_rate": 9.324744152046784e-06,
+      "loss": 1.74,
+      "step": 11683
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.418015241622925,
+      "learning_rate": 9.323830409356726e-06,
+      "loss": 1.3845,
+      "step": 11684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4779984951019287,
+      "learning_rate": 9.322916666666667e-06,
+      "loss": 1.8886,
+      "step": 11685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.123723030090332,
+      "learning_rate": 9.322002923976609e-06,
+      "loss": 1.3152,
+      "step": 11686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.309856653213501,
+      "learning_rate": 9.32108918128655e-06,
+      "loss": 1.0792,
+      "step": 11687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2715625762939453,
+      "learning_rate": 9.320175438596492e-06,
+      "loss": 1.44,
+      "step": 11688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.422868251800537,
+      "learning_rate": 9.319261695906432e-06,
+      "loss": 1.2744,
+      "step": 11689
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0998287200927734,
+      "learning_rate": 9.318347953216375e-06,
+      "loss": 1.1125,
+      "step": 11690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9176971912384033,
+      "learning_rate": 9.317434210526317e-06,
+      "loss": 1.7041,
+      "step": 11691
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.630908489227295,
+      "learning_rate": 9.316520467836257e-06,
+      "loss": 1.6638,
+      "step": 11692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4804911613464355,
+      "learning_rate": 9.3156067251462e-06,
+      "loss": 1.5086,
+      "step": 11693
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8879988193511963,
+      "learning_rate": 9.314692982456142e-06,
+      "loss": 1.6913,
+      "step": 11694
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.755826473236084,
+      "learning_rate": 9.313779239766082e-06,
+      "loss": 1.6806,
+      "step": 11695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.08054780960083,
+      "learning_rate": 9.312865497076024e-06,
+      "loss": 1.6408,
+      "step": 11696
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3495378494262695,
+      "learning_rate": 9.311951754385967e-06,
+      "loss": 1.3248,
+      "step": 11697
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3283679485321045,
+      "learning_rate": 9.311038011695907e-06,
+      "loss": 1.5983,
+      "step": 11698
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8248291015625,
+      "learning_rate": 9.31012426900585e-06,
+      "loss": 1.6636,
+      "step": 11699
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3628249168395996,
+      "learning_rate": 9.30921052631579e-06,
+      "loss": 1.2575,
+      "step": 11700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.603489398956299,
+      "learning_rate": 9.308296783625732e-06,
+      "loss": 1.7157,
+      "step": 11701
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.545719861984253,
+      "learning_rate": 9.307383040935673e-06,
+      "loss": 1.5593,
+      "step": 11702
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5602846145629883,
+      "learning_rate": 9.306469298245615e-06,
+      "loss": 1.3733,
+      "step": 11703
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4006786346435547,
+      "learning_rate": 9.305555555555557e-06,
+      "loss": 1.4149,
+      "step": 11704
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1385457515716553,
+      "learning_rate": 9.304641812865498e-06,
+      "loss": 1.5088,
+      "step": 11705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7847800254821777,
+      "learning_rate": 9.30372807017544e-06,
+      "loss": 1.6542,
+      "step": 11706
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.675170660018921,
+      "learning_rate": 9.30281432748538e-06,
+      "loss": 1.5355,
+      "step": 11707
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7258081436157227,
+      "learning_rate": 9.301900584795323e-06,
+      "loss": 1.4827,
+      "step": 11708
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.558631658554077,
+      "learning_rate": 9.300986842105263e-06,
+      "loss": 1.5587,
+      "step": 11709
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.731337070465088,
+      "learning_rate": 9.300073099415205e-06,
+      "loss": 1.6688,
+      "step": 11710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.499558925628662,
+      "learning_rate": 9.299159356725148e-06,
+      "loss": 1.2847,
+      "step": 11711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.207556962966919,
+      "learning_rate": 9.298245614035088e-06,
+      "loss": 1.0137,
+      "step": 11712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6061325073242188,
+      "learning_rate": 9.29733187134503e-06,
+      "loss": 1.9045,
+      "step": 11713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2977709770202637,
+      "learning_rate": 9.296418128654971e-06,
+      "loss": 1.4168,
+      "step": 11714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4780476093292236,
+      "learning_rate": 9.295504385964913e-06,
+      "loss": 1.595,
+      "step": 11715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.589730978012085,
+      "learning_rate": 9.294590643274854e-06,
+      "loss": 1.6722,
+      "step": 11716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0731399059295654,
+      "learning_rate": 9.293676900584796e-06,
+      "loss": 1.6429,
+      "step": 11717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.593205690383911,
+      "learning_rate": 9.292763157894738e-06,
+      "loss": 1.9711,
+      "step": 11718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.18733811378479,
+      "learning_rate": 9.291849415204679e-06,
+      "loss": 1.3025,
+      "step": 11719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3787965774536133,
+      "learning_rate": 9.290935672514621e-06,
+      "loss": 1.3554,
+      "step": 11720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7858097553253174,
+      "learning_rate": 9.290021929824561e-06,
+      "loss": 1.5315,
+      "step": 11721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1770222187042236,
+      "learning_rate": 9.289108187134504e-06,
+      "loss": 1.117,
+      "step": 11722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5887839794158936,
+      "learning_rate": 9.288194444444444e-06,
+      "loss": 1.6533,
+      "step": 11723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.952122926712036,
+      "learning_rate": 9.287280701754386e-06,
+      "loss": 1.8485,
+      "step": 11724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.35292911529541,
+      "learning_rate": 9.286366959064329e-06,
+      "loss": 1.5337,
+      "step": 11725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2884738445281982,
+      "learning_rate": 9.285453216374269e-06,
+      "loss": 1.1656,
+      "step": 11726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6312224864959717,
+      "learning_rate": 9.284539473684211e-06,
+      "loss": 1.4732,
+      "step": 11727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2718992233276367,
+      "learning_rate": 9.283625730994152e-06,
+      "loss": 1.439,
+      "step": 11728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4790897369384766,
+      "learning_rate": 9.282711988304094e-06,
+      "loss": 1.654,
+      "step": 11729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.396760940551758,
+      "learning_rate": 9.281798245614035e-06,
+      "loss": 1.1754,
+      "step": 11730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.41694712638855,
+      "learning_rate": 9.280884502923977e-06,
+      "loss": 1.3291,
+      "step": 11731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9715495109558105,
+      "learning_rate": 9.279970760233919e-06,
+      "loss": 1.7864,
+      "step": 11732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6347062587738037,
+      "learning_rate": 9.27905701754386e-06,
+      "loss": 1.5361,
+      "step": 11733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.16377329826355,
+      "learning_rate": 9.278143274853802e-06,
+      "loss": 1.1387,
+      "step": 11734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.050260066986084,
+      "learning_rate": 9.277229532163744e-06,
+      "loss": 1.1828,
+      "step": 11735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2054295539855957,
+      "learning_rate": 9.276315789473686e-06,
+      "loss": 1.6307,
+      "step": 11736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.783498525619507,
+      "learning_rate": 9.275402046783627e-06,
+      "loss": 1.4595,
+      "step": 11737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8796751499176025,
+      "learning_rate": 9.274488304093569e-06,
+      "loss": 1.7442,
+      "step": 11738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.722369432449341,
+      "learning_rate": 9.27357456140351e-06,
+      "loss": 1.7147,
+      "step": 11739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5966734886169434,
+      "learning_rate": 9.272660818713452e-06,
+      "loss": 1.2747,
+      "step": 11740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0508182048797607,
+      "learning_rate": 9.271747076023392e-06,
+      "loss": 1.7467,
+      "step": 11741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.691288471221924,
+      "learning_rate": 9.270833333333334e-06,
+      "loss": 1.6527,
+      "step": 11742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6510331630706787,
+      "learning_rate": 9.269919590643277e-06,
+      "loss": 1.4102,
+      "step": 11743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.516247272491455,
+      "learning_rate": 9.269005847953217e-06,
+      "loss": 1.4087,
+      "step": 11744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1456949710845947,
+      "learning_rate": 9.26809210526316e-06,
+      "loss": 1.6244,
+      "step": 11745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5511417388916016,
+      "learning_rate": 9.2671783625731e-06,
+      "loss": 1.447,
+      "step": 11746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5343875885009766,
+      "learning_rate": 9.266264619883042e-06,
+      "loss": 1.2365,
+      "step": 11747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3895270824432373,
+      "learning_rate": 9.265350877192983e-06,
+      "loss": 1.6255,
+      "step": 11748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.78511905670166,
+      "learning_rate": 9.264437134502925e-06,
+      "loss": 1.6199,
+      "step": 11749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.19179105758667,
+      "learning_rate": 9.263523391812867e-06,
+      "loss": 1.251,
+      "step": 11750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.461426258087158,
+      "learning_rate": 9.262609649122808e-06,
+      "loss": 1.3911,
+      "step": 11751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6172070503234863,
+      "learning_rate": 9.26169590643275e-06,
+      "loss": 1.5808,
+      "step": 11752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2877357006073,
+      "learning_rate": 9.26078216374269e-06,
+      "loss": 1.2787,
+      "step": 11753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.26753306388855,
+      "learning_rate": 9.259868421052633e-06,
+      "loss": 1.6995,
+      "step": 11754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.888519763946533,
+      "learning_rate": 9.258954678362573e-06,
+      "loss": 1.5565,
+      "step": 11755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.373067617416382,
+      "learning_rate": 9.258040935672515e-06,
+      "loss": 1.1528,
+      "step": 11756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.822049140930176,
+      "learning_rate": 9.257127192982458e-06,
+      "loss": 1.7953,
+      "step": 11757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8797354698181152,
+      "learning_rate": 9.256213450292398e-06,
+      "loss": 1.9895,
+      "step": 11758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.620560646057129,
+      "learning_rate": 9.25529970760234e-06,
+      "loss": 1.5902,
+      "step": 11759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.280526638031006,
+      "learning_rate": 9.254385964912281e-06,
+      "loss": 1.4806,
+      "step": 11760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7595930099487305,
+      "learning_rate": 9.253472222222223e-06,
+      "loss": 1.5447,
+      "step": 11761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4390313625335693,
+      "learning_rate": 9.252558479532164e-06,
+      "loss": 1.5571,
+      "step": 11762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7337067127227783,
+      "learning_rate": 9.251644736842106e-06,
+      "loss": 1.872,
+      "step": 11763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8228600025177,
+      "learning_rate": 9.250730994152048e-06,
+      "loss": 1.6282,
+      "step": 11764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.613245725631714,
+      "learning_rate": 9.249817251461989e-06,
+      "loss": 1.3164,
+      "step": 11765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8340413570404053,
+      "learning_rate": 9.248903508771931e-06,
+      "loss": 1.8423,
+      "step": 11766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.862196922302246,
+      "learning_rate": 9.247989766081871e-06,
+      "loss": 2.0861,
+      "step": 11767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6629550457000732,
+      "learning_rate": 9.247076023391814e-06,
+      "loss": 1.6203,
+      "step": 11768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.45347261428833,
+      "learning_rate": 9.246162280701754e-06,
+      "loss": 1.3207,
+      "step": 11769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.92004132270813,
+      "learning_rate": 9.245248538011696e-06,
+      "loss": 1.7835,
+      "step": 11770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.906874179840088,
+      "learning_rate": 9.244334795321639e-06,
+      "loss": 1.8652,
+      "step": 11771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.64937686920166,
+      "learning_rate": 9.243421052631579e-06,
+      "loss": 1.4584,
+      "step": 11772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.241950273513794,
+      "learning_rate": 9.242507309941521e-06,
+      "loss": 1.4934,
+      "step": 11773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9981166124343872,
+      "learning_rate": 9.241593567251462e-06,
+      "loss": 1.2267,
+      "step": 11774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.580260992050171,
+      "learning_rate": 9.240679824561404e-06,
+      "loss": 1.1999,
+      "step": 11775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.902480363845825,
+      "learning_rate": 9.239766081871345e-06,
+      "loss": 1.4103,
+      "step": 11776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5028088092803955,
+      "learning_rate": 9.238852339181287e-06,
+      "loss": 1.5946,
+      "step": 11777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3975069522857666,
+      "learning_rate": 9.237938596491229e-06,
+      "loss": 1.2729,
+      "step": 11778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4082584381103516,
+      "learning_rate": 9.237024853801171e-06,
+      "loss": 1.495,
+      "step": 11779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4452600479125977,
+      "learning_rate": 9.236111111111112e-06,
+      "loss": 1.2773,
+      "step": 11780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7365803718566895,
+      "learning_rate": 9.235197368421054e-06,
+      "loss": 1.5195,
+      "step": 11781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.68229079246521,
+      "learning_rate": 9.234283625730995e-06,
+      "loss": 1.5889,
+      "step": 11782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.927450180053711,
+      "learning_rate": 9.233369883040937e-06,
+      "loss": 1.5305,
+      "step": 11783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.897033214569092,
+      "learning_rate": 9.232456140350879e-06,
+      "loss": 1.5269,
+      "step": 11784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.430236339569092,
+      "learning_rate": 9.23154239766082e-06,
+      "loss": 1.5424,
+      "step": 11785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1441712379455566,
+      "learning_rate": 9.230628654970762e-06,
+      "loss": 1.956,
+      "step": 11786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3140172958374023,
+      "learning_rate": 9.229714912280702e-06,
+      "loss": 1.4458,
+      "step": 11787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.484534978866577,
+      "learning_rate": 9.228801169590644e-06,
+      "loss": 1.4246,
+      "step": 11788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0374088287353516,
+      "learning_rate": 9.227887426900585e-06,
+      "loss": 1.6137,
+      "step": 11789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.263437509536743,
+      "learning_rate": 9.226973684210527e-06,
+      "loss": 1.2768,
+      "step": 11790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.01111102104187,
+      "learning_rate": 9.22605994152047e-06,
+      "loss": 1.8906,
+      "step": 11791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.631157636642456,
+      "learning_rate": 9.22514619883041e-06,
+      "loss": 1.43,
+      "step": 11792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5473597049713135,
+      "learning_rate": 9.224232456140352e-06,
+      "loss": 1.4347,
+      "step": 11793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6992387771606445,
+      "learning_rate": 9.223318713450293e-06,
+      "loss": 1.9465,
+      "step": 11794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2773756980895996,
+      "learning_rate": 9.222404970760235e-06,
+      "loss": 1.3943,
+      "step": 11795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4303927421569824,
+      "learning_rate": 9.221491228070175e-06,
+      "loss": 1.4873,
+      "step": 11796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1708574295043945,
+      "learning_rate": 9.220577485380118e-06,
+      "loss": 1.184,
+      "step": 11797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5645298957824707,
+      "learning_rate": 9.21966374269006e-06,
+      "loss": 1.5106,
+      "step": 11798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.923349380493164,
+      "learning_rate": 9.21875e-06,
+      "loss": 1.5479,
+      "step": 11799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.115635633468628,
+      "learning_rate": 9.217836257309943e-06,
+      "loss": 1.6687,
+      "step": 11800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.616318941116333,
+      "learning_rate": 9.216922514619883e-06,
+      "loss": 1.5505,
+      "step": 11801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8302111625671387,
+      "learning_rate": 9.216008771929825e-06,
+      "loss": 1.7202,
+      "step": 11802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.595799207687378,
+      "learning_rate": 9.215095029239766e-06,
+      "loss": 1.3677,
+      "step": 11803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.708040237426758,
+      "learning_rate": 9.214181286549708e-06,
+      "loss": 1.5499,
+      "step": 11804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2907330989837646,
+      "learning_rate": 9.21326754385965e-06,
+      "loss": 1.547,
+      "step": 11805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.795814037322998,
+      "learning_rate": 9.212353801169591e-06,
+      "loss": 1.8317,
+      "step": 11806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.419442653656006,
+      "learning_rate": 9.211440058479533e-06,
+      "loss": 1.335,
+      "step": 11807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.486562490463257,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 1.549,
+      "step": 11808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6548736095428467,
+      "learning_rate": 9.209612573099416e-06,
+      "loss": 1.9903,
+      "step": 11809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.134044885635376,
+      "learning_rate": 9.208698830409356e-06,
+      "loss": 1.8683,
+      "step": 11810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.80924654006958,
+      "learning_rate": 9.207785087719299e-06,
+      "loss": 1.4505,
+      "step": 11811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9744476079940796,
+      "learning_rate": 9.206871345029241e-06,
+      "loss": 1.1967,
+      "step": 11812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3201205730438232,
+      "learning_rate": 9.205957602339181e-06,
+      "loss": 1.0639,
+      "step": 11813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4733710289001465,
+      "learning_rate": 9.205043859649124e-06,
+      "loss": 1.1974,
+      "step": 11814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.473935127258301,
+      "learning_rate": 9.204130116959064e-06,
+      "loss": 1.5215,
+      "step": 11815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9195573329925537,
+      "learning_rate": 9.203216374269006e-06,
+      "loss": 1.5067,
+      "step": 11816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.764434814453125,
+      "learning_rate": 9.202302631578947e-06,
+      "loss": 1.5793,
+      "step": 11817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.304046154022217,
+      "learning_rate": 9.201388888888889e-06,
+      "loss": 1.323,
+      "step": 11818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4615087509155273,
+      "learning_rate": 9.200475146198831e-06,
+      "loss": 1.2754,
+      "step": 11819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5553691387176514,
+      "learning_rate": 9.199561403508772e-06,
+      "loss": 1.3437,
+      "step": 11820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.346756935119629,
+      "learning_rate": 9.198647660818714e-06,
+      "loss": 1.3273,
+      "step": 11821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6111512184143066,
+      "learning_rate": 9.197733918128656e-06,
+      "loss": 1.396,
+      "step": 11822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3380963802337646,
+      "learning_rate": 9.196820175438599e-06,
+      "loss": 1.427,
+      "step": 11823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.654482364654541,
+      "learning_rate": 9.195906432748539e-06,
+      "loss": 1.3391,
+      "step": 11824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4960415363311768,
+      "learning_rate": 9.194992690058481e-06,
+      "loss": 1.4689,
+      "step": 11825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.381099224090576,
+      "learning_rate": 9.194078947368422e-06,
+      "loss": 1.5168,
+      "step": 11826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4045238494873047,
+      "learning_rate": 9.193165204678364e-06,
+      "loss": 1.3054,
+      "step": 11827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7664363384246826,
+      "learning_rate": 9.192251461988305e-06,
+      "loss": 1.7203,
+      "step": 11828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.168388843536377,
+      "learning_rate": 9.191337719298247e-06,
+      "loss": 1.7723,
+      "step": 11829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.554959774017334,
+      "learning_rate": 9.190423976608189e-06,
+      "loss": 1.9089,
+      "step": 11830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4074997901916504,
+      "learning_rate": 9.18951023391813e-06,
+      "loss": 1.2137,
+      "step": 11831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2778217792510986,
+      "learning_rate": 9.188596491228072e-06,
+      "loss": 1.2442,
+      "step": 11832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2642862796783447,
+      "learning_rate": 9.187682748538012e-06,
+      "loss": 2.1596,
+      "step": 11833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5803630352020264,
+      "learning_rate": 9.186769005847954e-06,
+      "loss": 1.5717,
+      "step": 11834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4632325172424316,
+      "learning_rate": 9.185855263157895e-06,
+      "loss": 1.5287,
+      "step": 11835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.755485773086548,
+      "learning_rate": 9.184941520467837e-06,
+      "loss": 1.7234,
+      "step": 11836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9631460905075073,
+      "learning_rate": 9.18402777777778e-06,
+      "loss": 1.131,
+      "step": 11837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3248963356018066,
+      "learning_rate": 9.18311403508772e-06,
+      "loss": 1.5962,
+      "step": 11838
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.813229560852051,
+      "learning_rate": 9.182200292397662e-06,
+      "loss": 1.9533,
+      "step": 11839
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.22330641746521,
+      "learning_rate": 9.181286549707603e-06,
+      "loss": 1.9211,
+      "step": 11840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6383676528930664,
+      "learning_rate": 9.180372807017545e-06,
+      "loss": 1.6193,
+      "step": 11841
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.83636736869812,
+      "learning_rate": 9.179459064327485e-06,
+      "loss": 1.5816,
+      "step": 11842
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0376081466674805,
+      "learning_rate": 9.178545321637428e-06,
+      "loss": 1.2184,
+      "step": 11843
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3944084644317627,
+      "learning_rate": 9.17763157894737e-06,
+      "loss": 1.3869,
+      "step": 11844
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.801952838897705,
+      "learning_rate": 9.17671783625731e-06,
+      "loss": 1.816,
+      "step": 11845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2516279220581055,
+      "learning_rate": 9.175804093567253e-06,
+      "loss": 1.8052,
+      "step": 11846
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2878520488739014,
+      "learning_rate": 9.174890350877193e-06,
+      "loss": 2.0204,
+      "step": 11847
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.368133306503296,
+      "learning_rate": 9.173976608187135e-06,
+      "loss": 1.4571,
+      "step": 11848
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5903759002685547,
+      "learning_rate": 9.173062865497076e-06,
+      "loss": 1.3859,
+      "step": 11849
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.084733009338379,
+      "learning_rate": 9.172149122807018e-06,
+      "loss": 1.2582,
+      "step": 11850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7405264377593994,
+      "learning_rate": 9.17123538011696e-06,
+      "loss": 1.8613,
+      "step": 11851
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.874934434890747,
+      "learning_rate": 9.170321637426901e-06,
+      "loss": 1.8531,
+      "step": 11852
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.037097692489624,
+      "learning_rate": 9.169407894736843e-06,
+      "loss": 1.392,
+      "step": 11853
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9588801860809326,
+      "learning_rate": 9.168494152046784e-06,
+      "loss": 1.6063,
+      "step": 11854
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5804121494293213,
+      "learning_rate": 9.167580409356726e-06,
+      "loss": 1.624,
+      "step": 11855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7704832553863525,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 1.2731,
+      "step": 11856
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4044487476348877,
+      "learning_rate": 9.165752923976609e-06,
+      "loss": 1.346,
+      "step": 11857
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.510451555252075,
+      "learning_rate": 9.164839181286551e-06,
+      "loss": 1.5754,
+      "step": 11858
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.503185987472534,
+      "learning_rate": 9.163925438596491e-06,
+      "loss": 1.4257,
+      "step": 11859
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6473495960235596,
+      "learning_rate": 9.163011695906434e-06,
+      "loss": 1.6531,
+      "step": 11860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.761094570159912,
+      "learning_rate": 9.162097953216374e-06,
+      "loss": 1.6064,
+      "step": 11861
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7969624996185303,
+      "learning_rate": 9.161184210526316e-06,
+      "loss": 1.7995,
+      "step": 11862
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.575047016143799,
+      "learning_rate": 9.160270467836257e-06,
+      "loss": 1.3566,
+      "step": 11863
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7944438457489014,
+      "learning_rate": 9.1593567251462e-06,
+      "loss": 1.9626,
+      "step": 11864
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7713046073913574,
+      "learning_rate": 9.158442982456141e-06,
+      "loss": 1.4214,
+      "step": 11865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9133570194244385,
+      "learning_rate": 9.157529239766084e-06,
+      "loss": 1.4224,
+      "step": 11866
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8159568309783936,
+      "learning_rate": 9.156615497076024e-06,
+      "loss": 1.6846,
+      "step": 11867
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.893854856491089,
+      "learning_rate": 9.155701754385966e-06,
+      "loss": 1.3365,
+      "step": 11868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.235751152038574,
+      "learning_rate": 9.154788011695907e-06,
+      "loss": 1.3142,
+      "step": 11869
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4809157848358154,
+      "learning_rate": 9.153874269005849e-06,
+      "loss": 1.6153,
+      "step": 11870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7880585193634033,
+      "learning_rate": 9.152960526315791e-06,
+      "loss": 1.6472,
+      "step": 11871
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.809889793395996,
+      "learning_rate": 9.152046783625732e-06,
+      "loss": 1.5987,
+      "step": 11872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6008694171905518,
+      "learning_rate": 9.151133040935674e-06,
+      "loss": 1.6603,
+      "step": 11873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.570415496826172,
+      "learning_rate": 9.150219298245615e-06,
+      "loss": 1.5609,
+      "step": 11874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.292773485183716,
+      "learning_rate": 9.149305555555557e-06,
+      "loss": 1.3312,
+      "step": 11875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8755874633789062,
+      "learning_rate": 9.148391812865497e-06,
+      "loss": 1.653,
+      "step": 11876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.598618507385254,
+      "learning_rate": 9.14747807017544e-06,
+      "loss": 1.3798,
+      "step": 11877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2568933963775635,
+      "learning_rate": 9.146564327485382e-06,
+      "loss": 1.1619,
+      "step": 11878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1912403106689453,
+      "learning_rate": 9.145650584795322e-06,
+      "loss": 1.4799,
+      "step": 11879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1662094593048096,
+      "learning_rate": 9.144736842105264e-06,
+      "loss": 1.2711,
+      "step": 11880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.612030506134033,
+      "learning_rate": 9.143823099415205e-06,
+      "loss": 1.4921,
+      "step": 11881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2713735103607178,
+      "learning_rate": 9.142909356725147e-06,
+      "loss": 1.3336,
+      "step": 11882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6055829524993896,
+      "learning_rate": 9.141995614035088e-06,
+      "loss": 1.3882,
+      "step": 11883
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.578669548034668,
+      "learning_rate": 9.14108187134503e-06,
+      "loss": 1.4277,
+      "step": 11884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.50988507270813,
+      "learning_rate": 9.140168128654972e-06,
+      "loss": 1.3567,
+      "step": 11885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.280165195465088,
+      "learning_rate": 9.139254385964913e-06,
+      "loss": 1.4637,
+      "step": 11886
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2889134883880615,
+      "learning_rate": 9.138340643274855e-06,
+      "loss": 1.5455,
+      "step": 11887
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.539059638977051,
+      "learning_rate": 9.137426900584795e-06,
+      "loss": 1.2912,
+      "step": 11888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3494012355804443,
+      "learning_rate": 9.136513157894738e-06,
+      "loss": 1.5353,
+      "step": 11889
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3183629512786865,
+      "learning_rate": 9.135599415204678e-06,
+      "loss": 1.4679,
+      "step": 11890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.554858684539795,
+      "learning_rate": 9.13468567251462e-06,
+      "loss": 1.4641,
+      "step": 11891
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.246641159057617,
+      "learning_rate": 9.133771929824563e-06,
+      "loss": 1.2156,
+      "step": 11892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8766586780548096,
+      "learning_rate": 9.132858187134503e-06,
+      "loss": 1.7175,
+      "step": 11893
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.001156806945801,
+      "learning_rate": 9.131944444444445e-06,
+      "loss": 1.4681,
+      "step": 11894
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7010490894317627,
+      "learning_rate": 9.131030701754386e-06,
+      "loss": 1.391,
+      "step": 11895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0974032878875732,
+      "learning_rate": 9.130116959064328e-06,
+      "loss": 1.5331,
+      "step": 11896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.601058006286621,
+      "learning_rate": 9.129203216374269e-06,
+      "loss": 1.3998,
+      "step": 11897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7866618633270264,
+      "learning_rate": 9.128289473684211e-06,
+      "loss": 1.7463,
+      "step": 11898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.44038462638855,
+      "learning_rate": 9.127375730994153e-06,
+      "loss": 1.4471,
+      "step": 11899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.268061637878418,
+      "learning_rate": 9.126461988304094e-06,
+      "loss": 1.369,
+      "step": 11900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9163146018981934,
+      "learning_rate": 9.125548245614036e-06,
+      "loss": 1.0913,
+      "step": 11901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2148706912994385,
+      "learning_rate": 9.124634502923976e-06,
+      "loss": 1.3851,
+      "step": 11902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8595168590545654,
+      "learning_rate": 9.123720760233919e-06,
+      "loss": 1.5564,
+      "step": 11903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.660464286804199,
+      "learning_rate": 9.12280701754386e-06,
+      "loss": 1.5197,
+      "step": 11904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7289483547210693,
+      "learning_rate": 9.121893274853801e-06,
+      "loss": 1.6992,
+      "step": 11905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.018449068069458,
+      "learning_rate": 9.120979532163744e-06,
+      "loss": 1.778,
+      "step": 11906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1039576530456543,
+      "learning_rate": 9.120065789473686e-06,
+      "loss": 1.1442,
+      "step": 11907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3999478816986084,
+      "learning_rate": 9.119152046783626e-06,
+      "loss": 1.6935,
+      "step": 11908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.975973129272461,
+      "learning_rate": 9.118238304093569e-06,
+      "loss": 1.8421,
+      "step": 11909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3178625106811523,
+      "learning_rate": 9.11732456140351e-06,
+      "loss": 1.2944,
+      "step": 11910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7454612255096436,
+      "learning_rate": 9.116410818713451e-06,
+      "loss": 1.7231,
+      "step": 11911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3607594966888428,
+      "learning_rate": 9.115497076023394e-06,
+      "loss": 1.4073,
+      "step": 11912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9551589488983154,
+      "learning_rate": 9.114583333333334e-06,
+      "loss": 1.6494,
+      "step": 11913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1848251819610596,
+      "learning_rate": 9.113669590643276e-06,
+      "loss": 1.7,
+      "step": 11914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8179874420166016,
+      "learning_rate": 9.112755847953217e-06,
+      "loss": 1.5249,
+      "step": 11915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.934446334838867,
+      "learning_rate": 9.111842105263159e-06,
+      "loss": 1.5396,
+      "step": 11916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4626400470733643,
+      "learning_rate": 9.1109283625731e-06,
+      "loss": 1.5828,
+      "step": 11917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8577632904052734,
+      "learning_rate": 9.110014619883042e-06,
+      "loss": 1.3805,
+      "step": 11918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.818542003631592,
+      "learning_rate": 9.109100877192984e-06,
+      "loss": 1.2788,
+      "step": 11919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.418463945388794,
+      "learning_rate": 9.108187134502925e-06,
+      "loss": 1.5631,
+      "step": 11920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5794565677642822,
+      "learning_rate": 9.107273391812867e-06,
+      "loss": 1.341,
+      "step": 11921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.020220994949341,
+      "learning_rate": 9.106359649122807e-06,
+      "loss": 1.2571,
+      "step": 11922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.422866106033325,
+      "learning_rate": 9.10544590643275e-06,
+      "loss": 1.466,
+      "step": 11923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.421417474746704,
+      "learning_rate": 9.10453216374269e-06,
+      "loss": 1.3353,
+      "step": 11924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4115333557128906,
+      "learning_rate": 9.103618421052632e-06,
+      "loss": 1.4671,
+      "step": 11925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.040747880935669,
+      "learning_rate": 9.102704678362574e-06,
+      "loss": 1.5111,
+      "step": 11926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.907780647277832,
+      "learning_rate": 9.101790935672515e-06,
+      "loss": 1.6901,
+      "step": 11927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2747175693511963,
+      "learning_rate": 9.100877192982457e-06,
+      "loss": 1.0645,
+      "step": 11928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8952200412750244,
+      "learning_rate": 9.099963450292398e-06,
+      "loss": 1.5313,
+      "step": 11929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3430585861206055,
+      "learning_rate": 9.09904970760234e-06,
+      "loss": 1.5332,
+      "step": 11930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.821134090423584,
+      "learning_rate": 9.09813596491228e-06,
+      "loss": 1.5409,
+      "step": 11931
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.287369966506958,
+      "learning_rate": 9.097222222222223e-06,
+      "loss": 1.3546,
+      "step": 11932
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.905579090118408,
+      "learning_rate": 9.096308479532165e-06,
+      "loss": 1.7274,
+      "step": 11933
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.366166830062866,
+      "learning_rate": 9.095394736842105e-06,
+      "loss": 1.6901,
+      "step": 11934
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7730026245117188,
+      "learning_rate": 9.094480994152048e-06,
+      "loss": 1.492,
+      "step": 11935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.307948589324951,
+      "learning_rate": 9.093567251461988e-06,
+      "loss": 1.3537,
+      "step": 11936
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2065649032592773,
+      "learning_rate": 9.09265350877193e-06,
+      "loss": 1.4255,
+      "step": 11937
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1845645904541016,
+      "learning_rate": 9.091739766081871e-06,
+      "loss": 1.7899,
+      "step": 11938
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.512906074523926,
+      "learning_rate": 9.090826023391813e-06,
+      "loss": 1.341,
+      "step": 11939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6936051845550537,
+      "learning_rate": 9.089912280701755e-06,
+      "loss": 1.6245,
+      "step": 11940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9250121116638184,
+      "learning_rate": 9.088998538011696e-06,
+      "loss": 1.5994,
+      "step": 11941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.231736421585083,
+      "learning_rate": 9.088084795321638e-06,
+      "loss": 1.3348,
+      "step": 11942
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.071162462234497,
+      "learning_rate": 9.087171052631579e-06,
+      "loss": 1.824,
+      "step": 11943
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.731173276901245,
+      "learning_rate": 9.086257309941521e-06,
+      "loss": 1.557,
+      "step": 11944
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5445408821105957,
+      "learning_rate": 9.085343567251461e-06,
+      "loss": 1.2619,
+      "step": 11945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4550576210021973,
+      "learning_rate": 9.084429824561404e-06,
+      "loss": 1.201,
+      "step": 11946
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.75665020942688,
+      "learning_rate": 9.083516081871346e-06,
+      "loss": 1.6748,
+      "step": 11947
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1111056804656982,
+      "learning_rate": 9.082602339181286e-06,
+      "loss": 1.279,
+      "step": 11948
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.574707508087158,
+      "learning_rate": 9.081688596491229e-06,
+      "loss": 1.3771,
+      "step": 11949
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.178809404373169,
+      "learning_rate": 9.080774853801171e-06,
+      "loss": 1.694,
+      "step": 11950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.009568691253662,
+      "learning_rate": 9.079861111111113e-06,
+      "loss": 1.5503,
+      "step": 11951
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4361486434936523,
+      "learning_rate": 9.078947368421054e-06,
+      "loss": 1.0902,
+      "step": 11952
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.598881483078003,
+      "learning_rate": 9.078033625730996e-06,
+      "loss": 1.5132,
+      "step": 11953
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.49810791015625,
+      "learning_rate": 9.077119883040936e-06,
+      "loss": 1.0812,
+      "step": 11954
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.294708013534546,
+      "learning_rate": 9.076206140350879e-06,
+      "loss": 1.4161,
+      "step": 11955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.627894401550293,
+      "learning_rate": 9.075292397660819e-06,
+      "loss": 1.4527,
+      "step": 11956
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.932476758956909,
+      "learning_rate": 9.074378654970761e-06,
+      "loss": 1.6387,
+      "step": 11957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.906651020050049,
+      "learning_rate": 9.073464912280704e-06,
+      "loss": 1.2118,
+      "step": 11958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3096115589141846,
+      "learning_rate": 9.072551169590644e-06,
+      "loss": 1.2781,
+      "step": 11959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7728376388549805,
+      "learning_rate": 9.071637426900586e-06,
+      "loss": 1.5309,
+      "step": 11960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.701648712158203,
+      "learning_rate": 9.070723684210527e-06,
+      "loss": 1.5569,
+      "step": 11961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6111881732940674,
+      "learning_rate": 9.069809941520469e-06,
+      "loss": 1.5761,
+      "step": 11962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6024043560028076,
+      "learning_rate": 9.06889619883041e-06,
+      "loss": 1.4507,
+      "step": 11963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.76543927192688,
+      "learning_rate": 9.067982456140352e-06,
+      "loss": 1.3876,
+      "step": 11964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8338265419006348,
+      "learning_rate": 9.067068713450294e-06,
+      "loss": 1.7142,
+      "step": 11965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.503364086151123,
+      "learning_rate": 9.066154970760235e-06,
+      "loss": 1.6097,
+      "step": 11966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.427215099334717,
+      "learning_rate": 9.065241228070177e-06,
+      "loss": 1.2832,
+      "step": 11967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9514758586883545,
+      "learning_rate": 9.064327485380117e-06,
+      "loss": 1.2844,
+      "step": 11968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.407259941101074,
+      "learning_rate": 9.06341374269006e-06,
+      "loss": 1.3279,
+      "step": 11969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5813417434692383,
+      "learning_rate": 9.0625e-06,
+      "loss": 1.5716,
+      "step": 11970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4094767570495605,
+      "learning_rate": 9.061586257309942e-06,
+      "loss": 1.5194,
+      "step": 11971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.636847972869873,
+      "learning_rate": 9.060672514619884e-06,
+      "loss": 1.5181,
+      "step": 11972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9379351139068604,
+      "learning_rate": 9.059758771929825e-06,
+      "loss": 1.4391,
+      "step": 11973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8831183910369873,
+      "learning_rate": 9.058845029239767e-06,
+      "loss": 1.7341,
+      "step": 11974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5486886501312256,
+      "learning_rate": 9.057931286549708e-06,
+      "loss": 1.3766,
+      "step": 11975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5746424198150635,
+      "learning_rate": 9.05701754385965e-06,
+      "loss": 1.4665,
+      "step": 11976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6857595443725586,
+      "learning_rate": 9.05610380116959e-06,
+      "loss": 1.4668,
+      "step": 11977
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4591562747955322,
+      "learning_rate": 9.055190058479533e-06,
+      "loss": 1.518,
+      "step": 11978
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.117690086364746,
+      "learning_rate": 9.054276315789475e-06,
+      "loss": 1.6657,
+      "step": 11979
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8438124656677246,
+      "learning_rate": 9.053362573099415e-06,
+      "loss": 1.5706,
+      "step": 11980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8086564540863037,
+      "learning_rate": 9.052448830409358e-06,
+      "loss": 1.6539,
+      "step": 11981
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.257551431655884,
+      "learning_rate": 9.051535087719298e-06,
+      "loss": 1.5329,
+      "step": 11982
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.738222599029541,
+      "learning_rate": 9.05062134502924e-06,
+      "loss": 1.4106,
+      "step": 11983
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1786160469055176,
+      "learning_rate": 9.049707602339181e-06,
+      "loss": 2.1658,
+      "step": 11984
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6455531120300293,
+      "learning_rate": 9.048793859649123e-06,
+      "loss": 1.632,
+      "step": 11985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.813887357711792,
+      "learning_rate": 9.047880116959065e-06,
+      "loss": 1.5197,
+      "step": 11986
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0566859245300293,
+      "learning_rate": 9.046966374269006e-06,
+      "loss": 1.7217,
+      "step": 11987
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3499505519866943,
+      "learning_rate": 9.046052631578948e-06,
+      "loss": 1.44,
+      "step": 11988
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8216919898986816,
+      "learning_rate": 9.045138888888889e-06,
+      "loss": 1.4993,
+      "step": 11989
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.766641616821289,
+      "learning_rate": 9.044225146198831e-06,
+      "loss": 1.6442,
+      "step": 11990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.50042986869812,
+      "learning_rate": 9.043311403508771e-06,
+      "loss": 1.6957,
+      "step": 11991
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5373282432556152,
+      "learning_rate": 9.042397660818715e-06,
+      "loss": 1.4359,
+      "step": 11992
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9323316812515259,
+      "learning_rate": 9.041483918128656e-06,
+      "loss": 1.2043,
+      "step": 11993
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.139030933380127,
+      "learning_rate": 9.040570175438598e-06,
+      "loss": 1.3351,
+      "step": 11994
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6221492290496826,
+      "learning_rate": 9.039656432748539e-06,
+      "loss": 1.5433,
+      "step": 11995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.189865827560425,
+      "learning_rate": 9.038742690058481e-06,
+      "loss": 1.2196,
+      "step": 11996
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.106976270675659,
+      "learning_rate": 9.037828947368421e-06,
+      "loss": 1.2376,
+      "step": 11997
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.017576217651367,
+      "learning_rate": 9.036915204678364e-06,
+      "loss": 1.1725,
+      "step": 11998
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6218061447143555,
+      "learning_rate": 9.036001461988306e-06,
+      "loss": 1.6037,
+      "step": 11999
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2857377529144287,
+      "learning_rate": 9.035087719298246e-06,
+      "loss": 1.3456,
+      "step": 12000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.417322874069214,
+      "learning_rate": 9.034173976608189e-06,
+      "loss": 1.2765,
+      "step": 12001
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0002353191375732,
+      "learning_rate": 9.033260233918129e-06,
+      "loss": 1.2356,
+      "step": 12002
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4176034927368164,
+      "learning_rate": 9.032346491228071e-06,
+      "loss": 1.1109,
+      "step": 12003
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.577981472015381,
+      "learning_rate": 9.031432748538012e-06,
+      "loss": 1.1948,
+      "step": 12004
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.815150499343872,
+      "learning_rate": 9.030519005847954e-06,
+      "loss": 1.6673,
+      "step": 12005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9545137882232666,
+      "learning_rate": 9.029605263157896e-06,
+      "loss": 1.691,
+      "step": 12006
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.402167320251465,
+      "learning_rate": 9.028691520467837e-06,
+      "loss": 1.3601,
+      "step": 12007
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.540590286254883,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 1.3945,
+      "step": 12008
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.640204429626465,
+      "learning_rate": 9.02686403508772e-06,
+      "loss": 1.4702,
+      "step": 12009
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4409661293029785,
+      "learning_rate": 9.025950292397662e-06,
+      "loss": 1.2834,
+      "step": 12010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.965930461883545,
+      "learning_rate": 9.025036549707602e-06,
+      "loss": 1.6647,
+      "step": 12011
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8204457759857178,
+      "learning_rate": 9.024122807017545e-06,
+      "loss": 1.2619,
+      "step": 12012
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.484431266784668,
+      "learning_rate": 9.023209064327487e-06,
+      "loss": 1.6757,
+      "step": 12013
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3125476837158203,
+      "learning_rate": 9.022295321637427e-06,
+      "loss": 1.4936,
+      "step": 12014
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.362539291381836,
+      "learning_rate": 9.02138157894737e-06,
+      "loss": 1.4083,
+      "step": 12015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0170531272888184,
+      "learning_rate": 9.02046783625731e-06,
+      "loss": 1.695,
+      "step": 12016
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9565651416778564,
+      "learning_rate": 9.019554093567252e-06,
+      "loss": 1.6217,
+      "step": 12017
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.712880849838257,
+      "learning_rate": 9.018640350877193e-06,
+      "loss": 1.5658,
+      "step": 12018
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1736464500427246,
+      "learning_rate": 9.017726608187135e-06,
+      "loss": 1.3355,
+      "step": 12019
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.131381034851074,
+      "learning_rate": 9.016812865497077e-06,
+      "loss": 2.1834,
+      "step": 12020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8069541454315186,
+      "learning_rate": 9.015899122807018e-06,
+      "loss": 1.616,
+      "step": 12021
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5625407695770264,
+      "learning_rate": 9.01498538011696e-06,
+      "loss": 1.2574,
+      "step": 12022
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2792763710021973,
+      "learning_rate": 9.0140716374269e-06,
+      "loss": 1.3278,
+      "step": 12023
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.034348726272583,
+      "learning_rate": 9.013157894736843e-06,
+      "loss": 1.1899,
+      "step": 12024
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6429529190063477,
+      "learning_rate": 9.012244152046783e-06,
+      "loss": 1.6667,
+      "step": 12025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3542747497558594,
+      "learning_rate": 9.011330409356725e-06,
+      "loss": 1.3482,
+      "step": 12026
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7250630855560303,
+      "learning_rate": 9.010416666666668e-06,
+      "loss": 1.9189,
+      "step": 12027
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.613887071609497,
+      "learning_rate": 9.009502923976608e-06,
+      "loss": 1.4732,
+      "step": 12028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.652946949005127,
+      "learning_rate": 9.00858918128655e-06,
+      "loss": 1.7256,
+      "step": 12029
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0738654136657715,
+      "learning_rate": 9.007675438596491e-06,
+      "loss": 1.65,
+      "step": 12030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6256771087646484,
+      "learning_rate": 9.006761695906433e-06,
+      "loss": 1.6395,
+      "step": 12031
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1069207191467285,
+      "learning_rate": 9.005847953216374e-06,
+      "loss": 1.4999,
+      "step": 12032
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4860005378723145,
+      "learning_rate": 9.004934210526316e-06,
+      "loss": 1.5073,
+      "step": 12033
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8765628337860107,
+      "learning_rate": 9.004020467836258e-06,
+      "loss": 2.0262,
+      "step": 12034
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.365112543106079,
+      "learning_rate": 9.0031067251462e-06,
+      "loss": 1.5399,
+      "step": 12035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.933641195297241,
+      "learning_rate": 9.002192982456141e-06,
+      "loss": 1.7777,
+      "step": 12036
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0650556087493896,
+      "learning_rate": 9.001279239766083e-06,
+      "loss": 1.1957,
+      "step": 12037
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.47188138961792,
+      "learning_rate": 9.000365497076025e-06,
+      "loss": 1.6969,
+      "step": 12038
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.654914140701294,
+      "learning_rate": 8.999451754385966e-06,
+      "loss": 1.5432,
+      "step": 12039
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.076658248901367,
+      "learning_rate": 8.998538011695908e-06,
+      "loss": 1.2965,
+      "step": 12040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.439317464828491,
+      "learning_rate": 8.997624269005849e-06,
+      "loss": 1.9076,
+      "step": 12041
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.411454677581787,
+      "learning_rate": 8.996710526315791e-06,
+      "loss": 1.4143,
+      "step": 12042
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.124197244644165,
+      "learning_rate": 8.995796783625731e-06,
+      "loss": 1.3002,
+      "step": 12043
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2147865295410156,
+      "learning_rate": 8.994883040935674e-06,
+      "loss": 1.6141,
+      "step": 12044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.439301013946533,
+      "learning_rate": 8.993969298245616e-06,
+      "loss": 1.2906,
+      "step": 12045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.439732551574707,
+      "learning_rate": 8.993055555555556e-06,
+      "loss": 1.5594,
+      "step": 12046
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.316892385482788,
+      "learning_rate": 8.992141812865499e-06,
+      "loss": 1.451,
+      "step": 12047
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0947484970092773,
+      "learning_rate": 8.991228070175439e-06,
+      "loss": 1.4316,
+      "step": 12048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.313016653060913,
+      "learning_rate": 8.990314327485381e-06,
+      "loss": 1.227,
+      "step": 12049
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6537110805511475,
+      "learning_rate": 8.989400584795322e-06,
+      "loss": 1.1197,
+      "step": 12050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8661582469940186,
+      "learning_rate": 8.988486842105264e-06,
+      "loss": 1.982,
+      "step": 12051
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.028895378112793,
+      "learning_rate": 8.987573099415206e-06,
+      "loss": 1.1328,
+      "step": 12052
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.271505355834961,
+      "learning_rate": 8.986659356725147e-06,
+      "loss": 1.4376,
+      "step": 12053
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1884243488311768,
+      "learning_rate": 8.985745614035089e-06,
+      "loss": 1.8949,
+      "step": 12054
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4825973510742188,
+      "learning_rate": 8.98483187134503e-06,
+      "loss": 1.4251,
+      "step": 12055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.095762014389038,
+      "learning_rate": 8.983918128654972e-06,
+      "loss": 1.7889,
+      "step": 12056
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3505115509033203,
+      "learning_rate": 8.983004385964912e-06,
+      "loss": 1.9705,
+      "step": 12057
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3716442584991455,
+      "learning_rate": 8.982090643274855e-06,
+      "loss": 0.9943,
+      "step": 12058
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.52962064743042,
+      "learning_rate": 8.981176900584797e-06,
+      "loss": 1.4978,
+      "step": 12059
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1655077934265137,
+      "learning_rate": 8.980263157894737e-06,
+      "loss": 1.2754,
+      "step": 12060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2791218757629395,
+      "learning_rate": 8.97934941520468e-06,
+      "loss": 1.3014,
+      "step": 12061
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4385271072387695,
+      "learning_rate": 8.97843567251462e-06,
+      "loss": 1.4144,
+      "step": 12062
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.911802291870117,
+      "learning_rate": 8.977521929824562e-06,
+      "loss": 1.1915,
+      "step": 12063
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1212289333343506,
+      "learning_rate": 8.976608187134503e-06,
+      "loss": 1.045,
+      "step": 12064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.218637704849243,
+      "learning_rate": 8.975694444444445e-06,
+      "loss": 1.4019,
+      "step": 12065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6773343086242676,
+      "learning_rate": 8.974780701754387e-06,
+      "loss": 1.165,
+      "step": 12066
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8169853687286377,
+      "learning_rate": 8.973866959064328e-06,
+      "loss": 1.7297,
+      "step": 12067
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8291752338409424,
+      "learning_rate": 8.97295321637427e-06,
+      "loss": 1.6661,
+      "step": 12068
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.462205648422241,
+      "learning_rate": 8.97203947368421e-06,
+      "loss": 1.2007,
+      "step": 12069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5445830821990967,
+      "learning_rate": 8.971125730994153e-06,
+      "loss": 1.4241,
+      "step": 12070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.344123601913452,
+      "learning_rate": 8.970211988304093e-06,
+      "loss": 1.396,
+      "step": 12071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5112996101379395,
+      "learning_rate": 8.969298245614035e-06,
+      "loss": 1.4853,
+      "step": 12072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.761460542678833,
+      "learning_rate": 8.968384502923978e-06,
+      "loss": 1.6147,
+      "step": 12073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4741322994232178,
+      "learning_rate": 8.967470760233918e-06,
+      "loss": 1.2689,
+      "step": 12074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7427499294281006,
+      "learning_rate": 8.96655701754386e-06,
+      "loss": 1.7271,
+      "step": 12075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.569270133972168,
+      "learning_rate": 8.965643274853801e-06,
+      "loss": 1.8712,
+      "step": 12076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.406451940536499,
+      "learning_rate": 8.964729532163743e-06,
+      "loss": 1.4309,
+      "step": 12077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2351975440979004,
+      "learning_rate": 8.963815789473685e-06,
+      "loss": 1.3183,
+      "step": 12078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6002230644226074,
+      "learning_rate": 8.962902046783628e-06,
+      "loss": 1.4455,
+      "step": 12079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7398412227630615,
+      "learning_rate": 8.961988304093568e-06,
+      "loss": 1.8217,
+      "step": 12080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5480234622955322,
+      "learning_rate": 8.96107456140351e-06,
+      "loss": 1.3146,
+      "step": 12081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0922915935516357,
+      "learning_rate": 8.960160818713451e-06,
+      "loss": 1.3068,
+      "step": 12082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.428903579711914,
+      "learning_rate": 8.959247076023393e-06,
+      "loss": 1.3202,
+      "step": 12083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6551311016082764,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 1.424,
+      "step": 12084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.72812557220459,
+      "learning_rate": 8.957419590643276e-06,
+      "loss": 1.6493,
+      "step": 12085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.21638560295105,
+      "learning_rate": 8.956505847953218e-06,
+      "loss": 1.5397,
+      "step": 12086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6325361728668213,
+      "learning_rate": 8.955592105263159e-06,
+      "loss": 1.4839,
+      "step": 12087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.572293519973755,
+      "learning_rate": 8.954678362573101e-06,
+      "loss": 1.8166,
+      "step": 12088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2424347400665283,
+      "learning_rate": 8.953764619883041e-06,
+      "loss": 1.435,
+      "step": 12089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8061673641204834,
+      "learning_rate": 8.952850877192984e-06,
+      "loss": 1.8667,
+      "step": 12090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.573720693588257,
+      "learning_rate": 8.951937134502924e-06,
+      "loss": 1.5195,
+      "step": 12091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7315638065338135,
+      "learning_rate": 8.951023391812866e-06,
+      "loss": 1.4537,
+      "step": 12092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.884221076965332,
+      "learning_rate": 8.950109649122809e-06,
+      "loss": 1.185,
+      "step": 12093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.224900484085083,
+      "learning_rate": 8.949195906432749e-06,
+      "loss": 1.3758,
+      "step": 12094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5961546897888184,
+      "learning_rate": 8.948282163742691e-06,
+      "loss": 1.5837,
+      "step": 12095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3354175090789795,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 1.7051,
+      "step": 12096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.731234312057495,
+      "learning_rate": 8.946454678362574e-06,
+      "loss": 1.508,
+      "step": 12097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9829477071762085,
+      "learning_rate": 8.945540935672515e-06,
+      "loss": 1.103,
+      "step": 12098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7973437309265137,
+      "learning_rate": 8.944627192982457e-06,
+      "loss": 1.5294,
+      "step": 12099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.139948606491089,
+      "learning_rate": 8.943713450292399e-06,
+      "loss": 1.6992,
+      "step": 12100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1842737197875977,
+      "learning_rate": 8.94279970760234e-06,
+      "loss": 1.2272,
+      "step": 12101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8276824951171875,
+      "learning_rate": 8.941885964912282e-06,
+      "loss": 1.8323,
+      "step": 12102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.029867172241211,
+      "learning_rate": 8.940972222222222e-06,
+      "loss": 1.875,
+      "step": 12103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4981894493103027,
+      "learning_rate": 8.940058479532165e-06,
+      "loss": 1.4352,
+      "step": 12104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.302189588546753,
+      "learning_rate": 8.939144736842105e-06,
+      "loss": 1.3471,
+      "step": 12105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7972989082336426,
+      "learning_rate": 8.938230994152047e-06,
+      "loss": 1.4571,
+      "step": 12106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.358865261077881,
+      "learning_rate": 8.93731725146199e-06,
+      "loss": 1.4536,
+      "step": 12107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4446380138397217,
+      "learning_rate": 8.93640350877193e-06,
+      "loss": 1.4675,
+      "step": 12108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.009838342666626,
+      "learning_rate": 8.935489766081872e-06,
+      "loss": 1.1003,
+      "step": 12109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.370797872543335,
+      "learning_rate": 8.934576023391813e-06,
+      "loss": 1.145,
+      "step": 12110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8888068199157715,
+      "learning_rate": 8.933662280701755e-06,
+      "loss": 1.808,
+      "step": 12111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7837250232696533,
+      "learning_rate": 8.932748538011696e-06,
+      "loss": 1.5686,
+      "step": 12112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.220154047012329,
+      "learning_rate": 8.931834795321638e-06,
+      "loss": 1.2851,
+      "step": 12113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9865903854370117,
+      "learning_rate": 8.93092105263158e-06,
+      "loss": 1.1366,
+      "step": 12114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.813976526260376,
+      "learning_rate": 8.93000730994152e-06,
+      "loss": 1.5514,
+      "step": 12115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.362391948699951,
+      "learning_rate": 8.929093567251463e-06,
+      "loss": 1.2367,
+      "step": 12116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8714964389801025,
+      "learning_rate": 8.928179824561403e-06,
+      "loss": 1.8971,
+      "step": 12117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.420403480529785,
+      "learning_rate": 8.927266081871345e-06,
+      "loss": 1.1127,
+      "step": 12118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.708503484725952,
+      "learning_rate": 8.926352339181286e-06,
+      "loss": 1.4684,
+      "step": 12119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9349381923675537,
+      "learning_rate": 8.92543859649123e-06,
+      "loss": 1.3328,
+      "step": 12120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.107671022415161,
+      "learning_rate": 8.92452485380117e-06,
+      "loss": 1.6411,
+      "step": 12121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1645045280456543,
+      "learning_rate": 8.923611111111113e-06,
+      "loss": 1.0962,
+      "step": 12122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7300190925598145,
+      "learning_rate": 8.922697368421053e-06,
+      "loss": 1.4395,
+      "step": 12123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.360269784927368,
+      "learning_rate": 8.921783625730995e-06,
+      "loss": 1.1884,
+      "step": 12124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.639227867126465,
+      "learning_rate": 8.920869883040938e-06,
+      "loss": 1.7323,
+      "step": 12125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.468086004257202,
+      "learning_rate": 8.919956140350878e-06,
+      "loss": 2.3738,
+      "step": 12126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4262866973876953,
+      "learning_rate": 8.91904239766082e-06,
+      "loss": 1.2642,
+      "step": 12127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.817093849182129,
+      "learning_rate": 8.918128654970761e-06,
+      "loss": 1.7998,
+      "step": 12128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7417850494384766,
+      "learning_rate": 8.917214912280703e-06,
+      "loss": 1.4788,
+      "step": 12129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2043159008026123,
+      "learning_rate": 8.916301169590644e-06,
+      "loss": 1.4388,
+      "step": 12130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.476452112197876,
+      "learning_rate": 8.915387426900586e-06,
+      "loss": 1.474,
+      "step": 12131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1425740718841553,
+      "learning_rate": 8.914473684210528e-06,
+      "loss": 2.0448,
+      "step": 12132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8716506958007812,
+      "learning_rate": 8.913559941520469e-06,
+      "loss": 1.5983,
+      "step": 12133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5260422229766846,
+      "learning_rate": 8.912646198830411e-06,
+      "loss": 1.4684,
+      "step": 12134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.17016339302063,
+      "learning_rate": 8.911732456140351e-06,
+      "loss": 1.4581,
+      "step": 12135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1958932876586914,
+      "learning_rate": 8.910818713450294e-06,
+      "loss": 1.3682,
+      "step": 12136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.426551103591919,
+      "learning_rate": 8.909904970760234e-06,
+      "loss": 1.9129,
+      "step": 12137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4259607791900635,
+      "learning_rate": 8.908991228070176e-06,
+      "loss": 1.6089,
+      "step": 12138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6334621906280518,
+      "learning_rate": 8.908077485380117e-06,
+      "loss": 1.6441,
+      "step": 12139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.543384313583374,
+      "learning_rate": 8.907163742690059e-06,
+      "loss": 1.5067,
+      "step": 12140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.17569899559021,
+      "learning_rate": 8.906250000000001e-06,
+      "loss": 1.3156,
+      "step": 12141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.250396966934204,
+      "learning_rate": 8.905336257309942e-06,
+      "loss": 1.6139,
+      "step": 12142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.522597312927246,
+      "learning_rate": 8.904422514619884e-06,
+      "loss": 1.56,
+      "step": 12143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.433488130569458,
+      "learning_rate": 8.903508771929825e-06,
+      "loss": 1.2542,
+      "step": 12144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6313703060150146,
+      "learning_rate": 8.902595029239767e-06,
+      "loss": 1.3739,
+      "step": 12145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.932112693786621,
+      "learning_rate": 8.901681286549707e-06,
+      "loss": 1.6983,
+      "step": 12146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2422091960906982,
+      "learning_rate": 8.90076754385965e-06,
+      "loss": 1.2306,
+      "step": 12147
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3174052238464355,
+      "learning_rate": 8.899853801169592e-06,
+      "loss": 1.2857,
+      "step": 12148
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.985459566116333,
+      "learning_rate": 8.898940058479532e-06,
+      "loss": 1.9136,
+      "step": 12149
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8564682006835938,
+      "learning_rate": 8.898026315789475e-06,
+      "loss": 1.5732,
+      "step": 12150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.976294755935669,
+      "learning_rate": 8.897112573099415e-06,
+      "loss": 1.2402,
+      "step": 12151
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6508965492248535,
+      "learning_rate": 8.896198830409357e-06,
+      "loss": 1.5119,
+      "step": 12152
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6672074794769287,
+      "learning_rate": 8.895285087719298e-06,
+      "loss": 1.6459,
+      "step": 12153
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.443406105041504,
+      "learning_rate": 8.89437134502924e-06,
+      "loss": 1.2111,
+      "step": 12154
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.631031036376953,
+      "learning_rate": 8.893457602339182e-06,
+      "loss": 1.7296,
+      "step": 12155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.265817165374756,
+      "learning_rate": 8.892543859649123e-06,
+      "loss": 1.3767,
+      "step": 12156
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5611510276794434,
+      "learning_rate": 8.891630116959065e-06,
+      "loss": 1.4288,
+      "step": 12157
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.510133743286133,
+      "learning_rate": 8.890716374269006e-06,
+      "loss": 1.4468,
+      "step": 12158
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.922666072845459,
+      "learning_rate": 8.889802631578948e-06,
+      "loss": 2.0631,
+      "step": 12159
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3755369186401367,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.3285,
+      "step": 12160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3848226070404053,
+      "learning_rate": 8.88797514619883e-06,
+      "loss": 1.1628,
+      "step": 12161
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4656829833984375,
+      "learning_rate": 8.887061403508773e-06,
+      "loss": 1.4897,
+      "step": 12162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5883281230926514,
+      "learning_rate": 8.886147660818715e-06,
+      "loss": 1.6501,
+      "step": 12163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.50791072845459,
+      "learning_rate": 8.885233918128655e-06,
+      "loss": 1.5387,
+      "step": 12164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.719663381576538,
+      "learning_rate": 8.884320175438598e-06,
+      "loss": 1.7582,
+      "step": 12165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9047083854675293,
+      "learning_rate": 8.88340643274854e-06,
+      "loss": 1.7218,
+      "step": 12166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0137829780578613,
+      "learning_rate": 8.88249269005848e-06,
+      "loss": 1.4498,
+      "step": 12167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.693997621536255,
+      "learning_rate": 8.881578947368423e-06,
+      "loss": 1.4806,
+      "step": 12168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4944117069244385,
+      "learning_rate": 8.880665204678363e-06,
+      "loss": 1.7684,
+      "step": 12169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3624393939971924,
+      "learning_rate": 8.879751461988305e-06,
+      "loss": 1.2779,
+      "step": 12170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.054023265838623,
+      "learning_rate": 8.878837719298246e-06,
+      "loss": 1.4279,
+      "step": 12171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.128059148788452,
+      "learning_rate": 8.877923976608188e-06,
+      "loss": 1.2078,
+      "step": 12172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2037527561187744,
+      "learning_rate": 8.87701023391813e-06,
+      "loss": 1.303,
+      "step": 12173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.398684024810791,
+      "learning_rate": 8.876096491228071e-06,
+      "loss": 1.2525,
+      "step": 12174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0149264335632324,
+      "learning_rate": 8.875182748538013e-06,
+      "loss": 2.1866,
+      "step": 12175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.316262722015381,
+      "learning_rate": 8.874269005847954e-06,
+      "loss": 1.0625,
+      "step": 12176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6471590995788574,
+      "learning_rate": 8.873355263157896e-06,
+      "loss": 1.3887,
+      "step": 12177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3483095169067383,
+      "learning_rate": 8.872441520467836e-06,
+      "loss": 1.6506,
+      "step": 12178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0936532020568848,
+      "learning_rate": 8.871527777777779e-06,
+      "loss": 1.1533,
+      "step": 12179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4835495948791504,
+      "learning_rate": 8.870614035087721e-06,
+      "loss": 1.6152,
+      "step": 12180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9792861938476562,
+      "learning_rate": 8.869700292397661e-06,
+      "loss": 1.6628,
+      "step": 12181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.666705846786499,
+      "learning_rate": 8.868786549707604e-06,
+      "loss": 1.4965,
+      "step": 12182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3643829822540283,
+      "learning_rate": 8.867872807017544e-06,
+      "loss": 1.4803,
+      "step": 12183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6047327518463135,
+      "learning_rate": 8.866959064327486e-06,
+      "loss": 1.562,
+      "step": 12184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8708889484405518,
+      "learning_rate": 8.866045321637427e-06,
+      "loss": 1.724,
+      "step": 12185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2188937664031982,
+      "learning_rate": 8.865131578947369e-06,
+      "loss": 1.8294,
+      "step": 12186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7070116996765137,
+      "learning_rate": 8.864217836257311e-06,
+      "loss": 1.3965,
+      "step": 12187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.63218092918396,
+      "learning_rate": 8.863304093567252e-06,
+      "loss": 1.5775,
+      "step": 12188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.361240863800049,
+      "learning_rate": 8.862390350877194e-06,
+      "loss": 1.3867,
+      "step": 12189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.389111042022705,
+      "learning_rate": 8.861476608187135e-06,
+      "loss": 2.0671,
+      "step": 12190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8728811740875244,
+      "learning_rate": 8.860562865497077e-06,
+      "loss": 1.7875,
+      "step": 12191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8197896480560303,
+      "learning_rate": 8.859649122807017e-06,
+      "loss": 1.9787,
+      "step": 12192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4322586059570312,
+      "learning_rate": 8.85873538011696e-06,
+      "loss": 1.0894,
+      "step": 12193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8174729347229004,
+      "learning_rate": 8.857821637426902e-06,
+      "loss": 1.7035,
+      "step": 12194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.311976194381714,
+      "learning_rate": 8.856907894736842e-06,
+      "loss": 1.3041,
+      "step": 12195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5588433742523193,
+      "learning_rate": 8.855994152046785e-06,
+      "loss": 1.4432,
+      "step": 12196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1596779823303223,
+      "learning_rate": 8.855080409356725e-06,
+      "loss": 1.3327,
+      "step": 12197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.839106798171997,
+      "learning_rate": 8.854166666666667e-06,
+      "loss": 1.202,
+      "step": 12198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1673102378845215,
+      "learning_rate": 8.853252923976608e-06,
+      "loss": 1.1761,
+      "step": 12199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.930577039718628,
+      "learning_rate": 8.85233918128655e-06,
+      "loss": 1.8598,
+      "step": 12200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4874141216278076,
+      "learning_rate": 8.851425438596492e-06,
+      "loss": 1.6248,
+      "step": 12201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4551970958709717,
+      "learning_rate": 8.850511695906433e-06,
+      "loss": 1.4893,
+      "step": 12202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9499611854553223,
+      "learning_rate": 8.849597953216375e-06,
+      "loss": 1.7275,
+      "step": 12203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5025382041931152,
+      "learning_rate": 8.848684210526316e-06,
+      "loss": 1.4384,
+      "step": 12204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.78574538230896,
+      "learning_rate": 8.847770467836258e-06,
+      "loss": 1.5281,
+      "step": 12205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.487825870513916,
+      "learning_rate": 8.8468567251462e-06,
+      "loss": 1.6811,
+      "step": 12206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6405763626098633,
+      "learning_rate": 8.845942982456142e-06,
+      "loss": 1.4528,
+      "step": 12207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.679586887359619,
+      "learning_rate": 8.845029239766083e-06,
+      "loss": 1.5058,
+      "step": 12208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.357792854309082,
+      "learning_rate": 8.844115497076025e-06,
+      "loss": 1.3862,
+      "step": 12209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.297072649002075,
+      "learning_rate": 8.843201754385965e-06,
+      "loss": 1.2618,
+      "step": 12210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3188650608062744,
+      "learning_rate": 8.842288011695908e-06,
+      "loss": 1.3954,
+      "step": 12211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.625236749649048,
+      "learning_rate": 8.841374269005848e-06,
+      "loss": 1.7912,
+      "step": 12212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4914603233337402,
+      "learning_rate": 8.84046052631579e-06,
+      "loss": 1.4468,
+      "step": 12213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.507124423980713,
+      "learning_rate": 8.839546783625733e-06,
+      "loss": 1.403,
+      "step": 12214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5273449420928955,
+      "learning_rate": 8.838633040935673e-06,
+      "loss": 2.2412,
+      "step": 12215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.319700002670288,
+      "learning_rate": 8.837719298245615e-06,
+      "loss": 1.3039,
+      "step": 12216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3319802284240723,
+      "learning_rate": 8.836805555555556e-06,
+      "loss": 1.4542,
+      "step": 12217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.343318223953247,
+      "learning_rate": 8.835891812865498e-06,
+      "loss": 1.4061,
+      "step": 12218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.924866199493408,
+      "learning_rate": 8.834978070175439e-06,
+      "loss": 2.129,
+      "step": 12219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.91090989112854,
+      "learning_rate": 8.834064327485381e-06,
+      "loss": 1.5013,
+      "step": 12220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.149465560913086,
+      "learning_rate": 8.833150584795323e-06,
+      "loss": 1.5758,
+      "step": 12221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5529754161834717,
+      "learning_rate": 8.832236842105264e-06,
+      "loss": 1.5363,
+      "step": 12222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0553064346313477,
+      "learning_rate": 8.831323099415206e-06,
+      "loss": 2.0357,
+      "step": 12223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8284401893615723,
+      "learning_rate": 8.830409356725146e-06,
+      "loss": 1.2077,
+      "step": 12224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.893237829208374,
+      "learning_rate": 8.829495614035089e-06,
+      "loss": 1.6131,
+      "step": 12225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4045519828796387,
+      "learning_rate": 8.82858187134503e-06,
+      "loss": 1.3166,
+      "step": 12226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3599936962127686,
+      "learning_rate": 8.827668128654971e-06,
+      "loss": 1.7918,
+      "step": 12227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3366031646728516,
+      "learning_rate": 8.826754385964914e-06,
+      "loss": 1.4546,
+      "step": 12228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2449746131896973,
+      "learning_rate": 8.825840643274854e-06,
+      "loss": 1.3618,
+      "step": 12229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3728713989257812,
+      "learning_rate": 8.824926900584796e-06,
+      "loss": 1.4382,
+      "step": 12230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6851654052734375,
+      "learning_rate": 8.824013157894737e-06,
+      "loss": 1.2109,
+      "step": 12231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.54148006439209,
+      "learning_rate": 8.823099415204679e-06,
+      "loss": 1.3562,
+      "step": 12232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.546332597732544,
+      "learning_rate": 8.82218567251462e-06,
+      "loss": 1.4296,
+      "step": 12233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4257655143737793,
+      "learning_rate": 8.821271929824562e-06,
+      "loss": 1.4099,
+      "step": 12234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.71799635887146,
+      "learning_rate": 8.820358187134504e-06,
+      "loss": 1.4905,
+      "step": 12235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.355051279067993,
+      "learning_rate": 8.819444444444445e-06,
+      "loss": 1.5392,
+      "step": 12236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9493212699890137,
+      "learning_rate": 8.818530701754387e-06,
+      "loss": 1.6757,
+      "step": 12237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0510809421539307,
+      "learning_rate": 8.817616959064327e-06,
+      "loss": 1.7108,
+      "step": 12238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.444295644760132,
+      "learning_rate": 8.81670321637427e-06,
+      "loss": 1.4253,
+      "step": 12239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8913888931274414,
+      "learning_rate": 8.81578947368421e-06,
+      "loss": 1.784,
+      "step": 12240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6152493953704834,
+      "learning_rate": 8.814875730994152e-06,
+      "loss": 1.6997,
+      "step": 12241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.07603120803833,
+      "learning_rate": 8.813961988304095e-06,
+      "loss": 1.753,
+      "step": 12242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5754811763763428,
+      "learning_rate": 8.813048245614035e-06,
+      "loss": 1.3964,
+      "step": 12243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.250760793685913,
+      "learning_rate": 8.812134502923977e-06,
+      "loss": 1.3954,
+      "step": 12244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.236705780029297,
+      "learning_rate": 8.811220760233918e-06,
+      "loss": 1.2451,
+      "step": 12245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5646262168884277,
+      "learning_rate": 8.81030701754386e-06,
+      "loss": 1.3671,
+      "step": 12246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8979475498199463,
+      "learning_rate": 8.8093932748538e-06,
+      "loss": 1.5297,
+      "step": 12247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.77170991897583,
+      "learning_rate": 8.808479532163744e-06,
+      "loss": 1.5532,
+      "step": 12248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.882380485534668,
+      "learning_rate": 8.807565789473685e-06,
+      "loss": 1.2525,
+      "step": 12249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5400147438049316,
+      "learning_rate": 8.806652046783627e-06,
+      "loss": 1.3728,
+      "step": 12250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6465280055999756,
+      "learning_rate": 8.805738304093568e-06,
+      "loss": 1.5099,
+      "step": 12251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3048923015594482,
+      "learning_rate": 8.80482456140351e-06,
+      "loss": 1.365,
+      "step": 12252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.439491033554077,
+      "learning_rate": 8.803910818713452e-06,
+      "loss": 1.4681,
+      "step": 12253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5029754638671875,
+      "learning_rate": 8.802997076023393e-06,
+      "loss": 1.4695,
+      "step": 12254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3357596397399902,
+      "learning_rate": 8.802083333333335e-06,
+      "loss": 1.6129,
+      "step": 12255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.126119613647461,
+      "learning_rate": 8.801169590643275e-06,
+      "loss": 1.0872,
+      "step": 12256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.800506591796875,
+      "learning_rate": 8.800255847953218e-06,
+      "loss": 1.804,
+      "step": 12257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6282460689544678,
+      "learning_rate": 8.799342105263158e-06,
+      "loss": 1.2692,
+      "step": 12258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0546646118164062,
+      "learning_rate": 8.7984283625731e-06,
+      "loss": 1.6996,
+      "step": 12259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.192852020263672,
+      "learning_rate": 8.797514619883043e-06,
+      "loss": 1.1737,
+      "step": 12260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.775444984436035,
+      "learning_rate": 8.796600877192983e-06,
+      "loss": 1.7759,
+      "step": 12261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.557229995727539,
+      "learning_rate": 8.795687134502925e-06,
+      "loss": 1.4846,
+      "step": 12262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.736785411834717,
+      "learning_rate": 8.794773391812866e-06,
+      "loss": 1.685,
+      "step": 12263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.279318332672119,
+      "learning_rate": 8.793859649122808e-06,
+      "loss": 1.4562,
+      "step": 12264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6708457469940186,
+      "learning_rate": 8.792945906432749e-06,
+      "loss": 1.5181,
+      "step": 12265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.787303924560547,
+      "learning_rate": 8.792032163742691e-06,
+      "loss": 1.4662,
+      "step": 12266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1531734466552734,
+      "learning_rate": 8.791118421052633e-06,
+      "loss": 1.123,
+      "step": 12267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.656184673309326,
+      "learning_rate": 8.790204678362574e-06,
+      "loss": 1.5173,
+      "step": 12268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4496781826019287,
+      "learning_rate": 8.789290935672516e-06,
+      "loss": 1.4263,
+      "step": 12269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.493692636489868,
+      "learning_rate": 8.788377192982456e-06,
+      "loss": 1.3071,
+      "step": 12270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5575084686279297,
+      "learning_rate": 8.787463450292399e-06,
+      "loss": 1.2786,
+      "step": 12271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.787281036376953,
+      "learning_rate": 8.78654970760234e-06,
+      "loss": 1.618,
+      "step": 12272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.534370183944702,
+      "learning_rate": 8.785635964912281e-06,
+      "loss": 1.4151,
+      "step": 12273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6699280738830566,
+      "learning_rate": 8.784722222222224e-06,
+      "loss": 1.5796,
+      "step": 12274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8174846172332764,
+      "learning_rate": 8.783808479532164e-06,
+      "loss": 1.5347,
+      "step": 12275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.62172269821167,
+      "learning_rate": 8.782894736842106e-06,
+      "loss": 1.3024,
+      "step": 12276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9857102632522583,
+      "learning_rate": 8.781980994152047e-06,
+      "loss": 1.303,
+      "step": 12277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5356979370117188,
+      "learning_rate": 8.781067251461989e-06,
+      "loss": 1.3098,
+      "step": 12278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0365138053894043,
+      "learning_rate": 8.78015350877193e-06,
+      "loss": 1.1335,
+      "step": 12279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4671435356140137,
+      "learning_rate": 8.779239766081872e-06,
+      "loss": 1.4992,
+      "step": 12280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.854781150817871,
+      "learning_rate": 8.778326023391814e-06,
+      "loss": 1.7645,
+      "step": 12281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3937385082244873,
+      "learning_rate": 8.777412280701755e-06,
+      "loss": 1.2935,
+      "step": 12282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.471308708190918,
+      "learning_rate": 8.776498538011697e-06,
+      "loss": 1.5338,
+      "step": 12283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0174343585968018,
+      "learning_rate": 8.775584795321637e-06,
+      "loss": 1.4712,
+      "step": 12284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.329843759536743,
+      "learning_rate": 8.77467105263158e-06,
+      "loss": 1.581,
+      "step": 12285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9338173866271973,
+      "learning_rate": 8.77375730994152e-06,
+      "loss": 1.5487,
+      "step": 12286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.065725326538086,
+      "learning_rate": 8.772843567251462e-06,
+      "loss": 1.0814,
+      "step": 12287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4417314529418945,
+      "learning_rate": 8.771929824561405e-06,
+      "loss": 1.5688,
+      "step": 12288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5971357822418213,
+      "learning_rate": 8.771016081871345e-06,
+      "loss": 1.3391,
+      "step": 12289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3988516330718994,
+      "learning_rate": 8.770102339181287e-06,
+      "loss": 1.416,
+      "step": 12290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.232395648956299,
+      "learning_rate": 8.76918859649123e-06,
+      "loss": 1.8363,
+      "step": 12291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4617919921875,
+      "learning_rate": 8.76827485380117e-06,
+      "loss": 2.1328,
+      "step": 12292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2419991493225098,
+      "learning_rate": 8.767361111111112e-06,
+      "loss": 1.436,
+      "step": 12293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7740273475646973,
+      "learning_rate": 8.766447368421054e-06,
+      "loss": 1.857,
+      "step": 12294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5933451652526855,
+      "learning_rate": 8.765533625730995e-06,
+      "loss": 1.3205,
+      "step": 12295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8158047199249268,
+      "learning_rate": 8.764619883040937e-06,
+      "loss": 1.6093,
+      "step": 12296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1543190479278564,
+      "learning_rate": 8.763706140350878e-06,
+      "loss": 1.2333,
+      "step": 12297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.942471742630005,
+      "learning_rate": 8.76279239766082e-06,
+      "loss": 1.8349,
+      "step": 12298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4409611225128174,
+      "learning_rate": 8.76187865497076e-06,
+      "loss": 1.2002,
+      "step": 12299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.822856903076172,
+      "learning_rate": 8.760964912280703e-06,
+      "loss": 1.5368,
+      "step": 12300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3976759910583496,
+      "learning_rate": 8.760051169590645e-06,
+      "loss": 1.2677,
+      "step": 12301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.645972967147827,
+      "learning_rate": 8.759137426900585e-06,
+      "loss": 1.6261,
+      "step": 12302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5386765003204346,
+      "learning_rate": 8.758223684210528e-06,
+      "loss": 1.4689,
+      "step": 12303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7315526008605957,
+      "learning_rate": 8.757309941520468e-06,
+      "loss": 1.5452,
+      "step": 12304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.689030885696411,
+      "learning_rate": 8.75639619883041e-06,
+      "loss": 1.6689,
+      "step": 12305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8438034057617188,
+      "learning_rate": 8.755482456140351e-06,
+      "loss": 1.6051,
+      "step": 12306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6234562397003174,
+      "learning_rate": 8.754568713450293e-06,
+      "loss": 1.9246,
+      "step": 12307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0802175998687744,
+      "learning_rate": 8.753654970760235e-06,
+      "loss": 1.2332,
+      "step": 12308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3275182247161865,
+      "learning_rate": 8.752741228070176e-06,
+      "loss": 1.4391,
+      "step": 12309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5213277339935303,
+      "learning_rate": 8.751827485380118e-06,
+      "loss": 1.6407,
+      "step": 12310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.440654754638672,
+      "learning_rate": 8.750913742690059e-06,
+      "loss": 1.1311,
+      "step": 12311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2390518188476562,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.9346,
+      "step": 12312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6964213848114014,
+      "learning_rate": 8.749086257309941e-06,
+      "loss": 1.867,
+      "step": 12313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6017372608184814,
+      "learning_rate": 8.748172514619884e-06,
+      "loss": 1.5452,
+      "step": 12314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.291807174682617,
+      "learning_rate": 8.747258771929826e-06,
+      "loss": 1.7032,
+      "step": 12315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.567901372909546,
+      "learning_rate": 8.746345029239766e-06,
+      "loss": 1.4921,
+      "step": 12316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6276419162750244,
+      "learning_rate": 8.745431286549709e-06,
+      "loss": 1.7424,
+      "step": 12317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5318922996520996,
+      "learning_rate": 8.74451754385965e-06,
+      "loss": 1.3527,
+      "step": 12318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4442572593688965,
+      "learning_rate": 8.743603801169591e-06,
+      "loss": 1.546,
+      "step": 12319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.703460454940796,
+      "learning_rate": 8.742690058479532e-06,
+      "loss": 1.5612,
+      "step": 12320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4804699420928955,
+      "learning_rate": 8.741776315789474e-06,
+      "loss": 1.5432,
+      "step": 12321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.806447744369507,
+      "learning_rate": 8.740862573099416e-06,
+      "loss": 1.4052,
+      "step": 12322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0680086612701416,
+      "learning_rate": 8.739948830409357e-06,
+      "loss": 1.4514,
+      "step": 12323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.572756767272949,
+      "learning_rate": 8.739035087719299e-06,
+      "loss": 1.6329,
+      "step": 12324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9395663738250732,
+      "learning_rate": 8.73812134502924e-06,
+      "loss": 1.9445,
+      "step": 12325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1762943267822266,
+      "learning_rate": 8.737207602339182e-06,
+      "loss": 1.2787,
+      "step": 12326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6248559951782227,
+      "learning_rate": 8.736293859649122e-06,
+      "loss": 1.3337,
+      "step": 12327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1750495433807373,
+      "learning_rate": 8.735380116959065e-06,
+      "loss": 1.2942,
+      "step": 12328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4081897735595703,
+      "learning_rate": 8.734466374269007e-06,
+      "loss": 1.6243,
+      "step": 12329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7511684894561768,
+      "learning_rate": 8.733552631578947e-06,
+      "loss": 1.8289,
+      "step": 12330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.256908893585205,
+      "learning_rate": 8.73263888888889e-06,
+      "loss": 1.7762,
+      "step": 12331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5582175254821777,
+      "learning_rate": 8.73172514619883e-06,
+      "loss": 1.3982,
+      "step": 12332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5144526958465576,
+      "learning_rate": 8.730811403508772e-06,
+      "loss": 1.3686,
+      "step": 12333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.245464324951172,
+      "learning_rate": 8.729897660818715e-06,
+      "loss": 1.9057,
+      "step": 12334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7416675090789795,
+      "learning_rate": 8.728983918128657e-06,
+      "loss": 1.583,
+      "step": 12335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4452261924743652,
+      "learning_rate": 8.728070175438597e-06,
+      "loss": 2.2131,
+      "step": 12336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.439483165740967,
+      "learning_rate": 8.72715643274854e-06,
+      "loss": 1.5171,
+      "step": 12337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9085023403167725,
+      "learning_rate": 8.72624269005848e-06,
+      "loss": 1.6892,
+      "step": 12338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.56309175491333,
+      "learning_rate": 8.725328947368422e-06,
+      "loss": 1.4774,
+      "step": 12339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8772854804992676,
+      "learning_rate": 8.724415204678364e-06,
+      "loss": 1.7454,
+      "step": 12340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6773252487182617,
+      "learning_rate": 8.723501461988305e-06,
+      "loss": 1.4285,
+      "step": 12341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.416989326477051,
+      "learning_rate": 8.722587719298247e-06,
+      "loss": 2.3465,
+      "step": 12342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9643611907958984,
+      "learning_rate": 8.721673976608188e-06,
+      "loss": 1.1027,
+      "step": 12343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.469930410385132,
+      "learning_rate": 8.72076023391813e-06,
+      "loss": 1.5108,
+      "step": 12344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7036943435668945,
+      "learning_rate": 8.71984649122807e-06,
+      "loss": 1.7526,
+      "step": 12345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.382654905319214,
+      "learning_rate": 8.718932748538013e-06,
+      "loss": 1.3687,
+      "step": 12346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4651622772216797,
+      "learning_rate": 8.718019005847955e-06,
+      "loss": 1.0101,
+      "step": 12347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0135271549224854,
+      "learning_rate": 8.717105263157895e-06,
+      "loss": 1.5247,
+      "step": 12348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.744100570678711,
+      "learning_rate": 8.716191520467838e-06,
+      "loss": 1.6104,
+      "step": 12349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.489108085632324,
+      "learning_rate": 8.715277777777778e-06,
+      "loss": 1.3469,
+      "step": 12350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.994062662124634,
+      "learning_rate": 8.71436403508772e-06,
+      "loss": 1.7095,
+      "step": 12351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.721712112426758,
+      "learning_rate": 8.713450292397661e-06,
+      "loss": 1.7758,
+      "step": 12352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8408710956573486,
+      "learning_rate": 8.712536549707603e-06,
+      "loss": 1.3604,
+      "step": 12353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.049006223678589,
+      "learning_rate": 8.711622807017545e-06,
+      "loss": 1.0982,
+      "step": 12354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.09920334815979,
+      "learning_rate": 8.710709064327486e-06,
+      "loss": 1.2053,
+      "step": 12355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.622419357299805,
+      "learning_rate": 8.709795321637428e-06,
+      "loss": 1.6173,
+      "step": 12356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7603111267089844,
+      "learning_rate": 8.708881578947369e-06,
+      "loss": 1.5705,
+      "step": 12357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5731260776519775,
+      "learning_rate": 8.707967836257311e-06,
+      "loss": 1.4478,
+      "step": 12358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5065109729766846,
+      "learning_rate": 8.707054093567251e-06,
+      "loss": 1.2841,
+      "step": 12359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6906650066375732,
+      "learning_rate": 8.706140350877194e-06,
+      "loss": 1.3626,
+      "step": 12360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7399473190307617,
+      "learning_rate": 8.705226608187136e-06,
+      "loss": 1.4681,
+      "step": 12361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4246459007263184,
+      "learning_rate": 8.704312865497076e-06,
+      "loss": 1.3487,
+      "step": 12362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0118212699890137,
+      "learning_rate": 8.703399122807019e-06,
+      "loss": 1.3705,
+      "step": 12363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1939899921417236,
+      "learning_rate": 8.70248538011696e-06,
+      "loss": 1.691,
+      "step": 12364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4222896099090576,
+      "learning_rate": 8.701571637426901e-06,
+      "loss": 1.5632,
+      "step": 12365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8308706283569336,
+      "learning_rate": 8.700657894736842e-06,
+      "loss": 1.5702,
+      "step": 12366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4431121349334717,
+      "learning_rate": 8.699744152046784e-06,
+      "loss": 1.3731,
+      "step": 12367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8078935146331787,
+      "learning_rate": 8.698830409356725e-06,
+      "loss": 1.4308,
+      "step": 12368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6994543075561523,
+      "learning_rate": 8.697916666666667e-06,
+      "loss": 1.4229,
+      "step": 12369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6757967472076416,
+      "learning_rate": 8.697002923976609e-06,
+      "loss": 1.6892,
+      "step": 12370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9260239601135254,
+      "learning_rate": 8.69608918128655e-06,
+      "loss": 1.6882,
+      "step": 12371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.28196120262146,
+      "learning_rate": 8.695175438596492e-06,
+      "loss": 1.413,
+      "step": 12372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5444326400756836,
+      "learning_rate": 8.694261695906432e-06,
+      "loss": 1.3691,
+      "step": 12373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1153452396392822,
+      "learning_rate": 8.693347953216375e-06,
+      "loss": 1.0832,
+      "step": 12374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8678219318389893,
+      "learning_rate": 8.692434210526315e-06,
+      "loss": 1.404,
+      "step": 12375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6466751098632812,
+      "learning_rate": 8.691520467836257e-06,
+      "loss": 1.6514,
+      "step": 12376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.940406084060669,
+      "learning_rate": 8.6906067251462e-06,
+      "loss": 1.904,
+      "step": 12377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2254912853240967,
+      "learning_rate": 8.689692982456142e-06,
+      "loss": 1.3513,
+      "step": 12378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.077760934829712,
+      "learning_rate": 8.688779239766082e-06,
+      "loss": 1.28,
+      "step": 12379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.728997230529785,
+      "learning_rate": 8.687865497076025e-06,
+      "loss": 1.643,
+      "step": 12380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.403489828109741,
+      "learning_rate": 8.686951754385967e-06,
+      "loss": 1.5391,
+      "step": 12381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7990620136260986,
+      "learning_rate": 8.686038011695907e-06,
+      "loss": 1.5818,
+      "step": 12382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.741762399673462,
+      "learning_rate": 8.68512426900585e-06,
+      "loss": 1.6349,
+      "step": 12383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0817105770111084,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 1.911,
+      "step": 12384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7056705951690674,
+      "learning_rate": 8.683296783625732e-06,
+      "loss": 1.6077,
+      "step": 12385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.035677671432495,
+      "learning_rate": 8.682383040935673e-06,
+      "loss": 1.6306,
+      "step": 12386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.659574270248413,
+      "learning_rate": 8.681469298245615e-06,
+      "loss": 1.4682,
+      "step": 12387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.452401638031006,
+      "learning_rate": 8.680555555555557e-06,
+      "loss": 1.3258,
+      "step": 12388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.042541980743408,
+      "learning_rate": 8.679641812865498e-06,
+      "loss": 1.8602,
+      "step": 12389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1672117710113525,
+      "learning_rate": 8.67872807017544e-06,
+      "loss": 1.4441,
+      "step": 12390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.319366455078125,
+      "learning_rate": 8.67781432748538e-06,
+      "loss": 1.5751,
+      "step": 12391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.023552656173706,
+      "learning_rate": 8.676900584795323e-06,
+      "loss": 1.7947,
+      "step": 12392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.782874345779419,
+      "learning_rate": 8.675986842105263e-06,
+      "loss": 1.5395,
+      "step": 12393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6774120330810547,
+      "learning_rate": 8.675073099415205e-06,
+      "loss": 1.65,
+      "step": 12394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9859615564346313,
+      "learning_rate": 8.674159356725148e-06,
+      "loss": 1.0867,
+      "step": 12395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.543417453765869,
+      "learning_rate": 8.673245614035088e-06,
+      "loss": 1.2579,
+      "step": 12396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5476667881011963,
+      "learning_rate": 8.67233187134503e-06,
+      "loss": 1.413,
+      "step": 12397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.593116044998169,
+      "learning_rate": 8.671418128654971e-06,
+      "loss": 1.4728,
+      "step": 12398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.088740110397339,
+      "learning_rate": 8.670504385964913e-06,
+      "loss": 1.5387,
+      "step": 12399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.303898811340332,
+      "learning_rate": 8.669590643274854e-06,
+      "loss": 1.2649,
+      "step": 12400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.285214900970459,
+      "learning_rate": 8.668676900584796e-06,
+      "loss": 1.59,
+      "step": 12401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.410156488418579,
+      "learning_rate": 8.667763157894738e-06,
+      "loss": 1.4645,
+      "step": 12402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.409168004989624,
+      "learning_rate": 8.666849415204679e-06,
+      "loss": 1.3799,
+      "step": 12403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8540890216827393,
+      "learning_rate": 8.665935672514621e-06,
+      "loss": 1.3496,
+      "step": 12404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.436034917831421,
+      "learning_rate": 8.665021929824561e-06,
+      "loss": 1.2146,
+      "step": 12405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.676832914352417,
+      "learning_rate": 8.664108187134504e-06,
+      "loss": 1.4953,
+      "step": 12406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5106749534606934,
+      "learning_rate": 8.663194444444444e-06,
+      "loss": 1.6064,
+      "step": 12407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.801609754562378,
+      "learning_rate": 8.662280701754386e-06,
+      "loss": 1.9504,
+      "step": 12408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0780274868011475,
+      "learning_rate": 8.661366959064329e-06,
+      "loss": 1.7695,
+      "step": 12409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9643058776855469,
+      "learning_rate": 8.66045321637427e-06,
+      "loss": 1.0952,
+      "step": 12410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6026291847229004,
+      "learning_rate": 8.659539473684211e-06,
+      "loss": 1.455,
+      "step": 12411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.431001663208008,
+      "learning_rate": 8.658625730994152e-06,
+      "loss": 1.5332,
+      "step": 12412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.354300022125244,
+      "learning_rate": 8.657711988304094e-06,
+      "loss": 1.3068,
+      "step": 12413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.830735445022583,
+      "learning_rate": 8.656798245614035e-06,
+      "loss": 1.7889,
+      "step": 12414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1566834449768066,
+      "learning_rate": 8.655884502923977e-06,
+      "loss": 1.3084,
+      "step": 12415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.659571647644043,
+      "learning_rate": 8.654970760233919e-06,
+      "loss": 1.5636,
+      "step": 12416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9600203037261963,
+      "learning_rate": 8.65405701754386e-06,
+      "loss": 1.6569,
+      "step": 12417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5163800716400146,
+      "learning_rate": 8.653143274853802e-06,
+      "loss": 1.8174,
+      "step": 12418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.631455183029175,
+      "learning_rate": 8.652229532163744e-06,
+      "loss": 1.4004,
+      "step": 12419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.934478282928467,
+      "learning_rate": 8.651315789473685e-06,
+      "loss": 1.5295,
+      "step": 12420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2237210273742676,
+      "learning_rate": 8.650402046783627e-06,
+      "loss": 1.9414,
+      "step": 12421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7695446014404297,
+      "learning_rate": 8.649488304093569e-06,
+      "loss": 1.6191,
+      "step": 12422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.692286968231201,
+      "learning_rate": 8.64857456140351e-06,
+      "loss": 1.5271,
+      "step": 12423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.391958713531494,
+      "learning_rate": 8.647660818713452e-06,
+      "loss": 1.3408,
+      "step": 12424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.110450267791748,
+      "learning_rate": 8.646747076023392e-06,
+      "loss": 1.2491,
+      "step": 12425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7112817764282227,
+      "learning_rate": 8.645833333333335e-06,
+      "loss": 1.4388,
+      "step": 12426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.310926914215088,
+      "learning_rate": 8.644919590643275e-06,
+      "loss": 1.845,
+      "step": 12427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.316188335418701,
+      "learning_rate": 8.644005847953217e-06,
+      "loss": 1.5498,
+      "step": 12428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.07753586769104,
+      "learning_rate": 8.64309210526316e-06,
+      "loss": 1.3797,
+      "step": 12429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.318432092666626,
+      "learning_rate": 8.6421783625731e-06,
+      "loss": 1.4253,
+      "step": 12430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.340914011001587,
+      "learning_rate": 8.641264619883042e-06,
+      "loss": 1.2406,
+      "step": 12431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7553133964538574,
+      "learning_rate": 8.640350877192983e-06,
+      "loss": 1.6078,
+      "step": 12432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1989171504974365,
+      "learning_rate": 8.639437134502925e-06,
+      "loss": 1.6767,
+      "step": 12433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.424077272415161,
+      "learning_rate": 8.638523391812866e-06,
+      "loss": 1.2351,
+      "step": 12434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5909016132354736,
+      "learning_rate": 8.637609649122808e-06,
+      "loss": 1.5541,
+      "step": 12435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9557578563690186,
+      "learning_rate": 8.63669590643275e-06,
+      "loss": 1.9463,
+      "step": 12436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3523647785186768,
+      "learning_rate": 8.63578216374269e-06,
+      "loss": 1.3274,
+      "step": 12437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7855045795440674,
+      "learning_rate": 8.634868421052633e-06,
+      "loss": 1.3848,
+      "step": 12438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.199399709701538,
+      "learning_rate": 8.633954678362573e-06,
+      "loss": 1.817,
+      "step": 12439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3920516967773438,
+      "learning_rate": 8.633040935672515e-06,
+      "loss": 1.4797,
+      "step": 12440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3090600967407227,
+      "learning_rate": 8.632127192982456e-06,
+      "loss": 1.3333,
+      "step": 12441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.145622968673706,
+      "learning_rate": 8.631213450292398e-06,
+      "loss": 1.7776,
+      "step": 12442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0040338039398193,
+      "learning_rate": 8.63029970760234e-06,
+      "loss": 1.5621,
+      "step": 12443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.380402088165283,
+      "learning_rate": 8.629385964912281e-06,
+      "loss": 1.9303,
+      "step": 12444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.824929714202881,
+      "learning_rate": 8.628472222222223e-06,
+      "loss": 1.505,
+      "step": 12445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.669670581817627,
+      "learning_rate": 8.627558479532164e-06,
+      "loss": 1.589,
+      "step": 12446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.726245641708374,
+      "learning_rate": 8.626644736842106e-06,
+      "loss": 1.6179,
+      "step": 12447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0048115253448486,
+      "learning_rate": 8.625730994152046e-06,
+      "loss": 1.1742,
+      "step": 12448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.245424509048462,
+      "learning_rate": 8.624817251461989e-06,
+      "loss": 1.3265,
+      "step": 12449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.837048053741455,
+      "learning_rate": 8.623903508771931e-06,
+      "loss": 1.4105,
+      "step": 12450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7544043064117432,
+      "learning_rate": 8.622989766081871e-06,
+      "loss": 0.9753,
+      "step": 12451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.653259038925171,
+      "learning_rate": 8.622076023391814e-06,
+      "loss": 1.4513,
+      "step": 12452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8850512504577637,
+      "learning_rate": 8.621162280701754e-06,
+      "loss": 1.4147,
+      "step": 12453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0408108234405518,
+      "learning_rate": 8.620248538011696e-06,
+      "loss": 1.5364,
+      "step": 12454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1405630111694336,
+      "learning_rate": 8.619334795321637e-06,
+      "loss": 1.2493,
+      "step": 12455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3081154823303223,
+      "learning_rate": 8.61842105263158e-06,
+      "loss": 1.3837,
+      "step": 12456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2718029022216797,
+      "learning_rate": 8.617507309941521e-06,
+      "loss": 1.2426,
+      "step": 12457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5468873977661133,
+      "learning_rate": 8.616593567251462e-06,
+      "loss": 1.502,
+      "step": 12458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5265274047851562,
+      "learning_rate": 8.615679824561404e-06,
+      "loss": 1.3586,
+      "step": 12459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.986088514328003,
+      "learning_rate": 8.614766081871345e-06,
+      "loss": 1.6178,
+      "step": 12460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.685009717941284,
+      "learning_rate": 8.613852339181287e-06,
+      "loss": 1.5402,
+      "step": 12461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.273677110671997,
+      "learning_rate": 8.612938596491229e-06,
+      "loss": 1.379,
+      "step": 12462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9641523361206055,
+      "learning_rate": 8.612024853801171e-06,
+      "loss": 2.1583,
+      "step": 12463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7644600868225098,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.2594,
+      "step": 12464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.56672739982605,
+      "learning_rate": 8.610197368421054e-06,
+      "loss": 1.4779,
+      "step": 12465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.321025848388672,
+      "learning_rate": 8.609283625730995e-06,
+      "loss": 1.2114,
+      "step": 12466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.951517105102539,
+      "learning_rate": 8.608369883040937e-06,
+      "loss": 1.2739,
+      "step": 12467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.656473398208618,
+      "learning_rate": 8.607456140350879e-06,
+      "loss": 1.4329,
+      "step": 12468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.718898296356201,
+      "learning_rate": 8.60654239766082e-06,
+      "loss": 1.6665,
+      "step": 12469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.676056146621704,
+      "learning_rate": 8.605628654970762e-06,
+      "loss": 1.5027,
+      "step": 12470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7016546726226807,
+      "learning_rate": 8.604714912280702e-06,
+      "loss": 1.8109,
+      "step": 12471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.252039909362793,
+      "learning_rate": 8.603801169590645e-06,
+      "loss": 1.3397,
+      "step": 12472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.505906343460083,
+      "learning_rate": 8.602887426900585e-06,
+      "loss": 1.4437,
+      "step": 12473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2392899990081787,
+      "learning_rate": 8.601973684210527e-06,
+      "loss": 1.2342,
+      "step": 12474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2595136165618896,
+      "learning_rate": 8.60105994152047e-06,
+      "loss": 1.2175,
+      "step": 12475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2623934745788574,
+      "learning_rate": 8.60014619883041e-06,
+      "loss": 1.2174,
+      "step": 12476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.353864908218384,
+      "learning_rate": 8.599232456140352e-06,
+      "loss": 1.4042,
+      "step": 12477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.894634485244751,
+      "learning_rate": 8.598318713450293e-06,
+      "loss": 1.3934,
+      "step": 12478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.283942222595215,
+      "learning_rate": 8.597404970760235e-06,
+      "loss": 1.1677,
+      "step": 12479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.55483341217041,
+      "learning_rate": 8.596491228070176e-06,
+      "loss": 1.598,
+      "step": 12480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6285035610198975,
+      "learning_rate": 8.595577485380118e-06,
+      "loss": 1.4038,
+      "step": 12481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.900261402130127,
+      "learning_rate": 8.59466374269006e-06,
+      "loss": 1.7632,
+      "step": 12482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5997283458709717,
+      "learning_rate": 8.59375e-06,
+      "loss": 1.6465,
+      "step": 12483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4812495708465576,
+      "learning_rate": 8.592836257309943e-06,
+      "loss": 1.3844,
+      "step": 12484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.075695276260376,
+      "learning_rate": 8.591922514619883e-06,
+      "loss": 1.823,
+      "step": 12485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9496161937713623,
+      "learning_rate": 8.591008771929825e-06,
+      "loss": 1.7527,
+      "step": 12486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4570701122283936,
+      "learning_rate": 8.590095029239766e-06,
+      "loss": 1.4459,
+      "step": 12487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.907656669616699,
+      "learning_rate": 8.589181286549708e-06,
+      "loss": 1.5122,
+      "step": 12488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.861804962158203,
+      "learning_rate": 8.58826754385965e-06,
+      "loss": 1.6647,
+      "step": 12489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3789517879486084,
+      "learning_rate": 8.587353801169591e-06,
+      "loss": 1.4951,
+      "step": 12490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5108962059020996,
+      "learning_rate": 8.586440058479533e-06,
+      "loss": 1.3737,
+      "step": 12491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.052360773086548,
+      "learning_rate": 8.585526315789474e-06,
+      "loss": 1.2558,
+      "step": 12492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.27380633354187,
+      "learning_rate": 8.584612573099416e-06,
+      "loss": 1.1439,
+      "step": 12493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1203668117523193,
+      "learning_rate": 8.583698830409356e-06,
+      "loss": 1.6901,
+      "step": 12494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5315115451812744,
+      "learning_rate": 8.582785087719299e-06,
+      "loss": 1.813,
+      "step": 12495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9184072017669678,
+      "learning_rate": 8.581871345029241e-06,
+      "loss": 1.4878,
+      "step": 12496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4821035861968994,
+      "learning_rate": 8.580957602339181e-06,
+      "loss": 1.3897,
+      "step": 12497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.181248664855957,
+      "learning_rate": 8.580043859649124e-06,
+      "loss": 1.0416,
+      "step": 12498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6006624698638916,
+      "learning_rate": 8.579130116959064e-06,
+      "loss": 1.6115,
+      "step": 12499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.181508779525757,
+      "learning_rate": 8.578216374269006e-06,
+      "loss": 1.3851,
+      "step": 12500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.46199107170105,
+      "learning_rate": 8.577302631578947e-06,
+      "loss": 1.1453,
+      "step": 12501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4530203342437744,
+      "learning_rate": 8.57638888888889e-06,
+      "loss": 1.4385,
+      "step": 12502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5582921504974365,
+      "learning_rate": 8.575475146198831e-06,
+      "loss": 1.4599,
+      "step": 12503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.714414119720459,
+      "learning_rate": 8.574561403508772e-06,
+      "loss": 1.4996,
+      "step": 12504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515726089477539,
+      "learning_rate": 8.573647660818714e-06,
+      "loss": 1.4664,
+      "step": 12505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8770101070404053,
+      "learning_rate": 8.572733918128656e-06,
+      "loss": 1.522,
+      "step": 12506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6866207122802734,
+      "learning_rate": 8.571820175438597e-06,
+      "loss": 1.7224,
+      "step": 12507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1131603717803955,
+      "learning_rate": 8.570906432748539e-06,
+      "loss": 0.9456,
+      "step": 12508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.612299680709839,
+      "learning_rate": 8.569992690058481e-06,
+      "loss": 1.3667,
+      "step": 12509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5816211700439453,
+      "learning_rate": 8.569078947368422e-06,
+      "loss": 1.7221,
+      "step": 12510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.547175407409668,
+      "learning_rate": 8.568165204678364e-06,
+      "loss": 1.1752,
+      "step": 12511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.987863779067993,
+      "learning_rate": 8.567251461988305e-06,
+      "loss": 1.8305,
+      "step": 12512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1147139072418213,
+      "learning_rate": 8.566337719298247e-06,
+      "loss": 1.3245,
+      "step": 12513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.625056266784668,
+      "learning_rate": 8.565423976608187e-06,
+      "loss": 1.5181,
+      "step": 12514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.726731300354004,
+      "learning_rate": 8.56451023391813e-06,
+      "loss": 1.5024,
+      "step": 12515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0504610538482666,
+      "learning_rate": 8.563596491228072e-06,
+      "loss": 1.8468,
+      "step": 12516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0288944244384766,
+      "learning_rate": 8.562682748538012e-06,
+      "loss": 1.323,
+      "step": 12517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6271979808807373,
+      "learning_rate": 8.561769005847955e-06,
+      "loss": 1.8529,
+      "step": 12518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1201815605163574,
+      "learning_rate": 8.560855263157895e-06,
+      "loss": 1.0217,
+      "step": 12519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5308079719543457,
+      "learning_rate": 8.559941520467837e-06,
+      "loss": 1.4027,
+      "step": 12520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.984736442565918,
+      "learning_rate": 8.559027777777778e-06,
+      "loss": 1.0898,
+      "step": 12521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.749114513397217,
+      "learning_rate": 8.55811403508772e-06,
+      "loss": 1.6652,
+      "step": 12522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9229323863983154,
+      "learning_rate": 8.557200292397662e-06,
+      "loss": 1.3418,
+      "step": 12523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5344905853271484,
+      "learning_rate": 8.556286549707603e-06,
+      "loss": 1.2,
+      "step": 12524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3236498832702637,
+      "learning_rate": 8.555372807017545e-06,
+      "loss": 1.1329,
+      "step": 12525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5334486961364746,
+      "learning_rate": 8.554459064327486e-06,
+      "loss": 1.5661,
+      "step": 12526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.27504301071167,
+      "learning_rate": 8.553545321637428e-06,
+      "loss": 1.4221,
+      "step": 12527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7074179649353027,
+      "learning_rate": 8.552631578947368e-06,
+      "loss": 1.3806,
+      "step": 12528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.648242950439453,
+      "learning_rate": 8.55171783625731e-06,
+      "loss": 1.452,
+      "step": 12529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.586939573287964,
+      "learning_rate": 8.550804093567253e-06,
+      "loss": 1.4276,
+      "step": 12530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.162900686264038,
+      "learning_rate": 8.549890350877193e-06,
+      "loss": 1.1822,
+      "step": 12531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2358109951019287,
+      "learning_rate": 8.548976608187135e-06,
+      "loss": 1.1773,
+      "step": 12532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.448153257369995,
+      "learning_rate": 8.548062865497076e-06,
+      "loss": 1.3206,
+      "step": 12533
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.05549955368042,
+      "learning_rate": 8.547149122807018e-06,
+      "loss": 1.0834,
+      "step": 12534
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4605791568756104,
+      "learning_rate": 8.546235380116959e-06,
+      "loss": 1.3203,
+      "step": 12535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1862633228302,
+      "learning_rate": 8.545321637426901e-06,
+      "loss": 1.2351,
+      "step": 12536
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.108531951904297,
+      "learning_rate": 8.544407894736843e-06,
+      "loss": 1.8104,
+      "step": 12537
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3919174671173096,
+      "learning_rate": 8.543494152046784e-06,
+      "loss": 1.2478,
+      "step": 12538
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4842324256896973,
+      "learning_rate": 8.542580409356726e-06,
+      "loss": 1.4735,
+      "step": 12539
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6567885875701904,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 1.5715,
+      "step": 12540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.375133514404297,
+      "learning_rate": 8.540752923976609e-06,
+      "loss": 1.3187,
+      "step": 12541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.487557888031006,
+      "learning_rate": 8.53983918128655e-06,
+      "loss": 1.5061,
+      "step": 12542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6328558921813965,
+      "learning_rate": 8.538925438596491e-06,
+      "loss": 1.1992,
+      "step": 12543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.333164930343628,
+      "learning_rate": 8.538011695906434e-06,
+      "loss": 1.9468,
+      "step": 12544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.275662660598755,
+      "learning_rate": 8.537097953216374e-06,
+      "loss": 1.7074,
+      "step": 12545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.539214611053467,
+      "learning_rate": 8.536184210526316e-06,
+      "loss": 1.4573,
+      "step": 12546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.713500499725342,
+      "learning_rate": 8.535270467836257e-06,
+      "loss": 1.6492,
+      "step": 12547
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8916213512420654,
+      "learning_rate": 8.534356725146201e-06,
+      "loss": 1.794,
+      "step": 12548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.570286273956299,
+      "learning_rate": 8.533442982456141e-06,
+      "loss": 1.4788,
+      "step": 12549
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0390357971191406,
+      "learning_rate": 8.532529239766084e-06,
+      "loss": 1.9781,
+      "step": 12550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4923152923583984,
+      "learning_rate": 8.531615497076024e-06,
+      "loss": 1.3669,
+      "step": 12551
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5328519344329834,
+      "learning_rate": 8.530701754385966e-06,
+      "loss": 1.3797,
+      "step": 12552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.484017848968506,
+      "learning_rate": 8.529788011695907e-06,
+      "loss": 1.2948,
+      "step": 12553
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.696767568588257,
+      "learning_rate": 8.528874269005849e-06,
+      "loss": 1.6203,
+      "step": 12554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7123775482177734,
+      "learning_rate": 8.527960526315791e-06,
+      "loss": 1.759,
+      "step": 12555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2405049800872803,
+      "learning_rate": 8.527046783625732e-06,
+      "loss": 1.103,
+      "step": 12556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2828376293182373,
+      "learning_rate": 8.526133040935674e-06,
+      "loss": 1.3909,
+      "step": 12557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4031689167022705,
+      "learning_rate": 8.525219298245615e-06,
+      "loss": 1.6256,
+      "step": 12558
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.839777946472168,
+      "learning_rate": 8.524305555555557e-06,
+      "loss": 1.2272,
+      "step": 12559
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4792983531951904,
+      "learning_rate": 8.523391812865497e-06,
+      "loss": 1.4507,
+      "step": 12560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.034722328186035,
+      "learning_rate": 8.52247807017544e-06,
+      "loss": 1.8427,
+      "step": 12561
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5678577423095703,
+      "learning_rate": 8.521564327485382e-06,
+      "loss": 1.1762,
+      "step": 12562
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9561309814453125,
+      "learning_rate": 8.520650584795322e-06,
+      "loss": 1.6631,
+      "step": 12563
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3864028453826904,
+      "learning_rate": 8.519736842105265e-06,
+      "loss": 1.3515,
+      "step": 12564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4349920749664307,
+      "learning_rate": 8.518823099415205e-06,
+      "loss": 1.2051,
+      "step": 12565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3375656604766846,
+      "learning_rate": 8.517909356725147e-06,
+      "loss": 1.9869,
+      "step": 12566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5170037746429443,
+      "learning_rate": 8.516995614035088e-06,
+      "loss": 1.1892,
+      "step": 12567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6554605960845947,
+      "learning_rate": 8.51608187134503e-06,
+      "loss": 1.294,
+      "step": 12568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.568723440170288,
+      "learning_rate": 8.515168128654972e-06,
+      "loss": 1.6776,
+      "step": 12569
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.045184850692749,
+      "learning_rate": 8.514254385964913e-06,
+      "loss": 1.7649,
+      "step": 12570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7472984790802,
+      "learning_rate": 8.513340643274855e-06,
+      "loss": 1.6467,
+      "step": 12571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8923747539520264,
+      "learning_rate": 8.512426900584796e-06,
+      "loss": 1.9159,
+      "step": 12572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0704946517944336,
+      "learning_rate": 8.511513157894738e-06,
+      "loss": 1.323,
+      "step": 12573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.28520131111145,
+      "learning_rate": 8.510599415204678e-06,
+      "loss": 1.9258,
+      "step": 12574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.72837233543396,
+      "learning_rate": 8.50968567251462e-06,
+      "loss": 1.6647,
+      "step": 12575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9212629795074463,
+      "learning_rate": 8.508771929824563e-06,
+      "loss": 1.4123,
+      "step": 12576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.496727228164673,
+      "learning_rate": 8.507858187134503e-06,
+      "loss": 1.0822,
+      "step": 12577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1292166709899902,
+      "learning_rate": 8.506944444444445e-06,
+      "loss": 1.2972,
+      "step": 12578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4213168621063232,
+      "learning_rate": 8.506030701754386e-06,
+      "loss": 1.2352,
+      "step": 12579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.804213047027588,
+      "learning_rate": 8.505116959064328e-06,
+      "loss": 1.3497,
+      "step": 12580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5725035667419434,
+      "learning_rate": 8.504203216374269e-06,
+      "loss": 1.3962,
+      "step": 12581
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9307618141174316,
+      "learning_rate": 8.503289473684211e-06,
+      "loss": 1.7826,
+      "step": 12582
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6282286643981934,
+      "learning_rate": 8.502375730994153e-06,
+      "loss": 1.4892,
+      "step": 12583
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2384555339813232,
+      "learning_rate": 8.501461988304094e-06,
+      "loss": 1.4307,
+      "step": 12584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8468871116638184,
+      "learning_rate": 8.500548245614036e-06,
+      "loss": 1.7208,
+      "step": 12585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.502683639526367,
+      "learning_rate": 8.499634502923976e-06,
+      "loss": 1.5946,
+      "step": 12586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9642398357391357,
+      "learning_rate": 8.498720760233919e-06,
+      "loss": 1.9434,
+      "step": 12587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.725154399871826,
+      "learning_rate": 8.49780701754386e-06,
+      "loss": 1.6949,
+      "step": 12588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.929051160812378,
+      "learning_rate": 8.496893274853801e-06,
+      "loss": 1.7794,
+      "step": 12589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.442869186401367,
+      "learning_rate": 8.495979532163744e-06,
+      "loss": 1.5591,
+      "step": 12590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.51586651802063,
+      "learning_rate": 8.495065789473686e-06,
+      "loss": 1.4075,
+      "step": 12591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.234384298324585,
+      "learning_rate": 8.494152046783626e-06,
+      "loss": 1.2142,
+      "step": 12592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.188941240310669,
+      "learning_rate": 8.493238304093569e-06,
+      "loss": 1.1214,
+      "step": 12593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7142174243927,
+      "learning_rate": 8.49232456140351e-06,
+      "loss": 1.4172,
+      "step": 12594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6613565683364868,
+      "learning_rate": 8.491410818713451e-06,
+      "loss": 1.0731,
+      "step": 12595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5984137058258057,
+      "learning_rate": 8.490497076023394e-06,
+      "loss": 1.6696,
+      "step": 12596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5351178646087646,
+      "learning_rate": 8.489583333333334e-06,
+      "loss": 1.4412,
+      "step": 12597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6236143112182617,
+      "learning_rate": 8.488669590643276e-06,
+      "loss": 1.5447,
+      "step": 12598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1079189777374268,
+      "learning_rate": 8.487755847953217e-06,
+      "loss": 2.1866,
+      "step": 12599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3601126670837402,
+      "learning_rate": 8.486842105263159e-06,
+      "loss": 1.2811,
+      "step": 12600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.462846279144287,
+      "learning_rate": 8.4859283625731e-06,
+      "loss": 1.4364,
+      "step": 12601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.682654857635498,
+      "learning_rate": 8.485014619883042e-06,
+      "loss": 1.5529,
+      "step": 12602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.386638879776001,
+      "learning_rate": 8.484100877192984e-06,
+      "loss": 1.2954,
+      "step": 12603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0134947299957275,
+      "learning_rate": 8.483187134502925e-06,
+      "loss": 1.2906,
+      "step": 12604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.73113751411438,
+      "learning_rate": 8.482273391812867e-06,
+      "loss": 1.5923,
+      "step": 12605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.192809820175171,
+      "learning_rate": 8.481359649122807e-06,
+      "loss": 1.4304,
+      "step": 12606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3422114849090576,
+      "learning_rate": 8.48044590643275e-06,
+      "loss": 1.4596,
+      "step": 12607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8301587104797363,
+      "learning_rate": 8.47953216374269e-06,
+      "loss": 1.3514,
+      "step": 12608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9836666584014893,
+      "learning_rate": 8.478618421052632e-06,
+      "loss": 1.6978,
+      "step": 12609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6733789443969727,
+      "learning_rate": 8.477704678362575e-06,
+      "loss": 1.5408,
+      "step": 12610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.362347364425659,
+      "learning_rate": 8.476790935672515e-06,
+      "loss": 1.3594,
+      "step": 12611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.559722661972046,
+      "learning_rate": 8.475877192982457e-06,
+      "loss": 1.254,
+      "step": 12612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5981266498565674,
+      "learning_rate": 8.474963450292398e-06,
+      "loss": 1.2812,
+      "step": 12613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6085968017578125,
+      "learning_rate": 8.47404970760234e-06,
+      "loss": 1.5472,
+      "step": 12614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4886884689331055,
+      "learning_rate": 8.47313596491228e-06,
+      "loss": 1.5063,
+      "step": 12615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6752054691314697,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 1.7461,
+      "step": 12616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6713802814483643,
+      "learning_rate": 8.471308479532165e-06,
+      "loss": 1.3543,
+      "step": 12617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.938707113265991,
+      "learning_rate": 8.470394736842106e-06,
+      "loss": 1.3743,
+      "step": 12618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4366462230682373,
+      "learning_rate": 8.469480994152048e-06,
+      "loss": 1.538,
+      "step": 12619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.420006513595581,
+      "learning_rate": 8.468567251461988e-06,
+      "loss": 1.5781,
+      "step": 12620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.739232063293457,
+      "learning_rate": 8.46765350877193e-06,
+      "loss": 1.4429,
+      "step": 12621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.50893235206604,
+      "learning_rate": 8.466739766081871e-06,
+      "loss": 1.4835,
+      "step": 12622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.669718027114868,
+      "learning_rate": 8.465826023391813e-06,
+      "loss": 1.3834,
+      "step": 12623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.242821455001831,
+      "learning_rate": 8.464912280701755e-06,
+      "loss": 1.5958,
+      "step": 12624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3537819385528564,
+      "learning_rate": 8.463998538011696e-06,
+      "loss": 1.5571,
+      "step": 12625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.931824207305908,
+      "learning_rate": 8.463084795321638e-06,
+      "loss": 1.3025,
+      "step": 12626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.377619743347168,
+      "learning_rate": 8.462171052631579e-06,
+      "loss": 1.1789,
+      "step": 12627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.292433023452759,
+      "learning_rate": 8.461257309941521e-06,
+      "loss": 1.4751,
+      "step": 12628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.548847198486328,
+      "learning_rate": 8.460343567251462e-06,
+      "loss": 1.4068,
+      "step": 12629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5041990280151367,
+      "learning_rate": 8.459429824561404e-06,
+      "loss": 1.4134,
+      "step": 12630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4618332386016846,
+      "learning_rate": 8.458516081871346e-06,
+      "loss": 1.3553,
+      "step": 12631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3695764541625977,
+      "learning_rate": 8.457602339181286e-06,
+      "loss": 1.6507,
+      "step": 12632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.85923171043396,
+      "learning_rate": 8.456688596491229e-06,
+      "loss": 1.5612,
+      "step": 12633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0356287956237793,
+      "learning_rate": 8.455774853801171e-06,
+      "loss": 1.0807,
+      "step": 12634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.402526617050171,
+      "learning_rate": 8.454861111111111e-06,
+      "loss": 1.4418,
+      "step": 12635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2630228996276855,
+      "learning_rate": 8.453947368421054e-06,
+      "loss": 1.3655,
+      "step": 12636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.315884828567505,
+      "learning_rate": 8.453033625730996e-06,
+      "loss": 1.4603,
+      "step": 12637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2369871139526367,
+      "learning_rate": 8.452119883040936e-06,
+      "loss": 1.2355,
+      "step": 12638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.587456703186035,
+      "learning_rate": 8.451206140350879e-06,
+      "loss": 1.6119,
+      "step": 12639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2108876705169678,
+      "learning_rate": 8.45029239766082e-06,
+      "loss": 1.3327,
+      "step": 12640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.519298791885376,
+      "learning_rate": 8.449378654970761e-06,
+      "loss": 1.2345,
+      "step": 12641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7444100379943848,
+      "learning_rate": 8.448464912280702e-06,
+      "loss": 1.0538,
+      "step": 12642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.997068166732788,
+      "learning_rate": 8.447551169590644e-06,
+      "loss": 1.5727,
+      "step": 12643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.464881181716919,
+      "learning_rate": 8.446637426900586e-06,
+      "loss": 1.6975,
+      "step": 12644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.426893472671509,
+      "learning_rate": 8.445723684210527e-06,
+      "loss": 1.3195,
+      "step": 12645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.349907159805298,
+      "learning_rate": 8.444809941520469e-06,
+      "loss": 1.2353,
+      "step": 12646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.639350414276123,
+      "learning_rate": 8.44389619883041e-06,
+      "loss": 1.6304,
+      "step": 12647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5553700923919678,
+      "learning_rate": 8.442982456140352e-06,
+      "loss": 1.372,
+      "step": 12648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.440117120742798,
+      "learning_rate": 8.442068713450292e-06,
+      "loss": 1.3011,
+      "step": 12649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6287529468536377,
+      "learning_rate": 8.441154970760235e-06,
+      "loss": 1.4595,
+      "step": 12650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3513035774230957,
+      "learning_rate": 8.440241228070177e-06,
+      "loss": 1.3625,
+      "step": 12651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8049957752227783,
+      "learning_rate": 8.439327485380117e-06,
+      "loss": 1.8513,
+      "step": 12652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.934016227722168,
+      "learning_rate": 8.43841374269006e-06,
+      "loss": 1.7683,
+      "step": 12653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7990047931671143,
+      "learning_rate": 8.4375e-06,
+      "loss": 1.5626,
+      "step": 12654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.573697805404663,
+      "learning_rate": 8.436586257309942e-06,
+      "loss": 1.5224,
+      "step": 12655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.70402455329895,
+      "learning_rate": 8.435672514619883e-06,
+      "loss": 1.2315,
+      "step": 12656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.524961471557617,
+      "learning_rate": 8.434758771929825e-06,
+      "loss": 1.4983,
+      "step": 12657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.507032871246338,
+      "learning_rate": 8.433845029239767e-06,
+      "loss": 1.4388,
+      "step": 12658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0714409351348877,
+      "learning_rate": 8.432931286549708e-06,
+      "loss": 1.102,
+      "step": 12659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7525181770324707,
+      "learning_rate": 8.43201754385965e-06,
+      "loss": 1.4437,
+      "step": 12660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.834744453430176,
+      "learning_rate": 8.43110380116959e-06,
+      "loss": 1.4851,
+      "step": 12661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2852728366851807,
+      "learning_rate": 8.430190058479533e-06,
+      "loss": 1.4778,
+      "step": 12662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5099570751190186,
+      "learning_rate": 8.429276315789473e-06,
+      "loss": 1.5238,
+      "step": 12663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7564079761505127,
+      "learning_rate": 8.428362573099416e-06,
+      "loss": 1.6277,
+      "step": 12664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2980380058288574,
+      "learning_rate": 8.427448830409358e-06,
+      "loss": 1.3009,
+      "step": 12665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6227588653564453,
+      "learning_rate": 8.426535087719298e-06,
+      "loss": 1.4536,
+      "step": 12666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9147212505340576,
+      "learning_rate": 8.42562134502924e-06,
+      "loss": 1.8429,
+      "step": 12667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5390541553497314,
+      "learning_rate": 8.424707602339181e-06,
+      "loss": 1.4545,
+      "step": 12668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6184425354003906,
+      "learning_rate": 8.423793859649123e-06,
+      "loss": 1.5351,
+      "step": 12669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9830374717712402,
+      "learning_rate": 8.422880116959064e-06,
+      "loss": 1.3039,
+      "step": 12670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2866694927215576,
+      "learning_rate": 8.421966374269006e-06,
+      "loss": 1.3518,
+      "step": 12671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.478238105773926,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 1.2353,
+      "step": 12672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6444365978240967,
+      "learning_rate": 8.420138888888889e-06,
+      "loss": 1.6808,
+      "step": 12673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6989171504974365,
+      "learning_rate": 8.419225146198831e-06,
+      "loss": 1.9599,
+      "step": 12674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5894105434417725,
+      "learning_rate": 8.418311403508772e-06,
+      "loss": 1.6622,
+      "step": 12675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.274312734603882,
+      "learning_rate": 8.417397660818715e-06,
+      "loss": 1.662,
+      "step": 12676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2483725547790527,
+      "learning_rate": 8.416483918128656e-06,
+      "loss": 1.3778,
+      "step": 12677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.294825792312622,
+      "learning_rate": 8.415570175438598e-06,
+      "loss": 1.2715,
+      "step": 12678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8768410682678223,
+      "learning_rate": 8.414656432748539e-06,
+      "loss": 1.6909,
+      "step": 12679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5347652435302734,
+      "learning_rate": 8.413742690058481e-06,
+      "loss": 1.6583,
+      "step": 12680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9514544010162354,
+      "learning_rate": 8.412828947368421e-06,
+      "loss": 1.7385,
+      "step": 12681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7068328857421875,
+      "learning_rate": 8.411915204678364e-06,
+      "loss": 1.425,
+      "step": 12682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9488909244537354,
+      "learning_rate": 8.411001461988306e-06,
+      "loss": 1.6843,
+      "step": 12683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.884199619293213,
+      "learning_rate": 8.410087719298246e-06,
+      "loss": 1.8234,
+      "step": 12684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5894737243652344,
+      "learning_rate": 8.409173976608189e-06,
+      "loss": 1.4383,
+      "step": 12685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.654891014099121,
+      "learning_rate": 8.40826023391813e-06,
+      "loss": 1.3309,
+      "step": 12686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8645782470703125,
+      "learning_rate": 8.407346491228071e-06,
+      "loss": 1.3677,
+      "step": 12687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7645246982574463,
+      "learning_rate": 8.406432748538012e-06,
+      "loss": 1.5731,
+      "step": 12688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6248650550842285,
+      "learning_rate": 8.405519005847954e-06,
+      "loss": 1.5354,
+      "step": 12689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.231651544570923,
+      "learning_rate": 8.404605263157896e-06,
+      "loss": 1.3987,
+      "step": 12690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1570053100585938,
+      "learning_rate": 8.403691520467837e-06,
+      "loss": 1.2245,
+      "step": 12691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.119187355041504,
+      "learning_rate": 8.402777777777779e-06,
+      "loss": 1.5166,
+      "step": 12692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0248489379882812,
+      "learning_rate": 8.40186403508772e-06,
+      "loss": 1.8721,
+      "step": 12693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.644367218017578,
+      "learning_rate": 8.400950292397662e-06,
+      "loss": 1.5921,
+      "step": 12694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4242756366729736,
+      "learning_rate": 8.400036549707602e-06,
+      "loss": 1.8437,
+      "step": 12695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.523317337036133,
+      "learning_rate": 8.399122807017545e-06,
+      "loss": 1.3717,
+      "step": 12696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.562551736831665,
+      "learning_rate": 8.398209064327487e-06,
+      "loss": 1.3439,
+      "step": 12697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.671900987625122,
+      "learning_rate": 8.397295321637427e-06,
+      "loss": 1.4873,
+      "step": 12698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.225806713104248,
+      "learning_rate": 8.39638157894737e-06,
+      "loss": 1.2019,
+      "step": 12699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.347059488296509,
+      "learning_rate": 8.39546783625731e-06,
+      "loss": 1.3454,
+      "step": 12700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4404451847076416,
+      "learning_rate": 8.394554093567252e-06,
+      "loss": 1.3006,
+      "step": 12701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.529944896697998,
+      "learning_rate": 8.393640350877193e-06,
+      "loss": 1.2844,
+      "step": 12702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.697324514389038,
+      "learning_rate": 8.392726608187135e-06,
+      "loss": 1.8676,
+      "step": 12703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.058346748352051,
+      "learning_rate": 8.391812865497077e-06,
+      "loss": 1.3092,
+      "step": 12704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4563558101654053,
+      "learning_rate": 8.390899122807018e-06,
+      "loss": 1.5277,
+      "step": 12705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.719409465789795,
+      "learning_rate": 8.38998538011696e-06,
+      "loss": 1.2041,
+      "step": 12706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5825843811035156,
+      "learning_rate": 8.3890716374269e-06,
+      "loss": 1.4986,
+      "step": 12707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.808178663253784,
+      "learning_rate": 8.388157894736843e-06,
+      "loss": 1.758,
+      "step": 12708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.738985061645508,
+      "learning_rate": 8.387244152046783e-06,
+      "loss": 1.9998,
+      "step": 12709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.017854928970337,
+      "learning_rate": 8.386330409356726e-06,
+      "loss": 1.1502,
+      "step": 12710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3483893871307373,
+      "learning_rate": 8.385416666666668e-06,
+      "loss": 1.6151,
+      "step": 12711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7072806358337402,
+      "learning_rate": 8.384502923976608e-06,
+      "loss": 1.824,
+      "step": 12712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.175941228866577,
+      "learning_rate": 8.38358918128655e-06,
+      "loss": 1.1401,
+      "step": 12713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1704514026641846,
+      "learning_rate": 8.382675438596491e-06,
+      "loss": 1.93,
+      "step": 12714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.777844190597534,
+      "learning_rate": 8.381761695906433e-06,
+      "loss": 1.5868,
+      "step": 12715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.981093645095825,
+      "learning_rate": 8.380847953216374e-06,
+      "loss": 1.2758,
+      "step": 12716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5078625679016113,
+      "learning_rate": 8.379934210526316e-06,
+      "loss": 1.5421,
+      "step": 12717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8015763759613037,
+      "learning_rate": 8.379020467836258e-06,
+      "loss": 1.7403,
+      "step": 12718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0679163932800293,
+      "learning_rate": 8.3781067251462e-06,
+      "loss": 1.3651,
+      "step": 12719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.54180645942688,
+      "learning_rate": 8.377192982456141e-06,
+      "loss": 1.2844,
+      "step": 12720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8455727100372314,
+      "learning_rate": 8.376279239766083e-06,
+      "loss": 1.4607,
+      "step": 12721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.554594039916992,
+      "learning_rate": 8.375365497076024e-06,
+      "loss": 1.5083,
+      "step": 12722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6399319171905518,
+      "learning_rate": 8.374451754385966e-06,
+      "loss": 1.1815,
+      "step": 12723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1570723056793213,
+      "learning_rate": 8.373538011695908e-06,
+      "loss": 1.2897,
+      "step": 12724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.082369804382324,
+      "learning_rate": 8.372624269005849e-06,
+      "loss": 1.8104,
+      "step": 12725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1192853450775146,
+      "learning_rate": 8.371710526315791e-06,
+      "loss": 1.2544,
+      "step": 12726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7910962104797363,
+      "learning_rate": 8.370796783625731e-06,
+      "loss": 1.615,
+      "step": 12727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7661519050598145,
+      "learning_rate": 8.369883040935674e-06,
+      "loss": 1.8343,
+      "step": 12728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.511488437652588,
+      "learning_rate": 8.368969298245614e-06,
+      "loss": 1.456,
+      "step": 12729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0215513706207275,
+      "learning_rate": 8.368055555555556e-06,
+      "loss": 1.0951,
+      "step": 12730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.345029354095459,
+      "learning_rate": 8.367141812865499e-06,
+      "loss": 1.5356,
+      "step": 12731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6339111328125,
+      "learning_rate": 8.36622807017544e-06,
+      "loss": 1.6932,
+      "step": 12732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3383986949920654,
+      "learning_rate": 8.365314327485381e-06,
+      "loss": 1.2846,
+      "step": 12733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9052915573120117,
+      "learning_rate": 8.364400584795322e-06,
+      "loss": 1.7785,
+      "step": 12734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9275734424591064,
+      "learning_rate": 8.363486842105264e-06,
+      "loss": 1.053,
+      "step": 12735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.413590669631958,
+      "learning_rate": 8.362573099415205e-06,
+      "loss": 1.5513,
+      "step": 12736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.603304862976074,
+      "learning_rate": 8.361659356725147e-06,
+      "loss": 1.3319,
+      "step": 12737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0787999629974365,
+      "learning_rate": 8.360745614035089e-06,
+      "loss": 1.9766,
+      "step": 12738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.283649444580078,
+      "learning_rate": 8.35983187134503e-06,
+      "loss": 1.5738,
+      "step": 12739
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3074357509613037,
+      "learning_rate": 8.358918128654972e-06,
+      "loss": 1.8904,
+      "step": 12740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5518360137939453,
+      "learning_rate": 8.358004385964912e-06,
+      "loss": 1.3274,
+      "step": 12741
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.453155517578125,
+      "learning_rate": 8.357090643274855e-06,
+      "loss": 1.5152,
+      "step": 12742
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4826431274414062,
+      "learning_rate": 8.356176900584795e-06,
+      "loss": 1.6175,
+      "step": 12743
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.077599048614502,
+      "learning_rate": 8.355263157894737e-06,
+      "loss": 1.1614,
+      "step": 12744
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4458913803100586,
+      "learning_rate": 8.35434941520468e-06,
+      "loss": 1.2093,
+      "step": 12745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.148676872253418,
+      "learning_rate": 8.35343567251462e-06,
+      "loss": 1.1529,
+      "step": 12746
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.190603733062744,
+      "learning_rate": 8.352521929824562e-06,
+      "loss": 1.1385,
+      "step": 12747
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0837013721466064,
+      "learning_rate": 8.351608187134503e-06,
+      "loss": 1.799,
+      "step": 12748
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.005418062210083,
+      "learning_rate": 8.350694444444445e-06,
+      "loss": 1.6169,
+      "step": 12749
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4495291709899902,
+      "learning_rate": 8.349780701754386e-06,
+      "loss": 1.3459,
+      "step": 12750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3627095222473145,
+      "learning_rate": 8.348866959064328e-06,
+      "loss": 1.2578,
+      "step": 12751
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6179256439208984,
+      "learning_rate": 8.34795321637427e-06,
+      "loss": 1.5146,
+      "step": 12752
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6858413219451904,
+      "learning_rate": 8.34703947368421e-06,
+      "loss": 1.6826,
+      "step": 12753
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2026126384735107,
+      "learning_rate": 8.346125730994153e-06,
+      "loss": 1.9211,
+      "step": 12754
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.289072275161743,
+      "learning_rate": 8.345211988304093e-06,
+      "loss": 1.3469,
+      "step": 12755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2064056396484375,
+      "learning_rate": 8.344298245614036e-06,
+      "loss": 1.6122,
+      "step": 12756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4655113220214844,
+      "learning_rate": 8.343384502923976e-06,
+      "loss": 1.4961,
+      "step": 12757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8511533737182617,
+      "learning_rate": 8.342470760233918e-06,
+      "loss": 1.5857,
+      "step": 12758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4887542724609375,
+      "learning_rate": 8.34155701754386e-06,
+      "loss": 1.6809,
+      "step": 12759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1485025882720947,
+      "learning_rate": 8.340643274853801e-06,
+      "loss": 2.185,
+      "step": 12760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9395270347595215,
+      "learning_rate": 8.339729532163743e-06,
+      "loss": 1.3514,
+      "step": 12761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7692716121673584,
+      "learning_rate": 8.338815789473685e-06,
+      "loss": 1.6255,
+      "step": 12762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.448875665664673,
+      "learning_rate": 8.337902046783628e-06,
+      "loss": 1.3222,
+      "step": 12763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8546483516693115,
+      "learning_rate": 8.336988304093568e-06,
+      "loss": 1.4741,
+      "step": 12764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0907092094421387,
+      "learning_rate": 8.33607456140351e-06,
+      "loss": 1.2957,
+      "step": 12765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.50126314163208,
+      "learning_rate": 8.335160818713451e-06,
+      "loss": 1.3887,
+      "step": 12766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.917520046234131,
+      "learning_rate": 8.334247076023393e-06,
+      "loss": 1.7778,
+      "step": 12767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.499324321746826,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.5407,
+      "step": 12768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2245752811431885,
+      "learning_rate": 8.332419590643276e-06,
+      "loss": 1.4522,
+      "step": 12769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.104250907897949,
+      "learning_rate": 8.331505847953218e-06,
+      "loss": 1.28,
+      "step": 12770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.94860577583313,
+      "learning_rate": 8.330592105263159e-06,
+      "loss": 1.9232,
+      "step": 12771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.482832431793213,
+      "learning_rate": 8.329678362573101e-06,
+      "loss": 1.3682,
+      "step": 12772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.623342990875244,
+      "learning_rate": 8.328764619883041e-06,
+      "loss": 1.6583,
+      "step": 12773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5072762966156006,
+      "learning_rate": 8.327850877192984e-06,
+      "loss": 1.2686,
+      "step": 12774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6765568256378174,
+      "learning_rate": 8.326937134502924e-06,
+      "loss": 1.132,
+      "step": 12775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7033486366271973,
+      "learning_rate": 8.326023391812866e-06,
+      "loss": 1.3736,
+      "step": 12776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3994789123535156,
+      "learning_rate": 8.325109649122809e-06,
+      "loss": 1.3762,
+      "step": 12777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.585986614227295,
+      "learning_rate": 8.32419590643275e-06,
+      "loss": 1.6443,
+      "step": 12778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.303879737854004,
+      "learning_rate": 8.323282163742691e-06,
+      "loss": 1.5997,
+      "step": 12779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.816514253616333,
+      "learning_rate": 8.322368421052632e-06,
+      "loss": 0.8646,
+      "step": 12780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7325949668884277,
+      "learning_rate": 8.321454678362574e-06,
+      "loss": 1.6631,
+      "step": 12781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5468344688415527,
+      "learning_rate": 8.320540935672515e-06,
+      "loss": 1.4362,
+      "step": 12782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6664371490478516,
+      "learning_rate": 8.319627192982457e-06,
+      "loss": 1.5258,
+      "step": 12783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.361743927001953,
+      "learning_rate": 8.318713450292399e-06,
+      "loss": 1.2282,
+      "step": 12784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.463547706604004,
+      "learning_rate": 8.31779970760234e-06,
+      "loss": 1.4094,
+      "step": 12785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9875712394714355,
+      "learning_rate": 8.316885964912282e-06,
+      "loss": 1.6875,
+      "step": 12786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.86206316947937,
+      "learning_rate": 8.315972222222222e-06,
+      "loss": 1.1536,
+      "step": 12787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0159311294555664,
+      "learning_rate": 8.315058479532165e-06,
+      "loss": 1.5463,
+      "step": 12788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.857604503631592,
+      "learning_rate": 8.314144736842105e-06,
+      "loss": 1.4852,
+      "step": 12789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.12091326713562,
+      "learning_rate": 8.313230994152047e-06,
+      "loss": 1.4575,
+      "step": 12790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.685293674468994,
+      "learning_rate": 8.31231725146199e-06,
+      "loss": 1.8161,
+      "step": 12791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3705015182495117,
+      "learning_rate": 8.31140350877193e-06,
+      "loss": 1.2455,
+      "step": 12792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5695996284484863,
+      "learning_rate": 8.310489766081872e-06,
+      "loss": 1.5139,
+      "step": 12793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.806349754333496,
+      "learning_rate": 8.309576023391813e-06,
+      "loss": 1.5541,
+      "step": 12794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4595866203308105,
+      "learning_rate": 8.308662280701755e-06,
+      "loss": 1.3717,
+      "step": 12795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.497943162918091,
+      "learning_rate": 8.307748538011696e-06,
+      "loss": 1.5445,
+      "step": 12796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.305845022201538,
+      "learning_rate": 8.306834795321638e-06,
+      "loss": 1.1894,
+      "step": 12797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6785902976989746,
+      "learning_rate": 8.30592105263158e-06,
+      "loss": 1.4026,
+      "step": 12798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5008902549743652,
+      "learning_rate": 8.30500730994152e-06,
+      "loss": 1.6018,
+      "step": 12799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.952669382095337,
+      "learning_rate": 8.304093567251463e-06,
+      "loss": 1.6835,
+      "step": 12800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.147198438644409,
+      "learning_rate": 8.303179824561403e-06,
+      "loss": 1.0161,
+      "step": 12801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.515573263168335,
+      "learning_rate": 8.302266081871346e-06,
+      "loss": 1.1509,
+      "step": 12802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.286440134048462,
+      "learning_rate": 8.301352339181286e-06,
+      "loss": 1.5687,
+      "step": 12803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.583897352218628,
+      "learning_rate": 8.30043859649123e-06,
+      "loss": 1.4604,
+      "step": 12804
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2662880420684814,
+      "learning_rate": 8.29952485380117e-06,
+      "loss": 1.2491,
+      "step": 12805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1316144466400146,
+      "learning_rate": 8.298611111111113e-06,
+      "loss": 1.1132,
+      "step": 12806
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4914300441741943,
+      "learning_rate": 8.297697368421053e-06,
+      "loss": 1.5776,
+      "step": 12807
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.311673879623413,
+      "learning_rate": 8.296783625730995e-06,
+      "loss": 1.4101,
+      "step": 12808
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9237101078033447,
+      "learning_rate": 8.295869883040936e-06,
+      "loss": 1.6144,
+      "step": 12809
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2636795043945312,
+      "learning_rate": 8.294956140350878e-06,
+      "loss": 1.3477,
+      "step": 12810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.162597894668579,
+      "learning_rate": 8.29404239766082e-06,
+      "loss": 1.1836,
+      "step": 12811
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3898613452911377,
+      "learning_rate": 8.293128654970761e-06,
+      "loss": 1.4279,
+      "step": 12812
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.892847776412964,
+      "learning_rate": 8.292214912280703e-06,
+      "loss": 1.7574,
+      "step": 12813
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.571895122528076,
+      "learning_rate": 8.291301169590644e-06,
+      "loss": 1.4191,
+      "step": 12814
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9919769763946533,
+      "learning_rate": 8.290387426900586e-06,
+      "loss": 1.6673,
+      "step": 12815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2006239891052246,
+      "learning_rate": 8.289473684210526e-06,
+      "loss": 1.0171,
+      "step": 12816
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7692880630493164,
+      "learning_rate": 8.288559941520469e-06,
+      "loss": 1.6942,
+      "step": 12817
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7747154235839844,
+      "learning_rate": 8.287646198830411e-06,
+      "loss": 1.6867,
+      "step": 12818
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.635115623474121,
+      "learning_rate": 8.286732456140351e-06,
+      "loss": 1.7501,
+      "step": 12819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.603945732116699,
+      "learning_rate": 8.285818713450294e-06,
+      "loss": 1.5881,
+      "step": 12820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6407253742218018,
+      "learning_rate": 8.284904970760234e-06,
+      "loss": 1.616,
+      "step": 12821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3830184936523438,
+      "learning_rate": 8.283991228070176e-06,
+      "loss": 1.5794,
+      "step": 12822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1930203437805176,
+      "learning_rate": 8.283077485380117e-06,
+      "loss": 1.7093,
+      "step": 12823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0504777431488037,
+      "learning_rate": 8.28216374269006e-06,
+      "loss": 1.8095,
+      "step": 12824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5607707500457764,
+      "learning_rate": 8.281250000000001e-06,
+      "loss": 1.557,
+      "step": 12825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8029208183288574,
+      "learning_rate": 8.280336257309942e-06,
+      "loss": 1.0058,
+      "step": 12826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5072195529937744,
+      "learning_rate": 8.279422514619884e-06,
+      "loss": 1.5497,
+      "step": 12827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7831687927246094,
+      "learning_rate": 8.278508771929825e-06,
+      "loss": 1.5186,
+      "step": 12828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0524415969848633,
+      "learning_rate": 8.277595029239767e-06,
+      "loss": 1.684,
+      "step": 12829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7293050289154053,
+      "learning_rate": 8.276681286549707e-06,
+      "loss": 1.4201,
+      "step": 12830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4616332054138184,
+      "learning_rate": 8.27576754385965e-06,
+      "loss": 1.5449,
+      "step": 12831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.451244354248047,
+      "learning_rate": 8.274853801169592e-06,
+      "loss": 1.424,
+      "step": 12832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.624652624130249,
+      "learning_rate": 8.273940058479532e-06,
+      "loss": 1.5733,
+      "step": 12833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9200735092163086,
+      "learning_rate": 8.273026315789475e-06,
+      "loss": 1.8286,
+      "step": 12834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4740822315216064,
+      "learning_rate": 8.272112573099415e-06,
+      "loss": 1.767,
+      "step": 12835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.854670524597168,
+      "learning_rate": 8.271198830409357e-06,
+      "loss": 1.6459,
+      "step": 12836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.686920642852783,
+      "learning_rate": 8.270285087719298e-06,
+      "loss": 1.6751,
+      "step": 12837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3764679431915283,
+      "learning_rate": 8.26937134502924e-06,
+      "loss": 1.5454,
+      "step": 12838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5146031379699707,
+      "learning_rate": 8.268457602339182e-06,
+      "loss": 1.3566,
+      "step": 12839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.590547800064087,
+      "learning_rate": 8.267543859649123e-06,
+      "loss": 1.7557,
+      "step": 12840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0163052082061768,
+      "learning_rate": 8.266630116959065e-06,
+      "loss": 1.2906,
+      "step": 12841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9912755489349365,
+      "learning_rate": 8.265716374269006e-06,
+      "loss": 1.8337,
+      "step": 12842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.16265869140625,
+      "learning_rate": 8.264802631578948e-06,
+      "loss": 1.4886,
+      "step": 12843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.926910400390625,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 1.777,
+      "step": 12844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7597599029541016,
+      "learning_rate": 8.26297514619883e-06,
+      "loss": 1.9952,
+      "step": 12845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7274386882781982,
+      "learning_rate": 8.262061403508773e-06,
+      "loss": 1.6896,
+      "step": 12846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.201767921447754,
+      "learning_rate": 8.261147660818715e-06,
+      "loss": 1.2453,
+      "step": 12847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8600070476531982,
+      "learning_rate": 8.260233918128656e-06,
+      "loss": 1.9179,
+      "step": 12848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3430254459381104,
+      "learning_rate": 8.259320175438598e-06,
+      "loss": 1.3055,
+      "step": 12849
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4826619625091553,
+      "learning_rate": 8.25840643274854e-06,
+      "loss": 1.5358,
+      "step": 12850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.783790111541748,
+      "learning_rate": 8.25749269005848e-06,
+      "loss": 1.6095,
+      "step": 12851
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4975545406341553,
+      "learning_rate": 8.256578947368423e-06,
+      "loss": 1.3505,
+      "step": 12852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2959933280944824,
+      "learning_rate": 8.255665204678363e-06,
+      "loss": 1.4038,
+      "step": 12853
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.038259267807007,
+      "learning_rate": 8.254751461988305e-06,
+      "loss": 1.8101,
+      "step": 12854
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0090389251708984,
+      "learning_rate": 8.253837719298246e-06,
+      "loss": 1.2547,
+      "step": 12855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4204611778259277,
+      "learning_rate": 8.252923976608188e-06,
+      "loss": 1.4415,
+      "step": 12856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.80377459526062,
+      "learning_rate": 8.25201023391813e-06,
+      "loss": 1.2521,
+      "step": 12857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.981017827987671,
+      "learning_rate": 8.251096491228071e-06,
+      "loss": 1.6368,
+      "step": 12858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7468905448913574,
+      "learning_rate": 8.250182748538013e-06,
+      "loss": 1.3487,
+      "step": 12859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2397682666778564,
+      "learning_rate": 8.249269005847954e-06,
+      "loss": 1.278,
+      "step": 12860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.781999349594116,
+      "learning_rate": 8.248355263157896e-06,
+      "loss": 1.2734,
+      "step": 12861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6326987743377686,
+      "learning_rate": 8.247441520467836e-06,
+      "loss": 1.54,
+      "step": 12862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6304171085357666,
+      "learning_rate": 8.246527777777779e-06,
+      "loss": 1.7506,
+      "step": 12863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.863593339920044,
+      "learning_rate": 8.24561403508772e-06,
+      "loss": 1.4497,
+      "step": 12864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.183415412902832,
+      "learning_rate": 8.244700292397661e-06,
+      "loss": 1.1948,
+      "step": 12865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.228766441345215,
+      "learning_rate": 8.243786549707604e-06,
+      "loss": 1.4941,
+      "step": 12866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.323258876800537,
+      "learning_rate": 8.242872807017544e-06,
+      "loss": 1.2834,
+      "step": 12867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.502002000808716,
+      "learning_rate": 8.241959064327486e-06,
+      "loss": 1.4418,
+      "step": 12868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8574790954589844,
+      "learning_rate": 8.241045321637427e-06,
+      "loss": 1.3219,
+      "step": 12869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9600181579589844,
+      "learning_rate": 8.24013157894737e-06,
+      "loss": 1.5833,
+      "step": 12870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.426407814025879,
+      "learning_rate": 8.23921783625731e-06,
+      "loss": 1.5453,
+      "step": 12871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5566439628601074,
+      "learning_rate": 8.238304093567252e-06,
+      "loss": 1.8204,
+      "step": 12872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.766568183898926,
+      "learning_rate": 8.237390350877194e-06,
+      "loss": 1.5285,
+      "step": 12873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.374087333679199,
+      "learning_rate": 8.236476608187135e-06,
+      "loss": 1.2191,
+      "step": 12874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.867266893386841,
+      "learning_rate": 8.235562865497077e-06,
+      "loss": 1.4173,
+      "step": 12875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.692495822906494,
+      "learning_rate": 8.234649122807017e-06,
+      "loss": 1.2511,
+      "step": 12876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3812625408172607,
+      "learning_rate": 8.23373538011696e-06,
+      "loss": 1.3634,
+      "step": 12877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7084717750549316,
+      "learning_rate": 8.2328216374269e-06,
+      "loss": 1.4339,
+      "step": 12878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5098588466644287,
+      "learning_rate": 8.231907894736842e-06,
+      "loss": 1.6009,
+      "step": 12879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5317726135253906,
+      "learning_rate": 8.230994152046785e-06,
+      "loss": 1.7627,
+      "step": 12880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.816819190979004,
+      "learning_rate": 8.230080409356725e-06,
+      "loss": 1.7364,
+      "step": 12881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.852259635925293,
+      "learning_rate": 8.229166666666667e-06,
+      "loss": 1.5877,
+      "step": 12882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.681049108505249,
+      "learning_rate": 8.228252923976608e-06,
+      "loss": 1.3987,
+      "step": 12883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.816866159439087,
+      "learning_rate": 8.22733918128655e-06,
+      "loss": 1.8321,
+      "step": 12884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.950636386871338,
+      "learning_rate": 8.22642543859649e-06,
+      "loss": 1.4553,
+      "step": 12885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6501221656799316,
+      "learning_rate": 8.225511695906433e-06,
+      "loss": 1.3089,
+      "step": 12886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1697895526885986,
+      "learning_rate": 8.224597953216375e-06,
+      "loss": 1.1165,
+      "step": 12887
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.204085350036621,
+      "learning_rate": 8.223684210526316e-06,
+      "loss": 1.3077,
+      "step": 12888
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3910884857177734,
+      "learning_rate": 8.222770467836258e-06,
+      "loss": 1.3819,
+      "step": 12889
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9358904361724854,
+      "learning_rate": 8.2218567251462e-06,
+      "loss": 1.7399,
+      "step": 12890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3571205139160156,
+      "learning_rate": 8.220942982456142e-06,
+      "loss": 1.4504,
+      "step": 12891
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4218590259552,
+      "learning_rate": 8.220029239766083e-06,
+      "loss": 1.6217,
+      "step": 12892
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5818161964416504,
+      "learning_rate": 8.219115497076025e-06,
+      "loss": 1.4732,
+      "step": 12893
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8526198863983154,
+      "learning_rate": 8.218201754385966e-06,
+      "loss": 1.4236,
+      "step": 12894
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8376474380493164,
+      "learning_rate": 8.217288011695908e-06,
+      "loss": 1.7062,
+      "step": 12895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.784459114074707,
+      "learning_rate": 8.216374269005848e-06,
+      "loss": 1.7108,
+      "step": 12896
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9348318576812744,
+      "learning_rate": 8.21546052631579e-06,
+      "loss": 1.7592,
+      "step": 12897
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.578838348388672,
+      "learning_rate": 8.214546783625733e-06,
+      "loss": 1.3195,
+      "step": 12898
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.287368059158325,
+      "learning_rate": 8.213633040935673e-06,
+      "loss": 1.4213,
+      "step": 12899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.859919548034668,
+      "learning_rate": 8.212719298245615e-06,
+      "loss": 1.3041,
+      "step": 12900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.840036392211914,
+      "learning_rate": 8.211805555555556e-06,
+      "loss": 1.9813,
+      "step": 12901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8193929195404053,
+      "learning_rate": 8.210891812865498e-06,
+      "loss": 1.5831,
+      "step": 12902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9917664527893066,
+      "learning_rate": 8.209978070175439e-06,
+      "loss": 1.5517,
+      "step": 12903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2248754501342773,
+      "learning_rate": 8.209064327485381e-06,
+      "loss": 1.3032,
+      "step": 12904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4295337200164795,
+      "learning_rate": 8.208150584795323e-06,
+      "loss": 1.5021,
+      "step": 12905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.634192943572998,
+      "learning_rate": 8.207236842105264e-06,
+      "loss": 1.3754,
+      "step": 12906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.972768783569336,
+      "learning_rate": 8.206323099415206e-06,
+      "loss": 1.7899,
+      "step": 12907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5415585041046143,
+      "learning_rate": 8.205409356725146e-06,
+      "loss": 1.4069,
+      "step": 12908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6689324378967285,
+      "learning_rate": 8.204495614035089e-06,
+      "loss": 1.7848,
+      "step": 12909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2562334537506104,
+      "learning_rate": 8.20358187134503e-06,
+      "loss": 2.0873,
+      "step": 12910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5941944122314453,
+      "learning_rate": 8.202668128654971e-06,
+      "loss": 1.5964,
+      "step": 12911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7148568630218506,
+      "learning_rate": 8.201754385964914e-06,
+      "loss": 1.4418,
+      "step": 12912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8187854290008545,
+      "learning_rate": 8.200840643274854e-06,
+      "loss": 1.2096,
+      "step": 12913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.684823751449585,
+      "learning_rate": 8.199926900584796e-06,
+      "loss": 1.6665,
+      "step": 12914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5824429988861084,
+      "learning_rate": 8.199013157894737e-06,
+      "loss": 1.2265,
+      "step": 12915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.628997802734375,
+      "learning_rate": 8.19809941520468e-06,
+      "loss": 1.5827,
+      "step": 12916
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3220958709716797,
+      "learning_rate": 8.19718567251462e-06,
+      "loss": 1.3346,
+      "step": 12917
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.056534767150879,
+      "learning_rate": 8.196271929824562e-06,
+      "loss": 1.1268,
+      "step": 12918
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.892461061477661,
+      "learning_rate": 8.195358187134504e-06,
+      "loss": 1.8112,
+      "step": 12919
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2279443740844727,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 1.4356,
+      "step": 12920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2855851650238037,
+      "learning_rate": 8.193530701754387e-06,
+      "loss": 1.3116,
+      "step": 12921
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4731597900390625,
+      "learning_rate": 8.192616959064327e-06,
+      "loss": 1.4552,
+      "step": 12922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7025671005249023,
+      "learning_rate": 8.19170321637427e-06,
+      "loss": 1.6002,
+      "step": 12923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.287541389465332,
+      "learning_rate": 8.19078947368421e-06,
+      "loss": 1.2452,
+      "step": 12924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4858920574188232,
+      "learning_rate": 8.189875730994152e-06,
+      "loss": 1.4473,
+      "step": 12925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2308387756347656,
+      "learning_rate": 8.188961988304095e-06,
+      "loss": 1.5007,
+      "step": 12926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4264700412750244,
+      "learning_rate": 8.188048245614035e-06,
+      "loss": 1.4017,
+      "step": 12927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7358224391937256,
+      "learning_rate": 8.187134502923977e-06,
+      "loss": 1.2694,
+      "step": 12928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2402565479278564,
+      "learning_rate": 8.186220760233918e-06,
+      "loss": 1.1648,
+      "step": 12929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7425527572631836,
+      "learning_rate": 8.18530701754386e-06,
+      "loss": 1.5168,
+      "step": 12930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.496171474456787,
+      "learning_rate": 8.1843932748538e-06,
+      "loss": 1.4544,
+      "step": 12931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1712777614593506,
+      "learning_rate": 8.183479532163745e-06,
+      "loss": 1.2863,
+      "step": 12932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3739430904388428,
+      "learning_rate": 8.182565789473685e-06,
+      "loss": 1.3523,
+      "step": 12933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6099441051483154,
+      "learning_rate": 8.181652046783627e-06,
+      "loss": 1.3764,
+      "step": 12934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.383955240249634,
+      "learning_rate": 8.180738304093568e-06,
+      "loss": 1.5691,
+      "step": 12935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.603916883468628,
+      "learning_rate": 8.17982456140351e-06,
+      "loss": 1.4175,
+      "step": 12936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.600620746612549,
+      "learning_rate": 8.17891081871345e-06,
+      "loss": 1.3674,
+      "step": 12937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.765310764312744,
+      "learning_rate": 8.177997076023393e-06,
+      "loss": 1.4399,
+      "step": 12938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5407357215881348,
+      "learning_rate": 8.177083333333335e-06,
+      "loss": 1.3147,
+      "step": 12939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1911544799804688,
+      "learning_rate": 8.176169590643276e-06,
+      "loss": 1.3518,
+      "step": 12940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.510918140411377,
+      "learning_rate": 8.175255847953218e-06,
+      "loss": 1.5374,
+      "step": 12941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.051360845565796,
+      "learning_rate": 8.174342105263158e-06,
+      "loss": 1.6605,
+      "step": 12942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6546854972839355,
+      "learning_rate": 8.1734283625731e-06,
+      "loss": 1.4174,
+      "step": 12943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5194637775421143,
+      "learning_rate": 8.172514619883041e-06,
+      "loss": 1.2934,
+      "step": 12944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7706966400146484,
+      "learning_rate": 8.171600877192983e-06,
+      "loss": 1.5987,
+      "step": 12945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4014854431152344,
+      "learning_rate": 8.170687134502925e-06,
+      "loss": 1.3969,
+      "step": 12946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.625469446182251,
+      "learning_rate": 8.169773391812866e-06,
+      "loss": 1.4875,
+      "step": 12947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0895066261291504,
+      "learning_rate": 8.168859649122808e-06,
+      "loss": 1.6491,
+      "step": 12948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0812084674835205,
+      "learning_rate": 8.167945906432749e-06,
+      "loss": 1.513,
+      "step": 12949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6268112659454346,
+      "learning_rate": 8.167032163742691e-06,
+      "loss": 1.5722,
+      "step": 12950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8457868099212646,
+      "learning_rate": 8.166118421052632e-06,
+      "loss": 1.4909,
+      "step": 12951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0285513401031494,
+      "learning_rate": 8.165204678362574e-06,
+      "loss": 1.6234,
+      "step": 12952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5057597160339355,
+      "learning_rate": 8.164290935672516e-06,
+      "loss": 1.7074,
+      "step": 12953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9605873823165894,
+      "learning_rate": 8.163377192982456e-06,
+      "loss": 1.0356,
+      "step": 12954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9394917488098145,
+      "learning_rate": 8.162463450292399e-06,
+      "loss": 1.3545,
+      "step": 12955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1565768718719482,
+      "learning_rate": 8.16154970760234e-06,
+      "loss": 1.3191,
+      "step": 12956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.703920841217041,
+      "learning_rate": 8.160635964912281e-06,
+      "loss": 1.7447,
+      "step": 12957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.92071533203125,
+      "learning_rate": 8.159722222222222e-06,
+      "loss": 1.6701,
+      "step": 12958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.389211654663086,
+      "learning_rate": 8.158808479532164e-06,
+      "loss": 1.2566,
+      "step": 12959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2921366691589355,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 1.3233,
+      "step": 12960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.147927761077881,
+      "learning_rate": 8.156980994152047e-06,
+      "loss": 1.2409,
+      "step": 12961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3480489253997803,
+      "learning_rate": 8.15606725146199e-06,
+      "loss": 1.385,
+      "step": 12962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.264270544052124,
+      "learning_rate": 8.15515350877193e-06,
+      "loss": 1.4141,
+      "step": 12963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.997927188873291,
+      "learning_rate": 8.154239766081872e-06,
+      "loss": 1.7812,
+      "step": 12964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.486302852630615,
+      "learning_rate": 8.153326023391812e-06,
+      "loss": 2.5232,
+      "step": 12965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2643067836761475,
+      "learning_rate": 8.152412280701755e-06,
+      "loss": 1.1411,
+      "step": 12966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1997056007385254,
+      "learning_rate": 8.151498538011697e-06,
+      "loss": 1.4917,
+      "step": 12967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.811305046081543,
+      "learning_rate": 8.150584795321637e-06,
+      "loss": 1.7929,
+      "step": 12968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2951772212982178,
+      "learning_rate": 8.14967105263158e-06,
+      "loss": 1.444,
+      "step": 12969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3100411891937256,
+      "learning_rate": 8.14875730994152e-06,
+      "loss": 1.3077,
+      "step": 12970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.542917251586914,
+      "learning_rate": 8.147843567251462e-06,
+      "loss": 1.8156,
+      "step": 12971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6072967052459717,
+      "learning_rate": 8.146929824561403e-06,
+      "loss": 1.6196,
+      "step": 12972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3362159729003906,
+      "learning_rate": 8.146016081871345e-06,
+      "loss": 1.3185,
+      "step": 12973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.31048583984375,
+      "learning_rate": 8.145102339181287e-06,
+      "loss": 1.4646,
+      "step": 12974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4440033435821533,
+      "learning_rate": 8.14418859649123e-06,
+      "loss": 1.4271,
+      "step": 12975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.704941987991333,
+      "learning_rate": 8.14327485380117e-06,
+      "loss": 1.6259,
+      "step": 12976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9009345769882202,
+      "learning_rate": 8.142361111111112e-06,
+      "loss": 1.0788,
+      "step": 12977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.333521604537964,
+      "learning_rate": 8.141447368421055e-06,
+      "loss": 1.4546,
+      "step": 12978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.024510622024536,
+      "learning_rate": 8.140533625730995e-06,
+      "loss": 1.704,
+      "step": 12979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9213666915893555,
+      "learning_rate": 8.139619883040937e-06,
+      "loss": 1.9633,
+      "step": 12980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.274805784225464,
+      "learning_rate": 8.138706140350878e-06,
+      "loss": 1.397,
+      "step": 12981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8650362491607666,
+      "learning_rate": 8.13779239766082e-06,
+      "loss": 1.7195,
+      "step": 12982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.385798692703247,
+      "learning_rate": 8.13687865497076e-06,
+      "loss": 1.3504,
+      "step": 12983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4056897163391113,
+      "learning_rate": 8.135964912280703e-06,
+      "loss": 1.3252,
+      "step": 12984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4298157691955566,
+      "learning_rate": 8.135051169590645e-06,
+      "loss": 1.2086,
+      "step": 12985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9372036457061768,
+      "learning_rate": 8.134137426900586e-06,
+      "loss": 1.7656,
+      "step": 12986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3153655529022217,
+      "learning_rate": 8.133223684210528e-06,
+      "loss": 1.2662,
+      "step": 12987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.451491355895996,
+      "learning_rate": 8.132309941520468e-06,
+      "loss": 1.5527,
+      "step": 12988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.599709987640381,
+      "learning_rate": 8.13139619883041e-06,
+      "loss": 1.5167,
+      "step": 12989
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4382495880126953,
+      "learning_rate": 8.130482456140351e-06,
+      "loss": 1.4925,
+      "step": 12990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.567168712615967,
+      "learning_rate": 8.129568713450293e-06,
+      "loss": 1.5158,
+      "step": 12991
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.115325450897217,
+      "learning_rate": 8.128654970760235e-06,
+      "loss": 1.15,
+      "step": 12992
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.614117383956909,
+      "learning_rate": 8.127741228070176e-06,
+      "loss": 1.5297,
+      "step": 12993
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9586710929870605,
+      "learning_rate": 8.126827485380118e-06,
+      "loss": 1.62,
+      "step": 12994
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8520593643188477,
+      "learning_rate": 8.125913742690059e-06,
+      "loss": 1.8464,
+      "step": 12995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.585953712463379,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.707,
+      "step": 12996
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.645620107650757,
+      "learning_rate": 8.124086257309942e-06,
+      "loss": 1.4807,
+      "step": 12997
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4690985679626465,
+      "learning_rate": 8.123172514619884e-06,
+      "loss": 1.6235,
+      "step": 12998
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4964747428894043,
+      "learning_rate": 8.122258771929826e-06,
+      "loss": 1.445,
+      "step": 12999
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6671342849731445,
+      "learning_rate": 8.121345029239766e-06,
+      "loss": 1.6424,
+      "step": 13000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6969151496887207,
+      "learning_rate": 8.120431286549709e-06,
+      "loss": 1.2908,
+      "step": 13001
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.532381296157837,
+      "learning_rate": 8.11951754385965e-06,
+      "loss": 1.7473,
+      "step": 13002
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.858764886856079,
+      "learning_rate": 8.118603801169591e-06,
+      "loss": 1.7457,
+      "step": 13003
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.861298680305481,
+      "learning_rate": 8.117690058479532e-06,
+      "loss": 1.1706,
+      "step": 13004
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3478543758392334,
+      "learning_rate": 8.116776315789474e-06,
+      "loss": 1.2895,
+      "step": 13005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5773346424102783,
+      "learning_rate": 8.115862573099416e-06,
+      "loss": 1.354,
+      "step": 13006
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7461864948272705,
+      "learning_rate": 8.114948830409357e-06,
+      "loss": 1.769,
+      "step": 13007
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4292917251586914,
+      "learning_rate": 8.1140350877193e-06,
+      "loss": 1.3279,
+      "step": 13008
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.061241388320923,
+      "learning_rate": 8.11312134502924e-06,
+      "loss": 1.7384,
+      "step": 13009
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9858150482177734,
+      "learning_rate": 8.112207602339182e-06,
+      "loss": 1.3356,
+      "step": 13010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6154024600982666,
+      "learning_rate": 8.111293859649122e-06,
+      "loss": 1.5744,
+      "step": 13011
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.061821460723877,
+      "learning_rate": 8.110380116959065e-06,
+      "loss": 1.6445,
+      "step": 13012
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.710965871810913,
+      "learning_rate": 8.109466374269007e-06,
+      "loss": 1.5347,
+      "step": 13013
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.403761625289917,
+      "learning_rate": 8.108552631578947e-06,
+      "loss": 1.4184,
+      "step": 13014
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.020473003387451,
+      "learning_rate": 8.10763888888889e-06,
+      "loss": 1.6402,
+      "step": 13015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3019402027130127,
+      "learning_rate": 8.10672514619883e-06,
+      "loss": 1.4782,
+      "step": 13016
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.619920492172241,
+      "learning_rate": 8.105811403508772e-06,
+      "loss": 1.6887,
+      "step": 13017
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4248554706573486,
+      "learning_rate": 8.104897660818715e-06,
+      "loss": 1.3228,
+      "step": 13018
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6615819931030273,
+      "learning_rate": 8.103983918128657e-06,
+      "loss": 1.5431,
+      "step": 13019
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.284217119216919,
+      "learning_rate": 8.103070175438597e-06,
+      "loss": 1.3247,
+      "step": 13020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8427557945251465,
+      "learning_rate": 8.10215643274854e-06,
+      "loss": 1.7489,
+      "step": 13021
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.119126796722412,
+      "learning_rate": 8.10124269005848e-06,
+      "loss": 1.2024,
+      "step": 13022
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3520970344543457,
+      "learning_rate": 8.100328947368422e-06,
+      "loss": 1.5475,
+      "step": 13023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.775474786758423,
+      "learning_rate": 8.099415204678363e-06,
+      "loss": 1.5619,
+      "step": 13024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.587817668914795,
+      "learning_rate": 8.098501461988305e-06,
+      "loss": 1.3988,
+      "step": 13025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.34509539604187,
+      "learning_rate": 8.097587719298247e-06,
+      "loss": 1.8384,
+      "step": 13026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4665849208831787,
+      "learning_rate": 8.096673976608188e-06,
+      "loss": 1.5147,
+      "step": 13027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3505704402923584,
+      "learning_rate": 8.09576023391813e-06,
+      "loss": 1.352,
+      "step": 13028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.769794225692749,
+      "learning_rate": 8.09484649122807e-06,
+      "loss": 1.0374,
+      "step": 13029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.942915201187134,
+      "learning_rate": 8.093932748538013e-06,
+      "loss": 1.955,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.468966007232666,
+      "learning_rate": 8.093019005847953e-06,
+      "loss": 1.5965,
+      "step": 13031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.445305109024048,
+      "learning_rate": 8.092105263157896e-06,
+      "loss": 1.404,
+      "step": 13032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0892179012298584,
+      "learning_rate": 8.091191520467838e-06,
+      "loss": 1.1676,
+      "step": 13033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9740660190582275,
+      "learning_rate": 8.090277777777778e-06,
+      "loss": 1.5238,
+      "step": 13034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.871316909790039,
+      "learning_rate": 8.08936403508772e-06,
+      "loss": 1.2301,
+      "step": 13035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8195531368255615,
+      "learning_rate": 8.088450292397661e-06,
+      "loss": 1.7928,
+      "step": 13036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.063410520553589,
+      "learning_rate": 8.087536549707603e-06,
+      "loss": 1.8371,
+      "step": 13037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5672926902770996,
+      "learning_rate": 8.086622807017544e-06,
+      "loss": 1.4269,
+      "step": 13038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.304682731628418,
+      "learning_rate": 8.085709064327486e-06,
+      "loss": 1.2718,
+      "step": 13039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.388928174972534,
+      "learning_rate": 8.084795321637428e-06,
+      "loss": 1.3742,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.250056743621826,
+      "learning_rate": 8.083881578947369e-06,
+      "loss": 1.6034,
+      "step": 13041
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3196609020233154,
+      "learning_rate": 8.082967836257311e-06,
+      "loss": 1.2663,
+      "step": 13042
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.671720504760742,
+      "learning_rate": 8.082054093567252e-06,
+      "loss": 1.5483,
+      "step": 13043
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.258446455001831,
+      "learning_rate": 8.081140350877194e-06,
+      "loss": 2.2867,
+      "step": 13044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.814725160598755,
+      "learning_rate": 8.080226608187134e-06,
+      "loss": 1.8856,
+      "step": 13045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.272162437438965,
+      "learning_rate": 8.079312865497076e-06,
+      "loss": 1.3973,
+      "step": 13046
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.66167950630188,
+      "learning_rate": 8.078399122807019e-06,
+      "loss": 1.6728,
+      "step": 13047
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5663082599639893,
+      "learning_rate": 8.07748538011696e-06,
+      "loss": 1.5974,
+      "step": 13048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9744715690612793,
+      "learning_rate": 8.076571637426901e-06,
+      "loss": 1.6955,
+      "step": 13049
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6237523555755615,
+      "learning_rate": 8.075657894736842e-06,
+      "loss": 1.6538,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0777554512023926,
+      "learning_rate": 8.074744152046784e-06,
+      "loss": 1.2111,
+      "step": 13051
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4740307331085205,
+      "learning_rate": 8.073830409356725e-06,
+      "loss": 1.5985,
+      "step": 13052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9638586044311523,
+      "learning_rate": 8.072916666666667e-06,
+      "loss": 1.9214,
+      "step": 13053
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4892313480377197,
+      "learning_rate": 8.07200292397661e-06,
+      "loss": 1.6184,
+      "step": 13054
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.999319076538086,
+      "learning_rate": 8.07108918128655e-06,
+      "loss": 1.0216,
+      "step": 13055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1085519790649414,
+      "learning_rate": 8.070175438596492e-06,
+      "loss": 1.167,
+      "step": 13056
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0564608573913574,
+      "learning_rate": 8.069261695906432e-06,
+      "loss": 1.1297,
+      "step": 13057
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5557758808135986,
+      "learning_rate": 8.068347953216375e-06,
+      "loss": 1.3945,
+      "step": 13058
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8508429527282715,
+      "learning_rate": 8.067434210526315e-06,
+      "loss": 1.6194,
+      "step": 13059
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7235894203186035,
+      "learning_rate": 8.066520467836257e-06,
+      "loss": 1.769,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9255430698394775,
+      "learning_rate": 8.0656067251462e-06,
+      "loss": 1.7752,
+      "step": 13061
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.715284585952759,
+      "learning_rate": 8.064692982456142e-06,
+      "loss": 1.4949,
+      "step": 13062
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.416264057159424,
+      "learning_rate": 8.063779239766082e-06,
+      "loss": 1.7942,
+      "step": 13063
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.708892822265625,
+      "learning_rate": 8.062865497076025e-06,
+      "loss": 1.5684,
+      "step": 13064
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.896399974822998,
+      "learning_rate": 8.061951754385967e-06,
+      "loss": 1.8145,
+      "step": 13065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6859142780303955,
+      "learning_rate": 8.061038011695907e-06,
+      "loss": 1.6771,
+      "step": 13066
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.979337453842163,
+      "learning_rate": 8.06012426900585e-06,
+      "loss": 1.8695,
+      "step": 13067
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5035316944122314,
+      "learning_rate": 8.05921052631579e-06,
+      "loss": 1.2988,
+      "step": 13068
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.483100175857544,
+      "learning_rate": 8.058296783625732e-06,
+      "loss": 1.3904,
+      "step": 13069
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.485539197921753,
+      "learning_rate": 8.057383040935673e-06,
+      "loss": 1.5327,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8400979042053223,
+      "learning_rate": 8.056469298245615e-06,
+      "loss": 1.5746,
+      "step": 13071
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6015801429748535,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 1.2683,
+      "step": 13072
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8197638988494873,
+      "learning_rate": 8.054641812865498e-06,
+      "loss": 1.5095,
+      "step": 13073
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.341756820678711,
+      "learning_rate": 8.05372807017544e-06,
+      "loss": 1.4552,
+      "step": 13074
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.284184217453003,
+      "learning_rate": 8.05281432748538e-06,
+      "loss": 1.2574,
+      "step": 13075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.753910779953003,
+      "learning_rate": 8.051900584795323e-06,
+      "loss": 1.5523,
+      "step": 13076
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3183538913726807,
+      "learning_rate": 8.050986842105263e-06,
+      "loss": 1.4062,
+      "step": 13077
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.525266408920288,
+      "learning_rate": 8.050073099415206e-06,
+      "loss": 1.7068,
+      "step": 13078
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5570647716522217,
+      "learning_rate": 8.049159356725148e-06,
+      "loss": 1.487,
+      "step": 13079
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2950823307037354,
+      "learning_rate": 8.048245614035088e-06,
+      "loss": 1.0798,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6586248874664307,
+      "learning_rate": 8.04733187134503e-06,
+      "loss": 1.2652,
+      "step": 13081
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2619431018829346,
+      "learning_rate": 8.046418128654971e-06,
+      "loss": 1.294,
+      "step": 13082
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8254520893096924,
+      "learning_rate": 8.045504385964913e-06,
+      "loss": 1.503,
+      "step": 13083
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.829364538192749,
+      "learning_rate": 8.044590643274854e-06,
+      "loss": 1.8618,
+      "step": 13084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.274298667907715,
+      "learning_rate": 8.043676900584796e-06,
+      "loss": 1.2988,
+      "step": 13085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0131123065948486,
+      "learning_rate": 8.042763157894737e-06,
+      "loss": 1.1412,
+      "step": 13086
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5855770111083984,
+      "learning_rate": 8.041849415204679e-06,
+      "loss": 1.7426,
+      "step": 13087
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4689207077026367,
+      "learning_rate": 8.040935672514621e-06,
+      "loss": 1.7215,
+      "step": 13088
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.824525833129883,
+      "learning_rate": 8.040021929824562e-06,
+      "loss": 1.3763,
+      "step": 13089
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4942543506622314,
+      "learning_rate": 8.039108187134504e-06,
+      "loss": 1.5242,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.042415142059326,
+      "learning_rate": 8.038194444444444e-06,
+      "loss": 1.222,
+      "step": 13091
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.670363426208496,
+      "learning_rate": 8.037280701754386e-06,
+      "loss": 1.5558,
+      "step": 13092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9370903968811035,
+      "learning_rate": 8.036366959064327e-06,
+      "loss": 1.8204,
+      "step": 13093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.595710277557373,
+      "learning_rate": 8.03545321637427e-06,
+      "loss": 2.2897,
+      "step": 13094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.091703414916992,
+      "learning_rate": 8.034539473684211e-06,
+      "loss": 1.7819,
+      "step": 13095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2959084510803223,
+      "learning_rate": 8.033625730994152e-06,
+      "loss": 1.2336,
+      "step": 13096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2136099338531494,
+      "learning_rate": 8.032711988304094e-06,
+      "loss": 1.3042,
+      "step": 13097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.699662446975708,
+      "learning_rate": 8.031798245614035e-06,
+      "loss": 1.3926,
+      "step": 13098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.279503345489502,
+      "learning_rate": 8.030884502923977e-06,
+      "loss": 1.4766,
+      "step": 13099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5216238498687744,
+      "learning_rate": 8.029970760233917e-06,
+      "loss": 1.6458,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5716850757598877,
+      "learning_rate": 8.02905701754386e-06,
+      "loss": 1.3741,
+      "step": 13101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7761409282684326,
+      "learning_rate": 8.028143274853802e-06,
+      "loss": 1.894,
+      "step": 13102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5264244079589844,
+      "learning_rate": 8.027229532163744e-06,
+      "loss": 1.6395,
+      "step": 13103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6863162517547607,
+      "learning_rate": 8.026315789473685e-06,
+      "loss": 1.5294,
+      "step": 13104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2278401851654053,
+      "learning_rate": 8.025402046783627e-06,
+      "loss": 1.2034,
+      "step": 13105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.038137674331665,
+      "learning_rate": 8.024488304093569e-06,
+      "loss": 1.4047,
+      "step": 13106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9347128868103027,
+      "learning_rate": 8.02357456140351e-06,
+      "loss": 1.2133,
+      "step": 13107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7971625328063965,
+      "learning_rate": 8.022660818713452e-06,
+      "loss": 1.5598,
+      "step": 13108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4059665203094482,
+      "learning_rate": 8.021747076023392e-06,
+      "loss": 1.3847,
+      "step": 13109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7304725646972656,
+      "learning_rate": 8.020833333333335e-06,
+      "loss": 1.4103,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.321918487548828,
+      "learning_rate": 8.019919590643275e-06,
+      "loss": 1.4203,
+      "step": 13111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.422335624694824,
+      "learning_rate": 8.019005847953217e-06,
+      "loss": 1.4309,
+      "step": 13112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.082444190979004,
+      "learning_rate": 8.01809210526316e-06,
+      "loss": 1.3463,
+      "step": 13113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.383012294769287,
+      "learning_rate": 8.0171783625731e-06,
+      "loss": 1.5244,
+      "step": 13114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0220589637756348,
+      "learning_rate": 8.016264619883042e-06,
+      "loss": 1.2893,
+      "step": 13115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7068161964416504,
+      "learning_rate": 8.015350877192983e-06,
+      "loss": 1.5544,
+      "step": 13116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.135524034500122,
+      "learning_rate": 8.014437134502925e-06,
+      "loss": 1.6603,
+      "step": 13117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.46549129486084,
+      "learning_rate": 8.013523391812866e-06,
+      "loss": 1.6795,
+      "step": 13118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3274998664855957,
+      "learning_rate": 8.012609649122808e-06,
+      "loss": 1.2587,
+      "step": 13119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.806034564971924,
+      "learning_rate": 8.01169590643275e-06,
+      "loss": 1.688,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.54091477394104,
+      "learning_rate": 8.01078216374269e-06,
+      "loss": 1.6826,
+      "step": 13121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5460073947906494,
+      "learning_rate": 8.009868421052633e-06,
+      "loss": 1.3681,
+      "step": 13122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.763768434524536,
+      "learning_rate": 8.008954678362573e-06,
+      "loss": 1.5041,
+      "step": 13123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1546823978424072,
+      "learning_rate": 8.008040935672516e-06,
+      "loss": 1.2866,
+      "step": 13124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.128286361694336,
+      "learning_rate": 8.007127192982456e-06,
+      "loss": 1.1672,
+      "step": 13125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1062235832214355,
+      "learning_rate": 8.006213450292398e-06,
+      "loss": 1.8218,
+      "step": 13126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.221019983291626,
+      "learning_rate": 8.00529970760234e-06,
+      "loss": 1.7845,
+      "step": 13127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.442172050476074,
+      "learning_rate": 8.004385964912281e-06,
+      "loss": 1.1671,
+      "step": 13128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6500015258789062,
+      "learning_rate": 8.003472222222223e-06,
+      "loss": 1.5729,
+      "step": 13129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.834575653076172,
+      "learning_rate": 8.002558479532164e-06,
+      "loss": 1.4396,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.534433603286743,
+      "learning_rate": 8.001644736842106e-06,
+      "loss": 1.1757,
+      "step": 13131
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.279629707336426,
+      "learning_rate": 8.000730994152047e-06,
+      "loss": 1.5816,
+      "step": 13132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4612362384796143,
+      "learning_rate": 7.999817251461989e-06,
+      "loss": 1.5012,
+      "step": 13133
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8424079418182373,
+      "learning_rate": 7.998903508771931e-06,
+      "loss": 1.6177,
+      "step": 13134
+    },
+    {
+      "epoch": 0.6,
+      "eval_train_bleu": 0.23059921478222575,
+      "eval_train_loss": 2.076962947845459,
+      "eval_train_runtime": 19.8586,
+      "eval_train_samples_per_second": 50.406,
+      "eval_train_steps_per_second": 1.611,
+      "step": 13134
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7147486209869385,
+      "learning_rate": 7.997989766081872e-06,
+      "loss": 1.6036,
+      "step": 13135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1220431327819824,
+      "learning_rate": 7.997076023391814e-06,
+      "loss": 1.3683,
+      "step": 13136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6677980422973633,
+      "learning_rate": 7.996162280701754e-06,
+      "loss": 1.677,
+      "step": 13137
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.211714744567871,
+      "learning_rate": 7.995248538011696e-06,
+      "loss": 1.4983,
+      "step": 13138
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2265732288360596,
+      "learning_rate": 7.994334795321637e-06,
+      "loss": 1.2256,
+      "step": 13139
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0723342895507812,
+      "learning_rate": 7.99342105263158e-06,
+      "loss": 1.1946,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.808145761489868,
+      "learning_rate": 7.992507309941521e-06,
+      "loss": 1.4897,
+      "step": 13141
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.636498212814331,
+      "learning_rate": 7.991593567251462e-06,
+      "loss": 1.5467,
+      "step": 13142
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.419830799102783,
+      "learning_rate": 7.990679824561404e-06,
+      "loss": 1.3152,
+      "step": 13143
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5129406452178955,
+      "learning_rate": 7.989766081871345e-06,
+      "loss": 1.3974,
+      "step": 13144
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.251268148422241,
+      "learning_rate": 7.988852339181287e-06,
+      "loss": 1.3542,
+      "step": 13145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4486749172210693,
+      "learning_rate": 7.98793859649123e-06,
+      "loss": 1.2854,
+      "step": 13146
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0512466430664062,
+      "learning_rate": 7.987024853801171e-06,
+      "loss": 1.7428,
+      "step": 13147
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8440322875976562,
+      "learning_rate": 7.986111111111112e-06,
+      "loss": 1.4828,
+      "step": 13148
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.525459051132202,
+      "learning_rate": 7.985197368421054e-06,
+      "loss": 1.2833,
+      "step": 13149
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7785121202468872,
+      "learning_rate": 7.984283625730995e-06,
+      "loss": 1.193,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6464743614196777,
+      "learning_rate": 7.983369883040937e-06,
+      "loss": 1.5852,
+      "step": 13151
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.331089496612549,
+      "learning_rate": 7.982456140350877e-06,
+      "loss": 1.1056,
+      "step": 13152
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6229987144470215,
+      "learning_rate": 7.98154239766082e-06,
+      "loss": 1.5595,
+      "step": 13153
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.584484577178955,
+      "learning_rate": 7.980628654970762e-06,
+      "loss": 1.414,
+      "step": 13154
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5325119495391846,
+      "learning_rate": 7.979714912280702e-06,
+      "loss": 1.4608,
+      "step": 13155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3276851177215576,
+      "learning_rate": 7.978801169590645e-06,
+      "loss": 1.1278,
+      "step": 13156
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.003124237060547,
+      "learning_rate": 7.977887426900585e-06,
+      "loss": 1.5526,
+      "step": 13157
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.023580551147461,
+      "learning_rate": 7.976973684210527e-06,
+      "loss": 1.2635,
+      "step": 13158
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.230120897293091,
+      "learning_rate": 7.976059941520468e-06,
+      "loss": 1.1953,
+      "step": 13159
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.273292303085327,
+      "learning_rate": 7.97514619883041e-06,
+      "loss": 1.4098,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3984501361846924,
+      "learning_rate": 7.974232456140352e-06,
+      "loss": 1.4823,
+      "step": 13161
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5746445655822754,
+      "learning_rate": 7.973318713450293e-06,
+      "loss": 1.7397,
+      "step": 13162
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.547208070755005,
+      "learning_rate": 7.972404970760235e-06,
+      "loss": 1.5615,
+      "step": 13163
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.774458885192871,
+      "learning_rate": 7.971491228070176e-06,
+      "loss": 1.4689,
+      "step": 13164
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.790109395980835,
+      "learning_rate": 7.970577485380118e-06,
+      "loss": 1.6031,
+      "step": 13165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.577390193939209,
+      "learning_rate": 7.969663742690058e-06,
+      "loss": 1.6312,
+      "step": 13166
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2973196506500244,
+      "learning_rate": 7.96875e-06,
+      "loss": 1.5758,
+      "step": 13167
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7401106357574463,
+      "learning_rate": 7.967836257309943e-06,
+      "loss": 1.199,
+      "step": 13168
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.736387014389038,
+      "learning_rate": 7.966922514619883e-06,
+      "loss": 1.6356,
+      "step": 13169
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.291210174560547,
+      "learning_rate": 7.966008771929826e-06,
+      "loss": 1.3861,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4428865909576416,
+      "learning_rate": 7.965095029239766e-06,
+      "loss": 1.4564,
+      "step": 13171
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3226587772369385,
+      "learning_rate": 7.964181286549708e-06,
+      "loss": 1.3516,
+      "step": 13172
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.998870849609375,
+      "learning_rate": 7.963267543859649e-06,
+      "loss": 1.4048,
+      "step": 13173
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7076914310455322,
+      "learning_rate": 7.962353801169591e-06,
+      "loss": 1.4361,
+      "step": 13174
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1538753509521484,
+      "learning_rate": 7.961440058479533e-06,
+      "loss": 1.9505,
+      "step": 13175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.750824451446533,
+      "learning_rate": 7.960526315789474e-06,
+      "loss": 1.8151,
+      "step": 13176
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.455643892288208,
+      "learning_rate": 7.959612573099416e-06,
+      "loss": 1.3553,
+      "step": 13177
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.223245859146118,
+      "learning_rate": 7.958698830409357e-06,
+      "loss": 1.1443,
+      "step": 13178
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.603794574737549,
+      "learning_rate": 7.957785087719299e-06,
+      "loss": 1.3368,
+      "step": 13179
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.513831853866577,
+      "learning_rate": 7.95687134502924e-06,
+      "loss": 1.3386,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5693278312683105,
+      "learning_rate": 7.955957602339182e-06,
+      "loss": 1.4068,
+      "step": 13181
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7854208946228027,
+      "learning_rate": 7.955043859649124e-06,
+      "loss": 1.6931,
+      "step": 13182
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.632650852203369,
+      "learning_rate": 7.954130116959064e-06,
+      "loss": 1.7215,
+      "step": 13183
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.983959436416626,
+      "learning_rate": 7.953216374269006e-06,
+      "loss": 1.6328,
+      "step": 13184
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7032530307769775,
+      "learning_rate": 7.952302631578947e-06,
+      "loss": 1.4995,
+      "step": 13185
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.60844349861145,
+      "learning_rate": 7.95138888888889e-06,
+      "loss": 1.2401,
+      "step": 13186
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.814687490463257,
+      "learning_rate": 7.95047514619883e-06,
+      "loss": 1.549,
+      "step": 13187
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7218258380889893,
+      "learning_rate": 7.949561403508772e-06,
+      "loss": 1.3074,
+      "step": 13188
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.936203956604004,
+      "learning_rate": 7.948647660818714e-06,
+      "loss": 1.4346,
+      "step": 13189
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.365806818008423,
+      "learning_rate": 7.947733918128656e-06,
+      "loss": 1.3033,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.298696279525757,
+      "learning_rate": 7.946820175438597e-06,
+      "loss": 1.4764,
+      "step": 13191
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.263077974319458,
+      "learning_rate": 7.94590643274854e-06,
+      "loss": 1.627,
+      "step": 13192
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0670759677886963,
+      "learning_rate": 7.944992690058481e-06,
+      "loss": 1.2853,
+      "step": 13193
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.149994373321533,
+      "learning_rate": 7.944078947368422e-06,
+      "loss": 1.6229,
+      "step": 13194
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6987996101379395,
+      "learning_rate": 7.943165204678364e-06,
+      "loss": 1.6488,
+      "step": 13195
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.196106433868408,
+      "learning_rate": 7.942251461988305e-06,
+      "loss": 1.6036,
+      "step": 13196
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.940239667892456,
+      "learning_rate": 7.941337719298247e-06,
+      "loss": 1.8368,
+      "step": 13197
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3983712196350098,
+      "learning_rate": 7.940423976608187e-06,
+      "loss": 1.4933,
+      "step": 13198
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.88041615486145,
+      "learning_rate": 7.93951023391813e-06,
+      "loss": 1.5971,
+      "step": 13199
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5585150718688965,
+      "learning_rate": 7.938596491228072e-06,
+      "loss": 1.2875,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9751434326171875,
+      "learning_rate": 7.937682748538012e-06,
+      "loss": 1.794,
+      "step": 13201
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.742849588394165,
+      "learning_rate": 7.936769005847955e-06,
+      "loss": 1.6083,
+      "step": 13202
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2368366718292236,
+      "learning_rate": 7.935855263157895e-06,
+      "loss": 1.4315,
+      "step": 13203
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.693474531173706,
+      "learning_rate": 7.934941520467837e-06,
+      "loss": 1.5128,
+      "step": 13204
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6412973403930664,
+      "learning_rate": 7.934027777777778e-06,
+      "loss": 1.6941,
+      "step": 13205
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3320889472961426,
+      "learning_rate": 7.93311403508772e-06,
+      "loss": 1.3596,
+      "step": 13206
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4448914527893066,
+      "learning_rate": 7.932200292397662e-06,
+      "loss": 1.8731,
+      "step": 13207
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.767507314682007,
+      "learning_rate": 7.931286549707603e-06,
+      "loss": 1.5604,
+      "step": 13208
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3433754444122314,
+      "learning_rate": 7.930372807017545e-06,
+      "loss": 1.6446,
+      "step": 13209
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4337165355682373,
+      "learning_rate": 7.929459064327486e-06,
+      "loss": 1.1819,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2909843921661377,
+      "learning_rate": 7.928545321637428e-06,
+      "loss": 1.4453,
+      "step": 13211
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5485427379608154,
+      "learning_rate": 7.927631578947368e-06,
+      "loss": 1.6035,
+      "step": 13212
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.774784564971924,
+      "learning_rate": 7.92671783625731e-06,
+      "loss": 1.4063,
+      "step": 13213
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0027310848236084,
+      "learning_rate": 7.925804093567253e-06,
+      "loss": 1.6407,
+      "step": 13214
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6098573207855225,
+      "learning_rate": 7.924890350877193e-06,
+      "loss": 1.514,
+      "step": 13215
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6573281288146973,
+      "learning_rate": 7.923976608187136e-06,
+      "loss": 1.2848,
+      "step": 13216
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3593761920928955,
+      "learning_rate": 7.923062865497076e-06,
+      "loss": 1.4058,
+      "step": 13217
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3210232257843018,
+      "learning_rate": 7.922149122807018e-06,
+      "loss": 1.4875,
+      "step": 13218
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.634173631668091,
+      "learning_rate": 7.921235380116959e-06,
+      "loss": 1.7642,
+      "step": 13219
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8062405586242676,
+      "learning_rate": 7.920321637426901e-06,
+      "loss": 1.2382,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.913170099258423,
+      "learning_rate": 7.919407894736843e-06,
+      "loss": 1.7603,
+      "step": 13221
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.537994146347046,
+      "learning_rate": 7.918494152046784e-06,
+      "loss": 1.7046,
+      "step": 13222
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6215760707855225,
+      "learning_rate": 7.917580409356726e-06,
+      "loss": 1.3558,
+      "step": 13223
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2894749641418457,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 1.3137,
+      "step": 13224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0244929790496826,
+      "learning_rate": 7.915752923976609e-06,
+      "loss": 1.9278,
+      "step": 13225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1287014484405518,
+      "learning_rate": 7.91483918128655e-06,
+      "loss": 1.3539,
+      "step": 13226
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.734849452972412,
+      "learning_rate": 7.913925438596492e-06,
+      "loss": 1.7521,
+      "step": 13227
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.463876247406006,
+      "learning_rate": 7.913011695906434e-06,
+      "loss": 1.4303,
+      "step": 13228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.136061429977417,
+      "learning_rate": 7.912097953216374e-06,
+      "loss": 1.4045,
+      "step": 13229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.639920473098755,
+      "learning_rate": 7.911184210526316e-06,
+      "loss": 1.5283,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6751644611358643,
+      "learning_rate": 7.910270467836257e-06,
+      "loss": 1.3408,
+      "step": 13231
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.015623092651367,
+      "learning_rate": 7.9093567251462e-06,
+      "loss": 1.6741,
+      "step": 13232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.655677556991577,
+      "learning_rate": 7.908442982456141e-06,
+      "loss": 1.6316,
+      "step": 13233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.493069648742676,
+      "learning_rate": 7.907529239766084e-06,
+      "loss": 1.4613,
+      "step": 13234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2471351623535156,
+      "learning_rate": 7.906615497076024e-06,
+      "loss": 1.2799,
+      "step": 13235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.504636764526367,
+      "learning_rate": 7.905701754385966e-06,
+      "loss": 1.5381,
+      "step": 13236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.665994167327881,
+      "learning_rate": 7.904788011695907e-06,
+      "loss": 1.4426,
+      "step": 13237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.910128116607666,
+      "learning_rate": 7.90387426900585e-06,
+      "loss": 1.5932,
+      "step": 13238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5235495567321777,
+      "learning_rate": 7.90296052631579e-06,
+      "loss": 1.1198,
+      "step": 13239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4432365894317627,
+      "learning_rate": 7.902046783625732e-06,
+      "loss": 1.3722,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4308881759643555,
+      "learning_rate": 7.901133040935674e-06,
+      "loss": 1.5208,
+      "step": 13241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.158963203430176,
+      "learning_rate": 7.900219298245615e-06,
+      "loss": 1.3511,
+      "step": 13242
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.860445022583008,
+      "learning_rate": 7.899305555555557e-06,
+      "loss": 1.6002,
+      "step": 13243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7625057697296143,
+      "learning_rate": 7.898391812865497e-06,
+      "loss": 1.8146,
+      "step": 13244
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.172497510910034,
+      "learning_rate": 7.89747807017544e-06,
+      "loss": 1.4523,
+      "step": 13245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.700942277908325,
+      "learning_rate": 7.89656432748538e-06,
+      "loss": 1.6766,
+      "step": 13246
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3701398372650146,
+      "learning_rate": 7.895650584795322e-06,
+      "loss": 1.3608,
+      "step": 13247
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6738080978393555,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 1.496,
+      "step": 13248
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6232101917266846,
+      "learning_rate": 7.893823099415205e-06,
+      "loss": 1.5677,
+      "step": 13249
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7446444034576416,
+      "learning_rate": 7.892909356725147e-06,
+      "loss": 1.6532,
+      "step": 13250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3976316452026367,
+      "learning_rate": 7.891995614035088e-06,
+      "loss": 1.797,
+      "step": 13251
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1338295936584473,
+      "learning_rate": 7.89108187134503e-06,
+      "loss": 1.383,
+      "step": 13252
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9626981019973755,
+      "learning_rate": 7.89016812865497e-06,
+      "loss": 1.0677,
+      "step": 13253
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.869813919067383,
+      "learning_rate": 7.889254385964913e-06,
+      "loss": 1.642,
+      "step": 13254
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9586894512176514,
+      "learning_rate": 7.888340643274855e-06,
+      "loss": 1.7007,
+      "step": 13255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1490862369537354,
+      "learning_rate": 7.887426900584796e-06,
+      "loss": 1.5607,
+      "step": 13256
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1575958728790283,
+      "learning_rate": 7.886513157894738e-06,
+      "loss": 2.1059,
+      "step": 13257
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3647520542144775,
+      "learning_rate": 7.885599415204678e-06,
+      "loss": 1.4466,
+      "step": 13258
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7670304775238037,
+      "learning_rate": 7.88468567251462e-06,
+      "loss": 1.3178,
+      "step": 13259
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.215308904647827,
+      "learning_rate": 7.883771929824561e-06,
+      "loss": 1.7711,
+      "step": 13260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.571049690246582,
+      "learning_rate": 7.882858187134503e-06,
+      "loss": 1.3053,
+      "step": 13261
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.881242036819458,
+      "learning_rate": 7.881944444444446e-06,
+      "loss": 1.503,
+      "step": 13262
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.736750602722168,
+      "learning_rate": 7.881030701754386e-06,
+      "loss": 1.8638,
+      "step": 13263
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6673624515533447,
+      "learning_rate": 7.880116959064328e-06,
+      "loss": 1.6246,
+      "step": 13264
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0917932987213135,
+      "learning_rate": 7.879203216374269e-06,
+      "loss": 1.68,
+      "step": 13265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3632349967956543,
+      "learning_rate": 7.878289473684211e-06,
+      "loss": 1.5017,
+      "step": 13266
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3861629962921143,
+      "learning_rate": 7.877375730994152e-06,
+      "loss": 1.305,
+      "step": 13267
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.219377279281616,
+      "learning_rate": 7.876461988304094e-06,
+      "loss": 1.405,
+      "step": 13268
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7659964561462402,
+      "learning_rate": 7.875548245614036e-06,
+      "loss": 1.6107,
+      "step": 13269
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3724098205566406,
+      "learning_rate": 7.874634502923977e-06,
+      "loss": 1.4715,
+      "step": 13270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3248555660247803,
+      "learning_rate": 7.873720760233919e-06,
+      "loss": 1.3206,
+      "step": 13271
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.309037446975708,
+      "learning_rate": 7.87280701754386e-06,
+      "loss": 1.3786,
+      "step": 13272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4863193035125732,
+      "learning_rate": 7.871893274853802e-06,
+      "loss": 1.5287,
+      "step": 13273
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3463125228881836,
+      "learning_rate": 7.870979532163744e-06,
+      "loss": 1.3747,
+      "step": 13274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.744783639907837,
+      "learning_rate": 7.870065789473686e-06,
+      "loss": 1.7603,
+      "step": 13275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.562101364135742,
+      "learning_rate": 7.869152046783626e-06,
+      "loss": 1.6278,
+      "step": 13276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1031711101531982,
+      "learning_rate": 7.868238304093569e-06,
+      "loss": 1.1838,
+      "step": 13277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8708407878875732,
+      "learning_rate": 7.86732456140351e-06,
+      "loss": 1.4526,
+      "step": 13278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3951728343963623,
+      "learning_rate": 7.866410818713451e-06,
+      "loss": 1.7752,
+      "step": 13279
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.321859121322632,
+      "learning_rate": 7.865497076023394e-06,
+      "loss": 1.1945,
+      "step": 13280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1232383251190186,
+      "learning_rate": 7.864583333333334e-06,
+      "loss": 1.3505,
+      "step": 13281
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.417556047439575,
+      "learning_rate": 7.863669590643276e-06,
+      "loss": 1.3487,
+      "step": 13282
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4318768978118896,
+      "learning_rate": 7.862755847953217e-06,
+      "loss": 1.4171,
+      "step": 13283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.314943790435791,
+      "learning_rate": 7.86184210526316e-06,
+      "loss": 1.4326,
+      "step": 13284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4189200401306152,
+      "learning_rate": 7.8609283625731e-06,
+      "loss": 1.2414,
+      "step": 13285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3962223529815674,
+      "learning_rate": 7.860014619883042e-06,
+      "loss": 1.4119,
+      "step": 13286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.29184627532959,
+      "learning_rate": 7.859100877192984e-06,
+      "loss": 1.6928,
+      "step": 13287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.562196969985962,
+      "learning_rate": 7.858187134502925e-06,
+      "loss": 1.5283,
+      "step": 13288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.797828435897827,
+      "learning_rate": 7.857273391812867e-06,
+      "loss": 1.5457,
+      "step": 13289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8413703441619873,
+      "learning_rate": 7.856359649122807e-06,
+      "loss": 1.4235,
+      "step": 13290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3615689277648926,
+      "learning_rate": 7.85544590643275e-06,
+      "loss": 1.3696,
+      "step": 13291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.670795440673828,
+      "learning_rate": 7.85453216374269e-06,
+      "loss": 1.5296,
+      "step": 13292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7207629680633545,
+      "learning_rate": 7.853618421052632e-06,
+      "loss": 1.4741,
+      "step": 13293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.574246406555176,
+      "learning_rate": 7.852704678362575e-06,
+      "loss": 1.6409,
+      "step": 13294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.536627769470215,
+      "learning_rate": 7.851790935672515e-06,
+      "loss": 1.4044,
+      "step": 13295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8425819873809814,
+      "learning_rate": 7.850877192982457e-06,
+      "loss": 1.9429,
+      "step": 13296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.835070848464966,
+      "learning_rate": 7.849963450292398e-06,
+      "loss": 1.5213,
+      "step": 13297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1225335597991943,
+      "learning_rate": 7.84904970760234e-06,
+      "loss": 1.9478,
+      "step": 13298
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.526285171508789,
+      "learning_rate": 7.84813596491228e-06,
+      "loss": 1.2332,
+      "step": 13299
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.440741777420044,
+      "learning_rate": 7.847222222222223e-06,
+      "loss": 1.2972,
+      "step": 13300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.024587392807007,
+      "learning_rate": 7.846308479532165e-06,
+      "loss": 1.3795,
+      "step": 13301
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.470673084259033,
+      "learning_rate": 7.845394736842106e-06,
+      "loss": 1.78,
+      "step": 13302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.617284059524536,
+      "learning_rate": 7.844480994152048e-06,
+      "loss": 1.5895,
+      "step": 13303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.593942165374756,
+      "learning_rate": 7.843567251461988e-06,
+      "loss": 1.5355,
+      "step": 13304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.012756824493408,
+      "learning_rate": 7.84265350877193e-06,
+      "loss": 1.5089,
+      "step": 13305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8026328086853027,
+      "learning_rate": 7.841739766081871e-06,
+      "loss": 1.6496,
+      "step": 13306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0552315711975098,
+      "learning_rate": 7.840826023391813e-06,
+      "loss": 1.7401,
+      "step": 13307
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.283633232116699,
+      "learning_rate": 7.839912280701756e-06,
+      "loss": 1.3598,
+      "step": 13308
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8759188652038574,
+      "learning_rate": 7.838998538011696e-06,
+      "loss": 1.6616,
+      "step": 13309
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.069528341293335,
+      "learning_rate": 7.838084795321638e-06,
+      "loss": 1.7864,
+      "step": 13310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8450019359588623,
+      "learning_rate": 7.837171052631579e-06,
+      "loss": 1.0183,
+      "step": 13311
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.317385673522949,
+      "learning_rate": 7.836257309941521e-06,
+      "loss": 1.3573,
+      "step": 13312
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.46630859375,
+      "learning_rate": 7.835343567251462e-06,
+      "loss": 1.3164,
+      "step": 13313
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.541166305541992,
+      "learning_rate": 7.834429824561404e-06,
+      "loss": 1.4967,
+      "step": 13314
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5025577545166016,
+      "learning_rate": 7.833516081871344e-06,
+      "loss": 1.4936,
+      "step": 13315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9494967460632324,
+      "learning_rate": 7.832602339181287e-06,
+      "loss": 1.5903,
+      "step": 13316
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.564540147781372,
+      "learning_rate": 7.831688596491229e-06,
+      "loss": 1.7864,
+      "step": 13317
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7732350826263428,
+      "learning_rate": 7.830774853801171e-06,
+      "loss": 1.6023,
+      "step": 13318
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.523068428039551,
+      "learning_rate": 7.829861111111112e-06,
+      "loss": 1.378,
+      "step": 13319
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.574772357940674,
+      "learning_rate": 7.828947368421054e-06,
+      "loss": 1.4834,
+      "step": 13320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.859374761581421,
+      "learning_rate": 7.828033625730996e-06,
+      "loss": 1.4996,
+      "step": 13321
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.32635498046875,
+      "learning_rate": 7.827119883040936e-06,
+      "loss": 1.2969,
+      "step": 13322
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6638307571411133,
+      "learning_rate": 7.826206140350879e-06,
+      "loss": 1.5002,
+      "step": 13323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.017390489578247,
+      "learning_rate": 7.82529239766082e-06,
+      "loss": 1.0401,
+      "step": 13324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6219942569732666,
+      "learning_rate": 7.824378654970761e-06,
+      "loss": 1.6479,
+      "step": 13325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.445406675338745,
+      "learning_rate": 7.823464912280702e-06,
+      "loss": 1.6048,
+      "step": 13326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.917067766189575,
+      "learning_rate": 7.822551169590644e-06,
+      "loss": 1.6214,
+      "step": 13327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6662797927856445,
+      "learning_rate": 7.821637426900586e-06,
+      "loss": 1.5399,
+      "step": 13328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.340254306793213,
+      "learning_rate": 7.820723684210527e-06,
+      "loss": 1.3986,
+      "step": 13329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1759021282196045,
+      "learning_rate": 7.81980994152047e-06,
+      "loss": 1.2048,
+      "step": 13330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.566722869873047,
+      "learning_rate": 7.81889619883041e-06,
+      "loss": 1.6673,
+      "step": 13331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2229084968566895,
+      "learning_rate": 7.817982456140352e-06,
+      "loss": 1.2682,
+      "step": 13332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9074347019195557,
+      "learning_rate": 7.817068713450292e-06,
+      "loss": 1.7531,
+      "step": 13333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.041098117828369,
+      "learning_rate": 7.816154970760235e-06,
+      "loss": 1.1827,
+      "step": 13334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4935460090637207,
+      "learning_rate": 7.815241228070177e-06,
+      "loss": 1.5219,
+      "step": 13335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8601155281066895,
+      "learning_rate": 7.814327485380117e-06,
+      "loss": 1.6276,
+      "step": 13336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3675074577331543,
+      "learning_rate": 7.81341374269006e-06,
+      "loss": 2.0144,
+      "step": 13337
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.065570592880249,
+      "learning_rate": 7.8125e-06,
+      "loss": 1.0785,
+      "step": 13338
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.144282579421997,
+      "learning_rate": 7.811586257309942e-06,
+      "loss": 1.8962,
+      "step": 13339
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.291027307510376,
+      "learning_rate": 7.810672514619883e-06,
+      "loss": 1.7739,
+      "step": 13340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.651411294937134,
+      "learning_rate": 7.809758771929825e-06,
+      "loss": 1.421,
+      "step": 13341
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6386425495147705,
+      "learning_rate": 7.808845029239767e-06,
+      "loss": 1.7402,
+      "step": 13342
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.707794666290283,
+      "learning_rate": 7.807931286549708e-06,
+      "loss": 1.6878,
+      "step": 13343
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5131747722625732,
+      "learning_rate": 7.80701754385965e-06,
+      "loss": 1.2304,
+      "step": 13344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2469515800476074,
+      "learning_rate": 7.80610380116959e-06,
+      "loss": 1.422,
+      "step": 13345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5352706909179688,
+      "learning_rate": 7.805190058479533e-06,
+      "loss": 1.6531,
+      "step": 13346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.601389169692993,
+      "learning_rate": 7.804276315789473e-06,
+      "loss": 1.6214,
+      "step": 13347
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.733415126800537,
+      "learning_rate": 7.803362573099416e-06,
+      "loss": 1.6731,
+      "step": 13348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.151634693145752,
+      "learning_rate": 7.802448830409358e-06,
+      "loss": 1.4536,
+      "step": 13349
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4131598472595215,
+      "learning_rate": 7.801535087719298e-06,
+      "loss": 1.1156,
+      "step": 13350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.919811248779297,
+      "learning_rate": 7.80062134502924e-06,
+      "loss": 2.0047,
+      "step": 13351
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5931410789489746,
+      "learning_rate": 7.799707602339181e-06,
+      "loss": 1.5091,
+      "step": 13352
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.453287363052368,
+      "learning_rate": 7.798793859649123e-06,
+      "loss": 1.4188,
+      "step": 13353
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4502065181732178,
+      "learning_rate": 7.797880116959064e-06,
+      "loss": 1.3482,
+      "step": 13354
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9079267978668213,
+      "learning_rate": 7.796966374269006e-06,
+      "loss": 1.421,
+      "step": 13355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4933812618255615,
+      "learning_rate": 7.796052631578948e-06,
+      "loss": 1.6286,
+      "step": 13356
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7511038780212402,
+      "learning_rate": 7.795138888888889e-06,
+      "loss": 1.6741,
+      "step": 13357
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.536208391189575,
+      "learning_rate": 7.794225146198831e-06,
+      "loss": 1.4993,
+      "step": 13358
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5688576698303223,
+      "learning_rate": 7.793311403508772e-06,
+      "loss": 1.1821,
+      "step": 13359
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.766169786453247,
+      "learning_rate": 7.792397660818714e-06,
+      "loss": 1.7308,
+      "step": 13360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.745168685913086,
+      "learning_rate": 7.791483918128656e-06,
+      "loss": 1.7312,
+      "step": 13361
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7272634506225586,
+      "learning_rate": 7.790570175438598e-06,
+      "loss": 1.763,
+      "step": 13362
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9750386476516724,
+      "learning_rate": 7.789656432748539e-06,
+      "loss": 1.1356,
+      "step": 13363
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.753966808319092,
+      "learning_rate": 7.788742690058481e-06,
+      "loss": 1.415,
+      "step": 13364
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9785752296447754,
+      "learning_rate": 7.787828947368422e-06,
+      "loss": 1.8521,
+      "step": 13365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.250532627105713,
+      "learning_rate": 7.786915204678364e-06,
+      "loss": 1.5679,
+      "step": 13366
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.778412342071533,
+      "learning_rate": 7.786001461988304e-06,
+      "loss": 1.5238,
+      "step": 13367
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.021334171295166,
+      "learning_rate": 7.785087719298246e-06,
+      "loss": 1.2816,
+      "step": 13368
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.751358985900879,
+      "learning_rate": 7.784173976608189e-06,
+      "loss": 1.9615,
+      "step": 13369
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7964367866516113,
+      "learning_rate": 7.78326023391813e-06,
+      "loss": 1.4846,
+      "step": 13370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5796046257019043,
+      "learning_rate": 7.782346491228071e-06,
+      "loss": 1.476,
+      "step": 13371
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2055044174194336,
+      "learning_rate": 7.781432748538012e-06,
+      "loss": 1.2829,
+      "step": 13372
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1415984630584717,
+      "learning_rate": 7.780519005847954e-06,
+      "loss": 1.7569,
+      "step": 13373
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.580296277999878,
+      "learning_rate": 7.779605263157895e-06,
+      "loss": 1.6216,
+      "step": 13374
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.426955461502075,
+      "learning_rate": 7.778691520467837e-06,
+      "loss": 1.8135,
+      "step": 13375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0630125999450684,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.1641,
+      "step": 13376
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.613093376159668,
+      "learning_rate": 7.77686403508772e-06,
+      "loss": 1.6057,
+      "step": 13377
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6670584678649902,
+      "learning_rate": 7.775950292397662e-06,
+      "loss": 1.6531,
+      "step": 13378
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9931609630584717,
+      "learning_rate": 7.775036549707602e-06,
+      "loss": 1.4255,
+      "step": 13379
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.865770101547241,
+      "learning_rate": 7.774122807017545e-06,
+      "loss": 1.2143,
+      "step": 13380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.715287923812866,
+      "learning_rate": 7.773209064327485e-06,
+      "loss": 1.5531,
+      "step": 13381
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.685127019882202,
+      "learning_rate": 7.772295321637427e-06,
+      "loss": 1.2842,
+      "step": 13382
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0995445251464844,
+      "learning_rate": 7.77138157894737e-06,
+      "loss": 1.2541,
+      "step": 13383
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1807680130004883,
+      "learning_rate": 7.77046783625731e-06,
+      "loss": 1.4915,
+      "step": 13384
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.068357229232788,
+      "learning_rate": 7.769554093567252e-06,
+      "loss": 1.2303,
+      "step": 13385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.578348398208618,
+      "learning_rate": 7.768640350877193e-06,
+      "loss": 1.6827,
+      "step": 13386
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.252218008041382,
+      "learning_rate": 7.767726608187135e-06,
+      "loss": 1.4361,
+      "step": 13387
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.391214609146118,
+      "learning_rate": 7.766812865497076e-06,
+      "loss": 1.1036,
+      "step": 13388
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.789785146713257,
+      "learning_rate": 7.765899122807018e-06,
+      "loss": 1.5177,
+      "step": 13389
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4070801734924316,
+      "learning_rate": 7.76498538011696e-06,
+      "loss": 1.3777,
+      "step": 13390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.230478525161743,
+      "learning_rate": 7.7640716374269e-06,
+      "loss": 1.2725,
+      "step": 13391
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6148953437805176,
+      "learning_rate": 7.763157894736843e-06,
+      "loss": 1.6357,
+      "step": 13392
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4993093013763428,
+      "learning_rate": 7.762244152046783e-06,
+      "loss": 1.5147,
+      "step": 13393
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.869849920272827,
+      "learning_rate": 7.761330409356726e-06,
+      "loss": 1.6172,
+      "step": 13394
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0662665367126465,
+      "learning_rate": 7.760416666666666e-06,
+      "loss": 1.4627,
+      "step": 13395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.357736349105835,
+      "learning_rate": 7.759502923976608e-06,
+      "loss": 1.4662,
+      "step": 13396
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7534732818603516,
+      "learning_rate": 7.75858918128655e-06,
+      "loss": 1.8318,
+      "step": 13397
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.260124444961548,
+      "learning_rate": 7.757675438596491e-06,
+      "loss": 1.2632,
+      "step": 13398
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5539050102233887,
+      "learning_rate": 7.756761695906433e-06,
+      "loss": 1.2944,
+      "step": 13399
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.935727596282959,
+      "learning_rate": 7.755847953216374e-06,
+      "loss": 1.6336,
+      "step": 13400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4932515621185303,
+      "learning_rate": 7.754934210526316e-06,
+      "loss": 1.4774,
+      "step": 13401
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.586113214492798,
+      "learning_rate": 7.754020467836257e-06,
+      "loss": 1.8039,
+      "step": 13402
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1171932220458984,
+      "learning_rate": 7.7531067251462e-06,
+      "loss": 1.7005,
+      "step": 13403
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.51168155670166,
+      "learning_rate": 7.752192982456141e-06,
+      "loss": 1.5578,
+      "step": 13404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0586535930633545,
+      "learning_rate": 7.751279239766083e-06,
+      "loss": 1.4567,
+      "step": 13405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9362385272979736,
+      "learning_rate": 7.750365497076024e-06,
+      "loss": 1.5067,
+      "step": 13406
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.555251359939575,
+      "learning_rate": 7.749451754385966e-06,
+      "loss": 1.5191,
+      "step": 13407
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.864678382873535,
+      "learning_rate": 7.748538011695908e-06,
+      "loss": 1.4554,
+      "step": 13408
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4860923290252686,
+      "learning_rate": 7.747624269005849e-06,
+      "loss": 1.164,
+      "step": 13409
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.412230968475342,
+      "learning_rate": 7.746710526315791e-06,
+      "loss": 1.4666,
+      "step": 13410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2849512100219727,
+      "learning_rate": 7.745796783625732e-06,
+      "loss": 2.1505,
+      "step": 13411
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7135708332061768,
+      "learning_rate": 7.744883040935674e-06,
+      "loss": 1.6534,
+      "step": 13412
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2135722637176514,
+      "learning_rate": 7.743969298245614e-06,
+      "loss": 1.6803,
+      "step": 13413
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.75519061088562,
+      "learning_rate": 7.743055555555556e-06,
+      "loss": 1.443,
+      "step": 13414
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5904531478881836,
+      "learning_rate": 7.742141812865499e-06,
+      "loss": 1.5852,
+      "step": 13415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0170347690582275,
+      "learning_rate": 7.74122807017544e-06,
+      "loss": 1.618,
+      "step": 13416
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.899057388305664,
+      "learning_rate": 7.740314327485381e-06,
+      "loss": 1.6741,
+      "step": 13417
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0334384441375732,
+      "learning_rate": 7.739400584795322e-06,
+      "loss": 1.4216,
+      "step": 13418
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5569756031036377,
+      "learning_rate": 7.738486842105264e-06,
+      "loss": 1.6574,
+      "step": 13419
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3062312602996826,
+      "learning_rate": 7.737573099415205e-06,
+      "loss": 1.4114,
+      "step": 13420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5234055519104004,
+      "learning_rate": 7.736659356725147e-06,
+      "loss": 1.2972,
+      "step": 13421
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8557896614074707,
+      "learning_rate": 7.73574561403509e-06,
+      "loss": 1.8752,
+      "step": 13422
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7664010524749756,
+      "learning_rate": 7.73483187134503e-06,
+      "loss": 1.5354,
+      "step": 13423
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0698041915893555,
+      "learning_rate": 7.733918128654972e-06,
+      "loss": 1.54,
+      "step": 13424
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3418800830841064,
+      "learning_rate": 7.733004385964912e-06,
+      "loss": 1.6591,
+      "step": 13425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4391162395477295,
+      "learning_rate": 7.732090643274855e-06,
+      "loss": 1.5804,
+      "step": 13426
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.702648639678955,
+      "learning_rate": 7.731176900584795e-06,
+      "loss": 1.8442,
+      "step": 13427
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3720526695251465,
+      "learning_rate": 7.730263157894737e-06,
+      "loss": 1.3792,
+      "step": 13428
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.384688377380371,
+      "learning_rate": 7.72934941520468e-06,
+      "loss": 1.1641,
+      "step": 13429
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.580354690551758,
+      "learning_rate": 7.72843567251462e-06,
+      "loss": 1.3992,
+      "step": 13430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8238394260406494,
+      "learning_rate": 7.727521929824562e-06,
+      "loss": 1.5301,
+      "step": 13431
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0470008850097656,
+      "learning_rate": 7.726608187134503e-06,
+      "loss": 1.7473,
+      "step": 13432
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3301093578338623,
+      "learning_rate": 7.725694444444445e-06,
+      "loss": 1.1738,
+      "step": 13433
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3621091842651367,
+      "learning_rate": 7.724780701754386e-06,
+      "loss": 1.9681,
+      "step": 13434
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5442659854888916,
+      "learning_rate": 7.723866959064328e-06,
+      "loss": 1.4808,
+      "step": 13435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.173427104949951,
+      "learning_rate": 7.72295321637427e-06,
+      "loss": 1.8277,
+      "step": 13436
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.478071451187134,
+      "learning_rate": 7.72203947368421e-06,
+      "loss": 1.3171,
+      "step": 13437
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5132999420166016,
+      "learning_rate": 7.721125730994153e-06,
+      "loss": 1.6183,
+      "step": 13438
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.244722366333008,
+      "learning_rate": 7.720211988304093e-06,
+      "loss": 1.2438,
+      "step": 13439
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.831852674484253,
+      "learning_rate": 7.719298245614036e-06,
+      "loss": 1.4995,
+      "step": 13440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7939395904541016,
+      "learning_rate": 7.718384502923976e-06,
+      "loss": 1.5254,
+      "step": 13441
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2861828804016113,
+      "learning_rate": 7.717470760233918e-06,
+      "loss": 1.3857,
+      "step": 13442
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.075277090072632,
+      "learning_rate": 7.71655701754386e-06,
+      "loss": 1.4325,
+      "step": 13443
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8563315868377686,
+      "learning_rate": 7.715643274853801e-06,
+      "loss": 1.4386,
+      "step": 13444
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3962841033935547,
+      "learning_rate": 7.714729532163743e-06,
+      "loss": 1.3575,
+      "step": 13445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4056057929992676,
+      "learning_rate": 7.713815789473686e-06,
+      "loss": 1.3287,
+      "step": 13446
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9932914972305298,
+      "learning_rate": 7.712902046783626e-06,
+      "loss": 1.2327,
+      "step": 13447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1900477409362793,
+      "learning_rate": 7.711988304093568e-06,
+      "loss": 0.9928,
+      "step": 13448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4532599449157715,
+      "learning_rate": 7.71107456140351e-06,
+      "loss": 0.9829,
+      "step": 13449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.246575117111206,
+      "learning_rate": 7.710160818713451e-06,
+      "loss": 1.3593,
+      "step": 13450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.561840295791626,
+      "learning_rate": 7.709247076023393e-06,
+      "loss": 1.5942,
+      "step": 13451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6858725547790527,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 1.3616,
+      "step": 13452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0095412731170654,
+      "learning_rate": 7.707419590643276e-06,
+      "loss": 1.029,
+      "step": 13453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9664275646209717,
+      "learning_rate": 7.706505847953217e-06,
+      "loss": 1.3672,
+      "step": 13454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4113380908966064,
+      "learning_rate": 7.705592105263159e-06,
+      "loss": 1.3826,
+      "step": 13455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.681004524230957,
+      "learning_rate": 7.704678362573101e-06,
+      "loss": 1.6804,
+      "step": 13456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.587735176086426,
+      "learning_rate": 7.703764619883042e-06,
+      "loss": 1.3614,
+      "step": 13457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.153939723968506,
+      "learning_rate": 7.702850877192984e-06,
+      "loss": 1.5876,
+      "step": 13458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4614145755767822,
+      "learning_rate": 7.701937134502924e-06,
+      "loss": 1.4802,
+      "step": 13459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.723857879638672,
+      "learning_rate": 7.701023391812866e-06,
+      "loss": 1.5212,
+      "step": 13460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.681100368499756,
+      "learning_rate": 7.700109649122807e-06,
+      "loss": 1.3548,
+      "step": 13461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.574694871902466,
+      "learning_rate": 7.69919590643275e-06,
+      "loss": 1.8079,
+      "step": 13462
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3665764331817627,
+      "learning_rate": 7.698282163742691e-06,
+      "loss": 1.3261,
+      "step": 13463
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1633474826812744,
+      "learning_rate": 7.697368421052632e-06,
+      "loss": 1.663,
+      "step": 13464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3054840564727783,
+      "learning_rate": 7.696454678362574e-06,
+      "loss": 2.0252,
+      "step": 13465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2981226444244385,
+      "learning_rate": 7.695540935672515e-06,
+      "loss": 1.3672,
+      "step": 13466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9708832502365112,
+      "learning_rate": 7.694627192982457e-06,
+      "loss": 1.2534,
+      "step": 13467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.428795576095581,
+      "learning_rate": 7.693713450292397e-06,
+      "loss": 1.3625,
+      "step": 13468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1678478717803955,
+      "learning_rate": 7.69279970760234e-06,
+      "loss": 1.071,
+      "step": 13469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0196802616119385,
+      "learning_rate": 7.691885964912282e-06,
+      "loss": 1.6599,
+      "step": 13470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2856533527374268,
+      "learning_rate": 7.690972222222222e-06,
+      "loss": 0.9827,
+      "step": 13471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.379639148712158,
+      "learning_rate": 7.690058479532165e-06,
+      "loss": 1.5149,
+      "step": 13472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7988383769989014,
+      "learning_rate": 7.689144736842105e-06,
+      "loss": 1.3757,
+      "step": 13473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2205286026000977,
+      "learning_rate": 7.688230994152047e-06,
+      "loss": 2.0067,
+      "step": 13474
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2105493545532227,
+      "learning_rate": 7.687317251461988e-06,
+      "loss": 1.2882,
+      "step": 13475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.270733594894409,
+      "learning_rate": 7.68640350877193e-06,
+      "loss": 1.3162,
+      "step": 13476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5839812755584717,
+      "learning_rate": 7.685489766081872e-06,
+      "loss": 1.6924,
+      "step": 13477
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.526827573776245,
+      "learning_rate": 7.684576023391813e-06,
+      "loss": 1.6521,
+      "step": 13478
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8326423168182373,
+      "learning_rate": 7.683662280701755e-06,
+      "loss": 1.5697,
+      "step": 13479
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.085170269012451,
+      "learning_rate": 7.682748538011696e-06,
+      "loss": 1.3631,
+      "step": 13480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4547579288482666,
+      "learning_rate": 7.681834795321638e-06,
+      "loss": 1.1805,
+      "step": 13481
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6243672370910645,
+      "learning_rate": 7.680921052631578e-06,
+      "loss": 1.5707,
+      "step": 13482
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.910284996032715,
+      "learning_rate": 7.68000730994152e-06,
+      "loss": 1.6743,
+      "step": 13483
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.694105386734009,
+      "learning_rate": 7.679093567251463e-06,
+      "loss": 1.5668,
+      "step": 13484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5310866832733154,
+      "learning_rate": 7.678179824561403e-06,
+      "loss": 1.2829,
+      "step": 13485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8807835578918457,
+      "learning_rate": 7.677266081871346e-06,
+      "loss": 1.1865,
+      "step": 13486
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0556790828704834,
+      "learning_rate": 7.676352339181286e-06,
+      "loss": 1.1583,
+      "step": 13487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3627588748931885,
+      "learning_rate": 7.67543859649123e-06,
+      "loss": 1.5419,
+      "step": 13488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1910712718963623,
+      "learning_rate": 7.67452485380117e-06,
+      "loss": 1.7859,
+      "step": 13489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.260176658630371,
+      "learning_rate": 7.673611111111113e-06,
+      "loss": 1.2103,
+      "step": 13490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6263978481292725,
+      "learning_rate": 7.672697368421053e-06,
+      "loss": 1.3545,
+      "step": 13491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.62794828414917,
+      "learning_rate": 7.671783625730996e-06,
+      "loss": 1.3312,
+      "step": 13492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.354069948196411,
+      "learning_rate": 7.670869883040936e-06,
+      "loss": 1.3832,
+      "step": 13493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.348435640335083,
+      "learning_rate": 7.669956140350878e-06,
+      "loss": 1.199,
+      "step": 13494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.412665605545044,
+      "learning_rate": 7.66904239766082e-06,
+      "loss": 1.3232,
+      "step": 13495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5430500507354736,
+      "learning_rate": 7.668128654970761e-06,
+      "loss": 1.4744,
+      "step": 13496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9015238285064697,
+      "learning_rate": 7.667214912280703e-06,
+      "loss": 1.5579,
+      "step": 13497
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.550546169281006,
+      "learning_rate": 7.666301169590644e-06,
+      "loss": 1.5171,
+      "step": 13498
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3476452827453613,
+      "learning_rate": 7.665387426900586e-06,
+      "loss": 1.4292,
+      "step": 13499
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1157541275024414,
+      "learning_rate": 7.664473684210527e-06,
+      "loss": 1.1386,
+      "step": 13500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5240187644958496,
+      "learning_rate": 7.663559941520469e-06,
+      "loss": 1.4407,
+      "step": 13501
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.664654493331909,
+      "learning_rate": 7.662646198830411e-06,
+      "loss": 1.5699,
+      "step": 13502
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.001457691192627,
+      "learning_rate": 7.661732456140352e-06,
+      "loss": 1.5701,
+      "step": 13503
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3309900760650635,
+      "learning_rate": 7.660818713450294e-06,
+      "loss": 1.5962,
+      "step": 13504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6736297607421875,
+      "learning_rate": 7.659904970760234e-06,
+      "loss": 1.1737,
+      "step": 13505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6216278076171875,
+      "learning_rate": 7.658991228070176e-06,
+      "loss": 1.5548,
+      "step": 13506
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3571176528930664,
+      "learning_rate": 7.658077485380117e-06,
+      "loss": 1.0938,
+      "step": 13507
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6203699111938477,
+      "learning_rate": 7.65716374269006e-06,
+      "loss": 1.8706,
+      "step": 13508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.712094306945801,
+      "learning_rate": 7.656250000000001e-06,
+      "loss": 1.6384,
+      "step": 13509
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.598701238632202,
+      "learning_rate": 7.655336257309942e-06,
+      "loss": 1.5972,
+      "step": 13510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.537827968597412,
+      "learning_rate": 7.654422514619884e-06,
+      "loss": 1.419,
+      "step": 13511
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6130809783935547,
+      "learning_rate": 7.653508771929825e-06,
+      "loss": 1.1687,
+      "step": 13512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4681951999664307,
+      "learning_rate": 7.652595029239767e-06,
+      "loss": 1.5117,
+      "step": 13513
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.543151378631592,
+      "learning_rate": 7.651681286549707e-06,
+      "loss": 1.546,
+      "step": 13514
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6269640922546387,
+      "learning_rate": 7.65076754385965e-06,
+      "loss": 1.4963,
+      "step": 13515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.419537305831909,
+      "learning_rate": 7.649853801169592e-06,
+      "loss": 1.5717,
+      "step": 13516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2916758060455322,
+      "learning_rate": 7.648940058479532e-06,
+      "loss": 1.161,
+      "step": 13517
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7289366722106934,
+      "learning_rate": 7.648026315789475e-06,
+      "loss": 1.4635,
+      "step": 13518
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.780362844467163,
+      "learning_rate": 7.647112573099415e-06,
+      "loss": 1.7961,
+      "step": 13519
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8035340309143066,
+      "learning_rate": 7.646198830409357e-06,
+      "loss": 1.3627,
+      "step": 13520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6099319458007812,
+      "learning_rate": 7.645285087719298e-06,
+      "loss": 1.5254,
+      "step": 13521
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9104461669921875,
+      "learning_rate": 7.64437134502924e-06,
+      "loss": 1.6797,
+      "step": 13522
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.399474620819092,
+      "learning_rate": 7.643457602339182e-06,
+      "loss": 1.4451,
+      "step": 13523
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3854000568389893,
+      "learning_rate": 7.642543859649123e-06,
+      "loss": 1.325,
+      "step": 13524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.160033702850342,
+      "learning_rate": 7.641630116959065e-06,
+      "loss": 1.4647,
+      "step": 13525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.654603958129883,
+      "learning_rate": 7.640716374269006e-06,
+      "loss": 1.672,
+      "step": 13526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.668165683746338,
+      "learning_rate": 7.639802631578948e-06,
+      "loss": 1.3276,
+      "step": 13527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1810927391052246,
+      "learning_rate": 7.638888888888888e-06,
+      "loss": 1.7622,
+      "step": 13528
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8264214992523193,
+      "learning_rate": 7.63797514619883e-06,
+      "loss": 1.5177,
+      "step": 13529
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.310886859893799,
+      "learning_rate": 7.637061403508773e-06,
+      "loss": 1.3599,
+      "step": 13530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8477025032043457,
+      "learning_rate": 7.636147660818715e-06,
+      "loss": 1.5329,
+      "step": 13531
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.440542459487915,
+      "learning_rate": 7.635233918128656e-06,
+      "loss": 1.3004,
+      "step": 13532
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.030919313430786,
+      "learning_rate": 7.634320175438598e-06,
+      "loss": 1.4755,
+      "step": 13533
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.740797996520996,
+      "learning_rate": 7.633406432748538e-06,
+      "loss": 1.6306,
+      "step": 13534
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2687222957611084,
+      "learning_rate": 7.63249269005848e-06,
+      "loss": 1.5936,
+      "step": 13535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5045740604400635,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 1.6869,
+      "step": 13536
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2306435108184814,
+      "learning_rate": 7.630665204678363e-06,
+      "loss": 1.263,
+      "step": 13537
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.397294521331787,
+      "learning_rate": 7.629751461988306e-06,
+      "loss": 1.3274,
+      "step": 13538
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7259092330932617,
+      "learning_rate": 7.628837719298247e-06,
+      "loss": 1.4451,
+      "step": 13539
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.445777416229248,
+      "learning_rate": 7.627923976608188e-06,
+      "loss": 1.3698,
+      "step": 13540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2411370277404785,
+      "learning_rate": 7.62701023391813e-06,
+      "loss": 1.1569,
+      "step": 13541
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9130845069885254,
+      "learning_rate": 7.626096491228071e-06,
+      "loss": 1.7749,
+      "step": 13542
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.292990207672119,
+      "learning_rate": 7.625182748538012e-06,
+      "loss": 1.2876,
+      "step": 13543
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.000784397125244,
+      "learning_rate": 7.624269005847954e-06,
+      "loss": 1.6286,
+      "step": 13544
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3359529972076416,
+      "learning_rate": 7.623355263157895e-06,
+      "loss": 1.3264,
+      "step": 13545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3273513317108154,
+      "learning_rate": 7.622441520467837e-06,
+      "loss": 1.9651,
+      "step": 13546
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.108830213546753,
+      "learning_rate": 7.621527777777779e-06,
+      "loss": 1.8886,
+      "step": 13547
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5745925903320312,
+      "learning_rate": 7.62061403508772e-06,
+      "loss": 1.5473,
+      "step": 13548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8462867736816406,
+      "learning_rate": 7.6197002923976615e-06,
+      "loss": 1.62,
+      "step": 13549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.424670934677124,
+      "learning_rate": 7.618786549707603e-06,
+      "loss": 1.4635,
+      "step": 13550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7993545532226562,
+      "learning_rate": 7.617872807017544e-06,
+      "loss": 1.8518,
+      "step": 13551
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2656025886535645,
+      "learning_rate": 7.616959064327486e-06,
+      "loss": 1.395,
+      "step": 13552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4913032054901123,
+      "learning_rate": 7.616045321637428e-06,
+      "loss": 1.2996,
+      "step": 13553
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.535783290863037,
+      "learning_rate": 7.615131578947369e-06,
+      "loss": 1.2349,
+      "step": 13554
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.664668321609497,
+      "learning_rate": 7.614217836257311e-06,
+      "loss": 1.3753,
+      "step": 13555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.539154052734375,
+      "learning_rate": 7.613304093567252e-06,
+      "loss": 1.3839,
+      "step": 13556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.19282603263855,
+      "learning_rate": 7.612390350877193e-06,
+      "loss": 1.5435,
+      "step": 13557
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7287020683288574,
+      "learning_rate": 7.611476608187135e-06,
+      "loss": 1.3837,
+      "step": 13558
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.838322162628174,
+      "learning_rate": 7.610562865497076e-06,
+      "loss": 1.6765,
+      "step": 13559
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5941755771636963,
+      "learning_rate": 7.609649122807018e-06,
+      "loss": 1.4272,
+      "step": 13560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7245054244995117,
+      "learning_rate": 7.60873538011696e-06,
+      "loss": 1.3646,
+      "step": 13561
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.528289556503296,
+      "learning_rate": 7.607821637426901e-06,
+      "loss": 1.5542,
+      "step": 13562
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.595663547515869,
+      "learning_rate": 7.6069078947368424e-06,
+      "loss": 1.0757,
+      "step": 13563
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8561787605285645,
+      "learning_rate": 7.605994152046784e-06,
+      "loss": 1.4282,
+      "step": 13564
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7328906059265137,
+      "learning_rate": 7.605080409356725e-06,
+      "loss": 1.4068,
+      "step": 13565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.046886682510376,
+      "learning_rate": 7.6041666666666666e-06,
+      "loss": 2.039,
+      "step": 13566
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.142585515975952,
+      "learning_rate": 7.603252923976609e-06,
+      "loss": 2.0456,
+      "step": 13567
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.506000280380249,
+      "learning_rate": 7.60233918128655e-06,
+      "loss": 1.7758,
+      "step": 13568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1983847618103027,
+      "learning_rate": 7.6014254385964915e-06,
+      "loss": 1.5042,
+      "step": 13569
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.941173553466797,
+      "learning_rate": 7.600511695906433e-06,
+      "loss": 1.4761,
+      "step": 13570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.412262201309204,
+      "learning_rate": 7.599597953216374e-06,
+      "loss": 1.4549,
+      "step": 13571
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8519484996795654,
+      "learning_rate": 7.598684210526316e-06,
+      "loss": 1.6642,
+      "step": 13572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.097074031829834,
+      "learning_rate": 7.597770467836257e-06,
+      "loss": 1.2318,
+      "step": 13573
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6957225799560547,
+      "learning_rate": 7.5968567251462e-06,
+      "loss": 1.5077,
+      "step": 13574
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3029274940490723,
+      "learning_rate": 7.5959429824561415e-06,
+      "loss": 1.1061,
+      "step": 13575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4503378868103027,
+      "learning_rate": 7.595029239766083e-06,
+      "loss": 1.3399,
+      "step": 13576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5357062816619873,
+      "learning_rate": 7.594115497076024e-06,
+      "loss": 1.1879,
+      "step": 13577
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4662787914276123,
+      "learning_rate": 7.593201754385966e-06,
+      "loss": 1.5047,
+      "step": 13578
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8603174686431885,
+      "learning_rate": 7.592288011695908e-06,
+      "loss": 1.3102,
+      "step": 13579
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5963456630706787,
+      "learning_rate": 7.591374269005849e-06,
+      "loss": 1.4249,
+      "step": 13580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8836162090301514,
+      "learning_rate": 7.5904605263157906e-06,
+      "loss": 1.6035,
+      "step": 13581
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2832326889038086,
+      "learning_rate": 7.589546783625732e-06,
+      "loss": 1.3736,
+      "step": 13582
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4061334133148193,
+      "learning_rate": 7.588633040935673e-06,
+      "loss": 1.6539,
+      "step": 13583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8141400814056396,
+      "learning_rate": 7.587719298245615e-06,
+      "loss": 1.6643,
+      "step": 13584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5037221908569336,
+      "learning_rate": 7.586805555555556e-06,
+      "loss": 1.5523,
+      "step": 13585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9988269805908203,
+      "learning_rate": 7.585891812865498e-06,
+      "loss": 1.736,
+      "step": 13586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8085618019104004,
+      "learning_rate": 7.58497807017544e-06,
+      "loss": 1.5498,
+      "step": 13587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.264897584915161,
+      "learning_rate": 7.584064327485381e-06,
+      "loss": 1.4073,
+      "step": 13588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6378347873687744,
+      "learning_rate": 7.583150584795322e-06,
+      "loss": 1.7265,
+      "step": 13589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9992189407348633,
+      "learning_rate": 7.582236842105264e-06,
+      "loss": 1.6215,
+      "step": 13590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.775970697402954,
+      "learning_rate": 7.581323099415205e-06,
+      "loss": 1.4209,
+      "step": 13591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.566577434539795,
+      "learning_rate": 7.5804093567251465e-06,
+      "loss": 1.6731,
+      "step": 13592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0583481788635254,
+      "learning_rate": 7.579495614035089e-06,
+      "loss": 1.7842,
+      "step": 13593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0006792545318604,
+      "learning_rate": 7.57858187134503e-06,
+      "loss": 1.8225,
+      "step": 13594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5472049713134766,
+      "learning_rate": 7.5776681286549715e-06,
+      "loss": 1.3874,
+      "step": 13595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7486908435821533,
+      "learning_rate": 7.576754385964913e-06,
+      "loss": 1.8203,
+      "step": 13596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4085867404937744,
+      "learning_rate": 7.575840643274854e-06,
+      "loss": 1.5639,
+      "step": 13597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8679754734039307,
+      "learning_rate": 7.574926900584796e-06,
+      "loss": 1.6061,
+      "step": 13598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.652646541595459,
+      "learning_rate": 7.574013157894737e-06,
+      "loss": 1.557,
+      "step": 13599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8456809520721436,
+      "learning_rate": 7.573099415204679e-06,
+      "loss": 1.7998,
+      "step": 13600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3945720195770264,
+      "learning_rate": 7.572185672514621e-06,
+      "loss": 1.9293,
+      "step": 13601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6920900344848633,
+      "learning_rate": 7.571271929824562e-06,
+      "loss": 1.6062,
+      "step": 13602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.182499885559082,
+      "learning_rate": 7.570358187134503e-06,
+      "loss": 1.2393,
+      "step": 13603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5809755325317383,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 1.3861,
+      "step": 13604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.673642635345459,
+      "learning_rate": 7.568530701754386e-06,
+      "loss": 1.3749,
+      "step": 13605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0763537883758545,
+      "learning_rate": 7.5676169590643275e-06,
+      "loss": 1.868,
+      "step": 13606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6180872917175293,
+      "learning_rate": 7.56670321637427e-06,
+      "loss": 1.3777,
+      "step": 13607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.414738893508911,
+      "learning_rate": 7.565789473684211e-06,
+      "loss": 1.3869,
+      "step": 13608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4084150791168213,
+      "learning_rate": 7.5648757309941524e-06,
+      "loss": 1.5797,
+      "step": 13609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2978689670562744,
+      "learning_rate": 7.563961988304094e-06,
+      "loss": 1.3389,
+      "step": 13610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.175370216369629,
+      "learning_rate": 7.563048245614035e-06,
+      "loss": 1.2866,
+      "step": 13611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4857466220855713,
+      "learning_rate": 7.5621345029239766e-06,
+      "loss": 1.5151,
+      "step": 13612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4575212001800537,
+      "learning_rate": 7.561220760233918e-06,
+      "loss": 1.223,
+      "step": 13613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.844373941421509,
+      "learning_rate": 7.56030701754386e-06,
+      "loss": 1.4068,
+      "step": 13614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5401906967163086,
+      "learning_rate": 7.5593932748538015e-06,
+      "loss": 1.5535,
+      "step": 13615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7115108966827393,
+      "learning_rate": 7.558479532163744e-06,
+      "loss": 1.4945,
+      "step": 13616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.779280662536621,
+      "learning_rate": 7.557565789473685e-06,
+      "loss": 1.5608,
+      "step": 13617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7817952632904053,
+      "learning_rate": 7.5566520467836265e-06,
+      "loss": 1.5936,
+      "step": 13618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2278542518615723,
+      "learning_rate": 7.555738304093569e-06,
+      "loss": 1.1639,
+      "step": 13619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.858476400375366,
+      "learning_rate": 7.55482456140351e-06,
+      "loss": 1.5388,
+      "step": 13620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0154082775115967,
+      "learning_rate": 7.5539108187134515e-06,
+      "loss": 1.6088,
+      "step": 13621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.749819755554199,
+      "learning_rate": 7.552997076023393e-06,
+      "loss": 1.9089,
+      "step": 13622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8699848651885986,
+      "learning_rate": 7.552083333333334e-06,
+      "loss": 1.5849,
+      "step": 13623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7159883975982666,
+      "learning_rate": 7.551169590643276e-06,
+      "loss": 1.8521,
+      "step": 13624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.451655387878418,
+      "learning_rate": 7.550255847953217e-06,
+      "loss": 1.4049,
+      "step": 13625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.434217929840088,
+      "learning_rate": 7.549342105263159e-06,
+      "loss": 1.5163,
+      "step": 13626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7697153091430664,
+      "learning_rate": 7.5484283625731006e-06,
+      "loss": 1.7679,
+      "step": 13627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5809919834136963,
+      "learning_rate": 7.547514619883042e-06,
+      "loss": 1.3531,
+      "step": 13628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.645742177963257,
+      "learning_rate": 7.546600877192983e-06,
+      "loss": 1.4538,
+      "step": 13629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.561124086380005,
+      "learning_rate": 7.545687134502925e-06,
+      "loss": 1.641,
+      "step": 13630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2443315982818604,
+      "learning_rate": 7.544773391812866e-06,
+      "loss": 1.361,
+      "step": 13631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4008188247680664,
+      "learning_rate": 7.5438596491228074e-06,
+      "loss": 1.5624,
+      "step": 13632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8436062335968018,
+      "learning_rate": 7.54294590643275e-06,
+      "loss": 1.5635,
+      "step": 13633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4549479484558105,
+      "learning_rate": 7.542032163742691e-06,
+      "loss": 1.5951,
+      "step": 13634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4548287391662598,
+      "learning_rate": 7.541118421052632e-06,
+      "loss": 1.3717,
+      "step": 13635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5329337120056152,
+      "learning_rate": 7.540204678362574e-06,
+      "loss": 1.6125,
+      "step": 13636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0518975257873535,
+      "learning_rate": 7.539290935672515e-06,
+      "loss": 1.1641,
+      "step": 13637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9254915714263916,
+      "learning_rate": 7.5383771929824565e-06,
+      "loss": 1.8981,
+      "step": 13638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0243797302246094,
+      "learning_rate": 7.537463450292398e-06,
+      "loss": 1.0574,
+      "step": 13639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8434925079345703,
+      "learning_rate": 7.53654970760234e-06,
+      "loss": 1.6566,
+      "step": 13640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5182814598083496,
+      "learning_rate": 7.5356359649122815e-06,
+      "loss": 1.2798,
+      "step": 13641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.792100191116333,
+      "learning_rate": 7.534722222222223e-06,
+      "loss": 1.3884,
+      "step": 13642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6665713787078857,
+      "learning_rate": 7.533808479532164e-06,
+      "loss": 1.4248,
+      "step": 13643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1027770042419434,
+      "learning_rate": 7.532894736842106e-06,
+      "loss": 1.6731,
+      "step": 13644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.151639699935913,
+      "learning_rate": 7.531980994152047e-06,
+      "loss": 1.7836,
+      "step": 13645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4385805130004883,
+      "learning_rate": 7.531067251461988e-06,
+      "loss": 1.5758,
+      "step": 13646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3759820461273193,
+      "learning_rate": 7.530153508771931e-06,
+      "loss": 1.2174,
+      "step": 13647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7685561180114746,
+      "learning_rate": 7.529239766081872e-06,
+      "loss": 1.4578,
+      "step": 13648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4428763389587402,
+      "learning_rate": 7.528326023391813e-06,
+      "loss": 1.3555,
+      "step": 13649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8420841693878174,
+      "learning_rate": 7.527412280701755e-06,
+      "loss": 1.7818,
+      "step": 13650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.545042037963867,
+      "learning_rate": 7.526498538011696e-06,
+      "loss": 1.5296,
+      "step": 13651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6429309844970703,
+      "learning_rate": 7.5255847953216375e-06,
+      "loss": 1.5776,
+      "step": 13652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4318013191223145,
+      "learning_rate": 7.524671052631579e-06,
+      "loss": 1.2052,
+      "step": 13653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.948350191116333,
+      "learning_rate": 7.523757309941521e-06,
+      "loss": 1.3863,
+      "step": 13654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.431351661682129,
+      "learning_rate": 7.5228435672514624e-06,
+      "loss": 1.4456,
+      "step": 13655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4336631298065186,
+      "learning_rate": 7.521929824561404e-06,
+      "loss": 1.4813,
+      "step": 13656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.455674886703491,
+      "learning_rate": 7.521016081871345e-06,
+      "loss": 1.3252,
+      "step": 13657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5855276584625244,
+      "learning_rate": 7.5201023391812866e-06,
+      "loss": 1.5152,
+      "step": 13658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8763701915740967,
+      "learning_rate": 7.51918859649123e-06,
+      "loss": 1.7276,
+      "step": 13659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.413194417953491,
+      "learning_rate": 7.518274853801171e-06,
+      "loss": 1.5429,
+      "step": 13660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0913350582122803,
+      "learning_rate": 7.517361111111112e-06,
+      "loss": 1.5936,
+      "step": 13661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3250620365142822,
+      "learning_rate": 7.516447368421054e-06,
+      "loss": 1.3178,
+      "step": 13662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.461554765701294,
+      "learning_rate": 7.515533625730995e-06,
+      "loss": 1.2461,
+      "step": 13663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.632842779159546,
+      "learning_rate": 7.5146198830409365e-06,
+      "loss": 1.3655,
+      "step": 13664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.759420871734619,
+      "learning_rate": 7.513706140350878e-06,
+      "loss": 1.4996,
+      "step": 13665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8265113830566406,
+      "learning_rate": 7.51279239766082e-06,
+      "loss": 1.6477,
+      "step": 13666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8504786491394043,
+      "learning_rate": 7.5118786549707615e-06,
+      "loss": 1.5013,
+      "step": 13667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.446258306503296,
+      "learning_rate": 7.510964912280703e-06,
+      "loss": 1.4797,
+      "step": 13668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1870923042297363,
+      "learning_rate": 7.510051169590644e-06,
+      "loss": 1.1507,
+      "step": 13669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7989635467529297,
+      "learning_rate": 7.509137426900586e-06,
+      "loss": 1.828,
+      "step": 13670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5459301471710205,
+      "learning_rate": 7.508223684210527e-06,
+      "loss": 1.3914,
+      "step": 13671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.937076449394226,
+      "learning_rate": 7.507309941520468e-06,
+      "loss": 1.1896,
+      "step": 13672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.587402105331421,
+      "learning_rate": 7.5063961988304106e-06,
+      "loss": 1.4694,
+      "step": 13673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4957714080810547,
+      "learning_rate": 7.505482456140352e-06,
+      "loss": 1.3416,
+      "step": 13674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.787240505218506,
+      "learning_rate": 7.504568713450293e-06,
+      "loss": 1.6575,
+      "step": 13675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9202321767807007,
+      "learning_rate": 7.503654970760235e-06,
+      "loss": 0.9415,
+      "step": 13676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3717398643493652,
+      "learning_rate": 7.502741228070176e-06,
+      "loss": 1.4256,
+      "step": 13677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.304351568222046,
+      "learning_rate": 7.5018274853801174e-06,
+      "loss": 1.3788,
+      "step": 13678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9547189474105835,
+      "learning_rate": 7.500913742690059e-06,
+      "loss": 1.1342,
+      "step": 13679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.199108362197876,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.2074,
+      "step": 13680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.400852918624878,
+      "learning_rate": 7.499086257309942e-06,
+      "loss": 1.4461,
+      "step": 13681
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7219460010528564,
+      "learning_rate": 7.498172514619884e-06,
+      "loss": 1.6409,
+      "step": 13682
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7739059925079346,
+      "learning_rate": 7.497258771929825e-06,
+      "loss": 1.7526,
+      "step": 13683
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.676921844482422,
+      "learning_rate": 7.4963450292397665e-06,
+      "loss": 1.7884,
+      "step": 13684
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1279423236846924,
+      "learning_rate": 7.495431286549708e-06,
+      "loss": 1.1814,
+      "step": 13685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7021446228027344,
+      "learning_rate": 7.494517543859649e-06,
+      "loss": 1.3966,
+      "step": 13686
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.573911428451538,
+      "learning_rate": 7.493603801169591e-06,
+      "loss": 1.641,
+      "step": 13687
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.959355354309082,
+      "learning_rate": 7.492690058479533e-06,
+      "loss": 1.6495,
+      "step": 13688
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5680437088012695,
+      "learning_rate": 7.491776315789474e-06,
+      "loss": 1.6271,
+      "step": 13689
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1968472003936768,
+      "learning_rate": 7.490862573099416e-06,
+      "loss": 1.9176,
+      "step": 13690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.539052724838257,
+      "learning_rate": 7.489948830409357e-06,
+      "loss": 1.7033,
+      "step": 13691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7955734729766846,
+      "learning_rate": 7.489035087719298e-06,
+      "loss": 1.5819,
+      "step": 13692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.275969982147217,
+      "learning_rate": 7.48812134502924e-06,
+      "loss": 1.3625,
+      "step": 13693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8776519298553467,
+      "learning_rate": 7.487207602339181e-06,
+      "loss": 1.4712,
+      "step": 13694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.350515604019165,
+      "learning_rate": 7.486293859649123e-06,
+      "loss": 1.65,
+      "step": 13695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0405123233795166,
+      "learning_rate": 7.485380116959065e-06,
+      "loss": 1.4975,
+      "step": 13696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.21075701713562,
+      "learning_rate": 7.484466374269006e-06,
+      "loss": 1.2747,
+      "step": 13697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.841688632965088,
+      "learning_rate": 7.4835526315789475e-06,
+      "loss": 1.6873,
+      "step": 13698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5030248165130615,
+      "learning_rate": 7.482638888888889e-06,
+      "loss": 1.4568,
+      "step": 13699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1975748538970947,
+      "learning_rate": 7.48172514619883e-06,
+      "loss": 1.2302,
+      "step": 13700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5145580768585205,
+      "learning_rate": 7.480811403508772e-06,
+      "loss": 1.4247,
+      "step": 13701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7923386096954346,
+      "learning_rate": 7.479897660818715e-06,
+      "loss": 1.6312,
+      "step": 13702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4257967472076416,
+      "learning_rate": 7.478983918128656e-06,
+      "loss": 1.2,
+      "step": 13703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.621748685836792,
+      "learning_rate": 7.478070175438597e-06,
+      "loss": 1.3589,
+      "step": 13704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3821640014648438,
+      "learning_rate": 7.477156432748539e-06,
+      "loss": 1.2791,
+      "step": 13705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4439167976379395,
+      "learning_rate": 7.47624269005848e-06,
+      "loss": 1.2634,
+      "step": 13706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3863346576690674,
+      "learning_rate": 7.475328947368422e-06,
+      "loss": 1.3369,
+      "step": 13707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.465642213821411,
+      "learning_rate": 7.474415204678364e-06,
+      "loss": 1.8028,
+      "step": 13708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.042771577835083,
+      "learning_rate": 7.473501461988305e-06,
+      "loss": 1.9493,
+      "step": 13709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2357096672058105,
+      "learning_rate": 7.4725877192982465e-06,
+      "loss": 1.133,
+      "step": 13710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2997467517852783,
+      "learning_rate": 7.471673976608188e-06,
+      "loss": 1.8744,
+      "step": 13711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7731733322143555,
+      "learning_rate": 7.470760233918129e-06,
+      "loss": 1.6168,
+      "step": 13712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8173632621765137,
+      "learning_rate": 7.469846491228071e-06,
+      "loss": 1.8541,
+      "step": 13713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3726673126220703,
+      "learning_rate": 7.468932748538013e-06,
+      "loss": 1.521,
+      "step": 13714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3628339767456055,
+      "learning_rate": 7.468019005847954e-06,
+      "loss": 1.5595,
+      "step": 13715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7147767543792725,
+      "learning_rate": 7.467105263157896e-06,
+      "loss": 1.6181,
+      "step": 13716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5921502113342285,
+      "learning_rate": 7.466191520467837e-06,
+      "loss": 1.2464,
+      "step": 13717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.401221752166748,
+      "learning_rate": 7.465277777777778e-06,
+      "loss": 1.3221,
+      "step": 13718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7839603424072266,
+      "learning_rate": 7.46436403508772e-06,
+      "loss": 1.9034,
+      "step": 13719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.408475637435913,
+      "learning_rate": 7.463450292397661e-06,
+      "loss": 1.5227,
+      "step": 13720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.582780122756958,
+      "learning_rate": 7.462536549707603e-06,
+      "loss": 1.4514,
+      "step": 13721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1186158657073975,
+      "learning_rate": 7.461622807017545e-06,
+      "loss": 1.1632,
+      "step": 13722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.442720413208008,
+      "learning_rate": 7.460709064327486e-06,
+      "loss": 1.363,
+      "step": 13723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.73915433883667,
+      "learning_rate": 7.4597953216374274e-06,
+      "loss": 1.535,
+      "step": 13724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4766552448272705,
+      "learning_rate": 7.458881578947369e-06,
+      "loss": 1.4854,
+      "step": 13725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.446476936340332,
+      "learning_rate": 7.45796783625731e-06,
+      "loss": 1.3485,
+      "step": 13726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3691115379333496,
+      "learning_rate": 7.4570540935672516e-06,
+      "loss": 1.3903,
+      "step": 13727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.134164810180664,
+      "learning_rate": 7.456140350877194e-06,
+      "loss": 1.1736,
+      "step": 13728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9100985527038574,
+      "learning_rate": 7.455226608187135e-06,
+      "loss": 1.6854,
+      "step": 13729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0504682064056396,
+      "learning_rate": 7.4543128654970765e-06,
+      "loss": 1.7546,
+      "step": 13730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.675295352935791,
+      "learning_rate": 7.453399122807018e-06,
+      "loss": 1.4621,
+      "step": 13731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.197629690170288,
+      "learning_rate": 7.452485380116959e-06,
+      "loss": 1.1061,
+      "step": 13732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.579514980316162,
+      "learning_rate": 7.451571637426901e-06,
+      "loss": 1.8454,
+      "step": 13733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.726041078567505,
+      "learning_rate": 7.450657894736842e-06,
+      "loss": 1.3384,
+      "step": 13734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5163967609405518,
+      "learning_rate": 7.449744152046784e-06,
+      "loss": 1.491,
+      "step": 13735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7062158584594727,
+      "learning_rate": 7.448830409356726e-06,
+      "loss": 1.4914,
+      "step": 13736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.092923402786255,
+      "learning_rate": 7.447916666666667e-06,
+      "loss": 1.8158,
+      "step": 13737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.534956693649292,
+      "learning_rate": 7.447002923976608e-06,
+      "loss": 1.6355,
+      "step": 13738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1860227584838867,
+      "learning_rate": 7.44608918128655e-06,
+      "loss": 1.3471,
+      "step": 13739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4061379432678223,
+      "learning_rate": 7.445175438596491e-06,
+      "loss": 1.3322,
+      "step": 13740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.531130790710449,
+      "learning_rate": 7.4442616959064325e-06,
+      "loss": 1.5512,
+      "step": 13741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.168487787246704,
+      "learning_rate": 7.443347953216375e-06,
+      "loss": 1.3997,
+      "step": 13742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6199092864990234,
+      "learning_rate": 7.442434210526316e-06,
+      "loss": 1.2586,
+      "step": 13743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.122060537338257,
+      "learning_rate": 7.4415204678362575e-06,
+      "loss": 1.3647,
+      "step": 13744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7125930786132812,
+      "learning_rate": 7.4406067251462e-06,
+      "loss": 1.6392,
+      "step": 13745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1929006576538086,
+      "learning_rate": 7.439692982456141e-06,
+      "loss": 1.3313,
+      "step": 13746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.009164810180664,
+      "learning_rate": 7.438779239766083e-06,
+      "loss": 1.0909,
+      "step": 13747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8219685554504395,
+      "learning_rate": 7.437865497076025e-06,
+      "loss": 1.6988,
+      "step": 13748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.268278121948242,
+      "learning_rate": 7.436951754385966e-06,
+      "loss": 1.8697,
+      "step": 13749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1101291179656982,
+      "learning_rate": 7.436038011695907e-06,
+      "loss": 1.2662,
+      "step": 13750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9959924221038818,
+      "learning_rate": 7.435124269005849e-06,
+      "loss": 1.0662,
+      "step": 13751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.371962308883667,
+      "learning_rate": 7.43421052631579e-06,
+      "loss": 1.3092,
+      "step": 13752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8879823684692383,
+      "learning_rate": 7.4332967836257315e-06,
+      "loss": 1.4221,
+      "step": 13753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71525502204895,
+      "learning_rate": 7.432383040935674e-06,
+      "loss": 1.5952,
+      "step": 13754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.280125379562378,
+      "learning_rate": 7.431469298245615e-06,
+      "loss": 1.567,
+      "step": 13755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6880435943603516,
+      "learning_rate": 7.4305555555555565e-06,
+      "loss": 1.4462,
+      "step": 13756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.544318199157715,
+      "learning_rate": 7.429641812865498e-06,
+      "loss": 1.5854,
+      "step": 13757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.572305917739868,
+      "learning_rate": 7.428728070175439e-06,
+      "loss": 1.4532,
+      "step": 13758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2112643718719482,
+      "learning_rate": 7.427814327485381e-06,
+      "loss": 1.2665,
+      "step": 13759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9345506429672241,
+      "learning_rate": 7.426900584795322e-06,
+      "loss": 1.0762,
+      "step": 13760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5969741344451904,
+      "learning_rate": 7.425986842105264e-06,
+      "loss": 1.5278,
+      "step": 13761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6826422214508057,
+      "learning_rate": 7.425073099415206e-06,
+      "loss": 1.4968,
+      "step": 13762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9983885288238525,
+      "learning_rate": 7.424159356725147e-06,
+      "loss": 1.6763,
+      "step": 13763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.346248149871826,
+      "learning_rate": 7.423245614035088e-06,
+      "loss": 1.5564,
+      "step": 13764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3024744987487793,
+      "learning_rate": 7.42233187134503e-06,
+      "loss": 1.3328,
+      "step": 13765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.656672477722168,
+      "learning_rate": 7.421418128654971e-06,
+      "loss": 1.757,
+      "step": 13766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7844746112823486,
+      "learning_rate": 7.4205043859649125e-06,
+      "loss": 1.4271,
+      "step": 13767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.917245864868164,
+      "learning_rate": 7.419590643274855e-06,
+      "loss": 1.2507,
+      "step": 13768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0884995460510254,
+      "learning_rate": 7.418676900584796e-06,
+      "loss": 1.727,
+      "step": 13769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.973803997039795,
+      "learning_rate": 7.4177631578947374e-06,
+      "loss": 1.7029,
+      "step": 13770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2438018321990967,
+      "learning_rate": 7.416849415204679e-06,
+      "loss": 1.3308,
+      "step": 13771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.99668288230896,
+      "learning_rate": 7.41593567251462e-06,
+      "loss": 1.6445,
+      "step": 13772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8415684700012207,
+      "learning_rate": 7.4150219298245616e-06,
+      "loss": 1.6233,
+      "step": 13773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.180443525314331,
+      "learning_rate": 7.414108187134503e-06,
+      "loss": 1.3446,
+      "step": 13774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6633450984954834,
+      "learning_rate": 7.413194444444445e-06,
+      "loss": 1.3983,
+      "step": 13775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.669793128967285,
+      "learning_rate": 7.4122807017543865e-06,
+      "loss": 1.6949,
+      "step": 13776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.79130482673645,
+      "learning_rate": 7.411366959064328e-06,
+      "loss": 1.7499,
+      "step": 13777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.338907241821289,
+      "learning_rate": 7.410453216374269e-06,
+      "loss": 1.1593,
+      "step": 13778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.455110788345337,
+      "learning_rate": 7.409539473684211e-06,
+      "loss": 1.4428,
+      "step": 13779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5101089477539062,
+      "learning_rate": 7.408625730994152e-06,
+      "loss": 1.4734,
+      "step": 13780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.663766860961914,
+      "learning_rate": 7.407711988304093e-06,
+      "loss": 1.4961,
+      "step": 13781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4034533500671387,
+      "learning_rate": 7.406798245614036e-06,
+      "loss": 1.3571,
+      "step": 13782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6048600673675537,
+      "learning_rate": 7.405884502923977e-06,
+      "loss": 1.6303,
+      "step": 13783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1683342456817627,
+      "learning_rate": 7.404970760233918e-06,
+      "loss": 1.4462,
+      "step": 13784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.228227376937866,
+      "learning_rate": 7.40405701754386e-06,
+      "loss": 1.3318,
+      "step": 13785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6566829681396484,
+      "learning_rate": 7.403143274853801e-06,
+      "loss": 1.3061,
+      "step": 13786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.868152141571045,
+      "learning_rate": 7.402229532163744e-06,
+      "loss": 1.4595,
+      "step": 13787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4116249084472656,
+      "learning_rate": 7.4013157894736856e-06,
+      "loss": 1.4143,
+      "step": 13788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5867738723754883,
+      "learning_rate": 7.400402046783627e-06,
+      "loss": 1.5025,
+      "step": 13789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4241645336151123,
+      "learning_rate": 7.399488304093568e-06,
+      "loss": 1.4614,
+      "step": 13790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6786835193634033,
+      "learning_rate": 7.39857456140351e-06,
+      "loss": 1.3429,
+      "step": 13791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5067827701568604,
+      "learning_rate": 7.397660818713451e-06,
+      "loss": 1.3721,
+      "step": 13792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.226579427719116,
+      "learning_rate": 7.3967470760233924e-06,
+      "loss": 1.4154,
+      "step": 13793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5311691761016846,
+      "learning_rate": 7.395833333333335e-06,
+      "loss": 1.445,
+      "step": 13794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6371302604675293,
+      "learning_rate": 7.394919590643276e-06,
+      "loss": 1.4849,
+      "step": 13795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.400496006011963,
+      "learning_rate": 7.394005847953217e-06,
+      "loss": 1.4808,
+      "step": 13796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5740458965301514,
+      "learning_rate": 7.393092105263159e-06,
+      "loss": 1.6547,
+      "step": 13797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.454127073287964,
+      "learning_rate": 7.3921783625731e-06,
+      "loss": 1.2945,
+      "step": 13798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.185922384262085,
+      "learning_rate": 7.3912646198830415e-06,
+      "loss": 1.8582,
+      "step": 13799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5178720951080322,
+      "learning_rate": 7.390350877192983e-06,
+      "loss": 1.4795,
+      "step": 13800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.328575372695923,
+      "learning_rate": 7.389437134502925e-06,
+      "loss": 1.5998,
+      "step": 13801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.831873893737793,
+      "learning_rate": 7.3885233918128665e-06,
+      "loss": 1.5554,
+      "step": 13802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.443133592605591,
+      "learning_rate": 7.387609649122808e-06,
+      "loss": 1.5278,
+      "step": 13803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.586348295211792,
+      "learning_rate": 7.386695906432749e-06,
+      "loss": 1.6036,
+      "step": 13804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8312294483184814,
+      "learning_rate": 7.385782163742691e-06,
+      "loss": 1.9383,
+      "step": 13805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.533778190612793,
+      "learning_rate": 7.384868421052632e-06,
+      "loss": 1.5817,
+      "step": 13806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4687998294830322,
+      "learning_rate": 7.383954678362573e-06,
+      "loss": 1.5741,
+      "step": 13807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.992228388786316,
+      "learning_rate": 7.383040935672516e-06,
+      "loss": 1.0926,
+      "step": 13808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.720335006713867,
+      "learning_rate": 7.382127192982457e-06,
+      "loss": 1.4754,
+      "step": 13809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4036598205566406,
+      "learning_rate": 7.381213450292398e-06,
+      "loss": 1.1872,
+      "step": 13810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7416043281555176,
+      "learning_rate": 7.38029970760234e-06,
+      "loss": 1.8526,
+      "step": 13811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4705562591552734,
+      "learning_rate": 7.379385964912281e-06,
+      "loss": 1.3366,
+      "step": 13812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7967522144317627,
+      "learning_rate": 7.3784722222222225e-06,
+      "loss": 1.6097,
+      "step": 13813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.822599411010742,
+      "learning_rate": 7.377558479532164e-06,
+      "loss": 1.7505,
+      "step": 13814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9388694763183594,
+      "learning_rate": 7.376644736842106e-06,
+      "loss": 1.6818,
+      "step": 13815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.638821840286255,
+      "learning_rate": 7.3757309941520474e-06,
+      "loss": 1.741,
+      "step": 13816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8702008724212646,
+      "learning_rate": 7.374817251461989e-06,
+      "loss": 1.5392,
+      "step": 13817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.844773054122925,
+      "learning_rate": 7.37390350877193e-06,
+      "loss": 1.703,
+      "step": 13818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6684682369232178,
+      "learning_rate": 7.3729897660818716e-06,
+      "loss": 1.3777,
+      "step": 13819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.849928379058838,
+      "learning_rate": 7.372076023391813e-06,
+      "loss": 1.3272,
+      "step": 13820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2126526832580566,
+      "learning_rate": 7.371162280701754e-06,
+      "loss": 1.3675,
+      "step": 13821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7243154048919678,
+      "learning_rate": 7.3702485380116965e-06,
+      "loss": 1.6238,
+      "step": 13822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3186326026916504,
+      "learning_rate": 7.369334795321638e-06,
+      "loss": 1.1954,
+      "step": 13823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6631011962890625,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 1.6341,
+      "step": 13824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.074652910232544,
+      "learning_rate": 7.367507309941521e-06,
+      "loss": 1.2338,
+      "step": 13825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4502196311950684,
+      "learning_rate": 7.366593567251462e-06,
+      "loss": 1.4314,
+      "step": 13826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.998501777648926,
+      "learning_rate": 7.365679824561403e-06,
+      "loss": 1.6125,
+      "step": 13827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.511984348297119,
+      "learning_rate": 7.364766081871345e-06,
+      "loss": 1.726,
+      "step": 13828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6849520206451416,
+      "learning_rate": 7.363852339181287e-06,
+      "loss": 1.354,
+      "step": 13829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3248729705810547,
+      "learning_rate": 7.362938596491229e-06,
+      "loss": 1.239,
+      "step": 13830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9396615028381348,
+      "learning_rate": 7.362024853801171e-06,
+      "loss": 1.459,
+      "step": 13831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.547630786895752,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 1.2869,
+      "step": 13832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.346137762069702,
+      "learning_rate": 7.360197368421053e-06,
+      "loss": 1.3165,
+      "step": 13833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1162195205688477,
+      "learning_rate": 7.3592836257309956e-06,
+      "loss": 1.1737,
+      "step": 13834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3126814365386963,
+      "learning_rate": 7.358369883040937e-06,
+      "loss": 1.3161,
+      "step": 13835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.271209239959717,
+      "learning_rate": 7.357456140350878e-06,
+      "loss": 1.2589,
+      "step": 13836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.046215295791626,
+      "learning_rate": 7.35654239766082e-06,
+      "loss": 1.369,
+      "step": 13837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.151190996170044,
+      "learning_rate": 7.355628654970761e-06,
+      "loss": 1.5667,
+      "step": 13838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9162211418151855,
+      "learning_rate": 7.3547149122807024e-06,
+      "loss": 1.7123,
+      "step": 13839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7650599479675293,
+      "learning_rate": 7.353801169590644e-06,
+      "loss": 1.4802,
+      "step": 13840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.032100200653076,
+      "learning_rate": 7.352887426900586e-06,
+      "loss": 1.1329,
+      "step": 13841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671800136566162,
+      "learning_rate": 7.351973684210527e-06,
+      "loss": 1.5498,
+      "step": 13842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.774129629135132,
+      "learning_rate": 7.351059941520469e-06,
+      "loss": 1.7404,
+      "step": 13843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.36529278755188,
+      "learning_rate": 7.35014619883041e-06,
+      "loss": 1.6224,
+      "step": 13844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.975659966468811,
+      "learning_rate": 7.3492324561403515e-06,
+      "loss": 1.3784,
+      "step": 13845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6017532348632812,
+      "learning_rate": 7.348318713450293e-06,
+      "loss": 1.3937,
+      "step": 13846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0877878665924072,
+      "learning_rate": 7.347404970760234e-06,
+      "loss": 1.7994,
+      "step": 13847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3586690425872803,
+      "learning_rate": 7.3464912280701765e-06,
+      "loss": 1.3107,
+      "step": 13848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7950788736343384,
+      "learning_rate": 7.345577485380118e-06,
+      "loss": 1.0616,
+      "step": 13849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5975146293640137,
+      "learning_rate": 7.344663742690059e-06,
+      "loss": 1.5825,
+      "step": 13850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3555116653442383,
+      "learning_rate": 7.343750000000001e-06,
+      "loss": 1.4781,
+      "step": 13851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.528935194015503,
+      "learning_rate": 7.342836257309942e-06,
+      "loss": 1.1897,
+      "step": 13852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.144657850265503,
+      "learning_rate": 7.341922514619883e-06,
+      "loss": 1.3343,
+      "step": 13853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.456437110900879,
+      "learning_rate": 7.341008771929825e-06,
+      "loss": 1.4314,
+      "step": 13854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0105197429656982,
+      "learning_rate": 7.340095029239767e-06,
+      "loss": 1.2166,
+      "step": 13855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.025127649307251,
+      "learning_rate": 7.339181286549708e-06,
+      "loss": 1.0711,
+      "step": 13856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.412297010421753,
+      "learning_rate": 7.33826754385965e-06,
+      "loss": 1.3621,
+      "step": 13857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.876343011856079,
+      "learning_rate": 7.337353801169591e-06,
+      "loss": 1.7146,
+      "step": 13858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8972370624542236,
+      "learning_rate": 7.3364400584795325e-06,
+      "loss": 1.5286,
+      "step": 13859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.804128885269165,
+      "learning_rate": 7.335526315789474e-06,
+      "loss": 1.4625,
+      "step": 13860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.291494131088257,
+      "learning_rate": 7.334612573099415e-06,
+      "loss": 1.2874,
+      "step": 13861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4759206771850586,
+      "learning_rate": 7.3336988304093574e-06,
+      "loss": 1.6546,
+      "step": 13862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7248520851135254,
+      "learning_rate": 7.332785087719299e-06,
+      "loss": 1.4345,
+      "step": 13863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3564562797546387,
+      "learning_rate": 7.33187134502924e-06,
+      "loss": 1.4745,
+      "step": 13864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.966580390930176,
+      "learning_rate": 7.3309576023391816e-06,
+      "loss": 1.8075,
+      "step": 13865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.627204418182373,
+      "learning_rate": 7.330043859649123e-06,
+      "loss": 1.4747,
+      "step": 13866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.069070816040039,
+      "learning_rate": 7.329130116959064e-06,
+      "loss": 1.287,
+      "step": 13867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8575973510742188,
+      "learning_rate": 7.328216374269006e-06,
+      "loss": 1.3692,
+      "step": 13868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.215773820877075,
+      "learning_rate": 7.327302631578948e-06,
+      "loss": 1.5497,
+      "step": 13869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4789881706237793,
+      "learning_rate": 7.326388888888889e-06,
+      "loss": 1.2907,
+      "step": 13870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.499483823776245,
+      "learning_rate": 7.325475146198831e-06,
+      "loss": 1.5363,
+      "step": 13871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8335068225860596,
+      "learning_rate": 7.324561403508772e-06,
+      "loss": 1.6946,
+      "step": 13872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.605348587036133,
+      "learning_rate": 7.323647660818714e-06,
+      "loss": 1.2534,
+      "step": 13873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4839532375335693,
+      "learning_rate": 7.3227339181286565e-06,
+      "loss": 2.2555,
+      "step": 13874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.464869499206543,
+      "learning_rate": 7.321820175438598e-06,
+      "loss": 1.7046,
+      "step": 13875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.591212272644043,
+      "learning_rate": 7.320906432748539e-06,
+      "loss": 1.607,
+      "step": 13876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7173943519592285,
+      "learning_rate": 7.319992690058481e-06,
+      "loss": 1.4675,
+      "step": 13877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.421110153198242,
+      "learning_rate": 7.319078947368422e-06,
+      "loss": 1.4069,
+      "step": 13878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8759169578552246,
+      "learning_rate": 7.318165204678363e-06,
+      "loss": 1.5419,
+      "step": 13879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8296945095062256,
+      "learning_rate": 7.317251461988305e-06,
+      "loss": 1.648,
+      "step": 13880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6599128246307373,
+      "learning_rate": 7.316337719298247e-06,
+      "loss": 1.5441,
+      "step": 13881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.954883575439453,
+      "learning_rate": 7.315423976608188e-06,
+      "loss": 1.2498,
+      "step": 13882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.701876640319824,
+      "learning_rate": 7.31451023391813e-06,
+      "loss": 1.5747,
+      "step": 13883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1779065132141113,
+      "learning_rate": 7.313596491228071e-06,
+      "loss": 1.1726,
+      "step": 13884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7224502563476562,
+      "learning_rate": 7.3126827485380124e-06,
+      "loss": 1.9069,
+      "step": 13885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7532331943511963,
+      "learning_rate": 7.311769005847954e-06,
+      "loss": 1.3272,
+      "step": 13886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.244326114654541,
+      "learning_rate": 7.310855263157895e-06,
+      "loss": 1.4516,
+      "step": 13887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.75453782081604,
+      "learning_rate": 7.309941520467837e-06,
+      "loss": 1.6861,
+      "step": 13888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.769958257675171,
+      "learning_rate": 7.309027777777779e-06,
+      "loss": 1.6003,
+      "step": 13889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6107966899871826,
+      "learning_rate": 7.30811403508772e-06,
+      "loss": 1.7488,
+      "step": 13890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5313284397125244,
+      "learning_rate": 7.3072002923976615e-06,
+      "loss": 1.5451,
+      "step": 13891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3611507415771484,
+      "learning_rate": 7.306286549707603e-06,
+      "loss": 1.3077,
+      "step": 13892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0339179039001465,
+      "learning_rate": 7.305372807017544e-06,
+      "loss": 1.853,
+      "step": 13893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3628504276275635,
+      "learning_rate": 7.304459064327486e-06,
+      "loss": 1.2293,
+      "step": 13894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.564542770385742,
+      "learning_rate": 7.303545321637428e-06,
+      "loss": 1.6079,
+      "step": 13895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.78096604347229,
+      "learning_rate": 7.302631578947369e-06,
+      "loss": 1.7894,
+      "step": 13896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.972888231277466,
+      "learning_rate": 7.301717836257311e-06,
+      "loss": 1.8911,
+      "step": 13897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1715805530548096,
+      "learning_rate": 7.300804093567252e-06,
+      "loss": 1.2825,
+      "step": 13898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1308557987213135,
+      "learning_rate": 7.299890350877193e-06,
+      "loss": 1.1747,
+      "step": 13899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.957216262817383,
+      "learning_rate": 7.298976608187135e-06,
+      "loss": 1.6112,
+      "step": 13900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.635333776473999,
+      "learning_rate": 7.298062865497076e-06,
+      "loss": 1.6262,
+      "step": 13901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.382312059402466,
+      "learning_rate": 7.297149122807018e-06,
+      "loss": 1.4801,
+      "step": 13902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.490206003189087,
+      "learning_rate": 7.29623538011696e-06,
+      "loss": 1.5557,
+      "step": 13903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.602240562438965,
+      "learning_rate": 7.295321637426901e-06,
+      "loss": 1.6313,
+      "step": 13904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.313554048538208,
+      "learning_rate": 7.2944078947368425e-06,
+      "loss": 1.3372,
+      "step": 13905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.328022003173828,
+      "learning_rate": 7.293494152046784e-06,
+      "loss": 1.0871,
+      "step": 13906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.485708236694336,
+      "learning_rate": 7.292580409356725e-06,
+      "loss": 1.2819,
+      "step": 13907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1978533267974854,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 1.138,
+      "step": 13908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6706600189208984,
+      "learning_rate": 7.290752923976608e-06,
+      "loss": 1.6177,
+      "step": 13909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.134840488433838,
+      "learning_rate": 7.28983918128655e-06,
+      "loss": 1.4081,
+      "step": 13910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7370736598968506,
+      "learning_rate": 7.2889254385964916e-06,
+      "loss": 1.5507,
+      "step": 13911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1681735515594482,
+      "learning_rate": 7.288011695906433e-06,
+      "loss": 1.2514,
+      "step": 13912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.334115982055664,
+      "learning_rate": 7.287097953216374e-06,
+      "loss": 1.4616,
+      "step": 13913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8197531700134277,
+      "learning_rate": 7.286184210526316e-06,
+      "loss": 1.6187,
+      "step": 13914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.707378387451172,
+      "learning_rate": 7.285270467836257e-06,
+      "loss": 1.1104,
+      "step": 13915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.031139612197876,
+      "learning_rate": 7.2843567251462e-06,
+      "loss": 1.5999,
+      "step": 13916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3219974040985107,
+      "learning_rate": 7.2834429824561415e-06,
+      "loss": 1.4327,
+      "step": 13917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.190732955932617,
+      "learning_rate": 7.282529239766083e-06,
+      "loss": 1.4372,
+      "step": 13918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.41792893409729,
+      "learning_rate": 7.281615497076024e-06,
+      "loss": 1.3201,
+      "step": 13919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0637731552124023,
+      "learning_rate": 7.280701754385966e-06,
+      "loss": 1.5905,
+      "step": 13920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.604810953140259,
+      "learning_rate": 7.279788011695908e-06,
+      "loss": 1.7051,
+      "step": 13921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5114123821258545,
+      "learning_rate": 7.278874269005849e-06,
+      "loss": 1.3421,
+      "step": 13922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.531942844390869,
+      "learning_rate": 7.277960526315791e-06,
+      "loss": 1.2834,
+      "step": 13923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.728544235229492,
+      "learning_rate": 7.277046783625732e-06,
+      "loss": 1.5048,
+      "step": 13924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.622285842895508,
+      "learning_rate": 7.276133040935673e-06,
+      "loss": 1.4154,
+      "step": 13925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1364152431488037,
+      "learning_rate": 7.275219298245615e-06,
+      "loss": 1.1396,
+      "step": 13926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.035156488418579,
+      "learning_rate": 7.274305555555556e-06,
+      "loss": 1.6989,
+      "step": 13927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5508861541748047,
+      "learning_rate": 7.2733918128654975e-06,
+      "loss": 1.6219,
+      "step": 13928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0152361392974854,
+      "learning_rate": 7.27247807017544e-06,
+      "loss": 1.2173,
+      "step": 13929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.205348491668701,
+      "learning_rate": 7.271564327485381e-06,
+      "loss": 1.3191,
+      "step": 13930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7067341804504395,
+      "learning_rate": 7.2706505847953224e-06,
+      "loss": 1.7142,
+      "step": 13931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.156867027282715,
+      "learning_rate": 7.269736842105264e-06,
+      "loss": 1.9396,
+      "step": 13932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2977778911590576,
+      "learning_rate": 7.268823099415205e-06,
+      "loss": 1.3749,
+      "step": 13933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.028965950012207,
+      "learning_rate": 7.2679093567251466e-06,
+      "loss": 1.7021,
+      "step": 13934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.401048421859741,
+      "learning_rate": 7.266995614035088e-06,
+      "loss": 1.4157,
+      "step": 13935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.19661545753479,
+      "learning_rate": 7.26608187134503e-06,
+      "loss": 1.2461,
+      "step": 13936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7313926219940186,
+      "learning_rate": 7.2651681286549715e-06,
+      "loss": 1.4945,
+      "step": 13937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4562783241271973,
+      "learning_rate": 7.264254385964913e-06,
+      "loss": 1.4055,
+      "step": 13938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0449936389923096,
+      "learning_rate": 7.263340643274854e-06,
+      "loss": 1.1939,
+      "step": 13939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.405360460281372,
+      "learning_rate": 7.262426900584796e-06,
+      "loss": 1.5405,
+      "step": 13940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.394313335418701,
+      "learning_rate": 7.261513157894737e-06,
+      "loss": 1.2742,
+      "step": 13941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.559204339981079,
+      "learning_rate": 7.260599415204678e-06,
+      "loss": 1.5222,
+      "step": 13942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.651465892791748,
+      "learning_rate": 7.259685672514621e-06,
+      "loss": 1.5987,
+      "step": 13943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.450875997543335,
+      "learning_rate": 7.258771929824562e-06,
+      "loss": 1.3469,
+      "step": 13944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.181394577026367,
+      "learning_rate": 7.257858187134503e-06,
+      "loss": 1.7501,
+      "step": 13945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5101609230041504,
+      "learning_rate": 7.256944444444445e-06,
+      "loss": 1.4398,
+      "step": 13946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3056843280792236,
+      "learning_rate": 7.256030701754386e-06,
+      "loss": 1.2996,
+      "step": 13947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4051766395568848,
+      "learning_rate": 7.2551169590643275e-06,
+      "loss": 1.6002,
+      "step": 13948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5886576175689697,
+      "learning_rate": 7.254203216374269e-06,
+      "loss": 1.7608,
+      "step": 13949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2836380004882812,
+      "learning_rate": 7.253289473684211e-06,
+      "loss": 1.4023,
+      "step": 13950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9791474342346191,
+      "learning_rate": 7.2523757309941525e-06,
+      "loss": 1.1373,
+      "step": 13951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.119645118713379,
+      "learning_rate": 7.251461988304094e-06,
+      "loss": 1.2677,
+      "step": 13952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.109754800796509,
+      "learning_rate": 7.250548245614035e-06,
+      "loss": 1.5194,
+      "step": 13953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.870804786682129,
+      "learning_rate": 7.249634502923977e-06,
+      "loss": 1.7667,
+      "step": 13954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5209100246429443,
+      "learning_rate": 7.248720760233918e-06,
+      "loss": 1.4074,
+      "step": 13955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.009793758392334,
+      "learning_rate": 7.247807017543859e-06,
+      "loss": 1.7147,
+      "step": 13956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6683762073516846,
+      "learning_rate": 7.2468932748538016e-06,
+      "loss": 1.6007,
+      "step": 13957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3336760997772217,
+      "learning_rate": 7.245979532163744e-06,
+      "loss": 1.3714,
+      "step": 13958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3862688541412354,
+      "learning_rate": 7.245065789473685e-06,
+      "loss": 1.4334,
+      "step": 13959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0073764324188232,
+      "learning_rate": 7.2441520467836265e-06,
+      "loss": 1.0142,
+      "step": 13960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.829716444015503,
+      "learning_rate": 7.243238304093568e-06,
+      "loss": 1.5554,
+      "step": 13961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.238757610321045,
+      "learning_rate": 7.24232456140351e-06,
+      "loss": 1.4066,
+      "step": 13962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9249813556671143,
+      "learning_rate": 7.2414108187134515e-06,
+      "loss": 1.2051,
+      "step": 13963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6755831241607666,
+      "learning_rate": 7.240497076023393e-06,
+      "loss": 1.5233,
+      "step": 13964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0914506912231445,
+      "learning_rate": 7.239583333333334e-06,
+      "loss": 1.1999,
+      "step": 13965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.667466640472412,
+      "learning_rate": 7.238669590643276e-06,
+      "loss": 1.5271,
+      "step": 13966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3508055210113525,
+      "learning_rate": 7.237755847953217e-06,
+      "loss": 1.5425,
+      "step": 13967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0416057109832764,
+      "learning_rate": 7.236842105263158e-06,
+      "loss": 1.062,
+      "step": 13968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8264973163604736,
+      "learning_rate": 7.235928362573101e-06,
+      "loss": 1.464,
+      "step": 13969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3398447036743164,
+      "learning_rate": 7.235014619883042e-06,
+      "loss": 1.2854,
+      "step": 13970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.323138475418091,
+      "learning_rate": 7.234100877192983e-06,
+      "loss": 1.1642,
+      "step": 13971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2732958793640137,
+      "learning_rate": 7.233187134502925e-06,
+      "loss": 1.4394,
+      "step": 13972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.502732753753662,
+      "learning_rate": 7.232273391812866e-06,
+      "loss": 1.5793,
+      "step": 13973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3524205684661865,
+      "learning_rate": 7.2313596491228075e-06,
+      "loss": 1.2506,
+      "step": 13974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1089823246002197,
+      "learning_rate": 7.230445906432749e-06,
+      "loss": 1.0707,
+      "step": 13975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6212754249572754,
+      "learning_rate": 7.229532163742691e-06,
+      "loss": 1.5099,
+      "step": 13976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.165191411972046,
+      "learning_rate": 7.2286184210526324e-06,
+      "loss": 1.3771,
+      "step": 13977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.388972520828247,
+      "learning_rate": 7.227704678362574e-06,
+      "loss": 1.4348,
+      "step": 13978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2371790409088135,
+      "learning_rate": 7.226790935672515e-06,
+      "loss": 1.3517,
+      "step": 13979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3830885887145996,
+      "learning_rate": 7.2258771929824566e-06,
+      "loss": 1.309,
+      "step": 13980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.228996753692627,
+      "learning_rate": 7.224963450292398e-06,
+      "loss": 1.2078,
+      "step": 13981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.649794578552246,
+      "learning_rate": 7.224049707602339e-06,
+      "loss": 1.4313,
+      "step": 13982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8073965311050415,
+      "learning_rate": 7.2231359649122815e-06,
+      "loss": 0.9776,
+      "step": 13983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4410860538482666,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 1.6871,
+      "step": 13984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.94427490234375,
+      "learning_rate": 7.221308479532164e-06,
+      "loss": 1.5447,
+      "step": 13985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5330281257629395,
+      "learning_rate": 7.220394736842106e-06,
+      "loss": 1.4592,
+      "step": 13986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.423778533935547,
+      "learning_rate": 7.219480994152047e-06,
+      "loss": 1.2378,
+      "step": 13987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.600778102874756,
+      "learning_rate": 7.218567251461988e-06,
+      "loss": 1.4648,
+      "step": 13988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4628560543060303,
+      "learning_rate": 7.21765350877193e-06,
+      "loss": 1.3549,
+      "step": 13989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7146801948547363,
+      "learning_rate": 7.216739766081872e-06,
+      "loss": 1.5822,
+      "step": 13990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6343789100646973,
+      "learning_rate": 7.215826023391813e-06,
+      "loss": 1.4993,
+      "step": 13991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.634121894836426,
+      "learning_rate": 7.214912280701755e-06,
+      "loss": 1.6548,
+      "step": 13992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.851747512817383,
+      "learning_rate": 7.213998538011696e-06,
+      "loss": 1.5798,
+      "step": 13993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6124117374420166,
+      "learning_rate": 7.2130847953216375e-06,
+      "loss": 1.6675,
+      "step": 13994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.122764825820923,
+      "learning_rate": 7.212171052631579e-06,
+      "loss": 1.2729,
+      "step": 13995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3177106380462646,
+      "learning_rate": 7.21125730994152e-06,
+      "loss": 1.4151,
+      "step": 13996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.887986898422241,
+      "learning_rate": 7.2103435672514625e-06,
+      "loss": 1.792,
+      "step": 13997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.618446111679077,
+      "learning_rate": 7.209429824561404e-06,
+      "loss": 1.3077,
+      "step": 13998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4473659992218018,
+      "learning_rate": 7.208516081871345e-06,
+      "loss": 1.5877,
+      "step": 13999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8699305057525635,
+      "learning_rate": 7.207602339181287e-06,
+      "loss": 1.7929,
+      "step": 14000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.490877389907837,
+      "learning_rate": 7.206688596491229e-06,
+      "loss": 1.3021,
+      "step": 14001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.220082998275757,
+      "learning_rate": 7.205774853801171e-06,
+      "loss": 1.1997,
+      "step": 14002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8435020446777344,
+      "learning_rate": 7.204861111111112e-06,
+      "loss": 1.5761,
+      "step": 14003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4428937435150146,
+      "learning_rate": 7.203947368421054e-06,
+      "loss": 1.5323,
+      "step": 14004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.880016326904297,
+      "learning_rate": 7.203033625730995e-06,
+      "loss": 1.6516,
+      "step": 14005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1895225048065186,
+      "learning_rate": 7.2021198830409365e-06,
+      "loss": 1.3128,
+      "step": 14006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.320530414581299,
+      "learning_rate": 7.201206140350878e-06,
+      "loss": 1.3981,
+      "step": 14007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8548150062561035,
+      "learning_rate": 7.200292397660819e-06,
+      "loss": 1.468,
+      "step": 14008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.214665174484253,
+      "learning_rate": 7.1993786549707615e-06,
+      "loss": 1.2656,
+      "step": 14009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.230970621109009,
+      "learning_rate": 7.198464912280703e-06,
+      "loss": 1.9397,
+      "step": 14010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4452319145202637,
+      "learning_rate": 7.197551169590644e-06,
+      "loss": 1.4315,
+      "step": 14011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0861566066741943,
+      "learning_rate": 7.196637426900586e-06,
+      "loss": 1.6294,
+      "step": 14012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.539731740951538,
+      "learning_rate": 7.195723684210527e-06,
+      "loss": 1.3176,
+      "step": 14013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7072958946228027,
+      "learning_rate": 7.194809941520468e-06,
+      "loss": 1.9092,
+      "step": 14014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9220306873321533,
+      "learning_rate": 7.19389619883041e-06,
+      "loss": 1.7551,
+      "step": 14015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1957249641418457,
+      "learning_rate": 7.192982456140352e-06,
+      "loss": 1.3681,
+      "step": 14016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.610914945602417,
+      "learning_rate": 7.192068713450293e-06,
+      "loss": 1.4313,
+      "step": 14017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5149803161621094,
+      "learning_rate": 7.191154970760235e-06,
+      "loss": 1.4766,
+      "step": 14018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.314777374267578,
+      "learning_rate": 7.190241228070176e-06,
+      "loss": 1.3885,
+      "step": 14019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.642444372177124,
+      "learning_rate": 7.1893274853801175e-06,
+      "loss": 1.3851,
+      "step": 14020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.561309814453125,
+      "learning_rate": 7.188413742690059e-06,
+      "loss": 1.505,
+      "step": 14021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8552403450012207,
+      "learning_rate": 7.1875e-06,
+      "loss": 1.9008,
+      "step": 14022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.203688144683838,
+      "learning_rate": 7.1865862573099424e-06,
+      "loss": 1.7236,
+      "step": 14023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4591927528381348,
+      "learning_rate": 7.185672514619884e-06,
+      "loss": 1.6528,
+      "step": 14024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9637885093688965,
+      "learning_rate": 7.184758771929825e-06,
+      "loss": 1.7096,
+      "step": 14025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3123373985290527,
+      "learning_rate": 7.1838450292397666e-06,
+      "loss": 1.1959,
+      "step": 14026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4488565921783447,
+      "learning_rate": 7.182931286549708e-06,
+      "loss": 1.1375,
+      "step": 14027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.647333860397339,
+      "learning_rate": 7.182017543859649e-06,
+      "loss": 1.292,
+      "step": 14028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2235758304595947,
+      "learning_rate": 7.181103801169591e-06,
+      "loss": 1.5061,
+      "step": 14029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.543175458908081,
+      "learning_rate": 7.180190058479533e-06,
+      "loss": 1.5649,
+      "step": 14030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.712188243865967,
+      "learning_rate": 7.179276315789474e-06,
+      "loss": 1.6387,
+      "step": 14031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5710365772247314,
+      "learning_rate": 7.178362573099416e-06,
+      "loss": 1.5162,
+      "step": 14032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.990760564804077,
+      "learning_rate": 7.177448830409357e-06,
+      "loss": 1.1664,
+      "step": 14033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5300791263580322,
+      "learning_rate": 7.176535087719298e-06,
+      "loss": 1.3542,
+      "step": 14034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.235630512237549,
+      "learning_rate": 7.17562134502924e-06,
+      "loss": 1.7217,
+      "step": 14035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7939980030059814,
+      "learning_rate": 7.174707602339181e-06,
+      "loss": 1.894,
+      "step": 14036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5166783332824707,
+      "learning_rate": 7.173793859649123e-06,
+      "loss": 1.2666,
+      "step": 14037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6884002685546875,
+      "learning_rate": 7.172880116959065e-06,
+      "loss": 1.7897,
+      "step": 14038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2029407024383545,
+      "learning_rate": 7.171966374269006e-06,
+      "loss": 1.3769,
+      "step": 14039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2885096073150635,
+      "learning_rate": 7.1710526315789475e-06,
+      "loss": 1.2462,
+      "step": 14040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.436405897140503,
+      "learning_rate": 7.170138888888889e-06,
+      "loss": 1.4599,
+      "step": 14041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1247482299804688,
+      "learning_rate": 7.16922514619883e-06,
+      "loss": 1.37,
+      "step": 14042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9191852807998657,
+      "learning_rate": 7.168311403508772e-06,
+      "loss": 1.0519,
+      "step": 14043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5503084659576416,
+      "learning_rate": 7.167397660818715e-06,
+      "loss": 1.6506,
+      "step": 14044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9668431282043457,
+      "learning_rate": 7.166483918128656e-06,
+      "loss": 1.3328,
+      "step": 14045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.794480323791504,
+      "learning_rate": 7.1655701754385974e-06,
+      "loss": 1.1949,
+      "step": 14046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.458418130874634,
+      "learning_rate": 7.164656432748539e-06,
+      "loss": 1.6948,
+      "step": 14047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2612247467041016,
+      "learning_rate": 7.16374269005848e-06,
+      "loss": 1.0457,
+      "step": 14048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6419618129730225,
+      "learning_rate": 7.162828947368422e-06,
+      "loss": 1.5628,
+      "step": 14049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2916951179504395,
+      "learning_rate": 7.161915204678364e-06,
+      "loss": 1.549,
+      "step": 14050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4681594371795654,
+      "learning_rate": 7.161001461988305e-06,
+      "loss": 1.3945,
+      "step": 14051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.293952703475952,
+      "learning_rate": 7.1600877192982465e-06,
+      "loss": 1.4518,
+      "step": 14052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5961155891418457,
+      "learning_rate": 7.159173976608188e-06,
+      "loss": 1.2904,
+      "step": 14053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.416638135910034,
+      "learning_rate": 7.158260233918129e-06,
+      "loss": 1.5094,
+      "step": 14054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7915666103363037,
+      "learning_rate": 7.157346491228071e-06,
+      "loss": 1.6841,
+      "step": 14055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6035354137420654,
+      "learning_rate": 7.156432748538013e-06,
+      "loss": 1.5797,
+      "step": 14056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7670750617980957,
+      "learning_rate": 7.155519005847954e-06,
+      "loss": 1.6128,
+      "step": 14057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.424424648284912,
+      "learning_rate": 7.154605263157896e-06,
+      "loss": 1.4332,
+      "step": 14058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8816654682159424,
+      "learning_rate": 7.153691520467837e-06,
+      "loss": 1.735,
+      "step": 14059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8639332056045532,
+      "learning_rate": 7.152777777777778e-06,
+      "loss": 1.0617,
+      "step": 14060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5535078048706055,
+      "learning_rate": 7.15186403508772e-06,
+      "loss": 1.5925,
+      "step": 14061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4095089435577393,
+      "learning_rate": 7.150950292397661e-06,
+      "loss": 1.3168,
+      "step": 14062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6132848262786865,
+      "learning_rate": 7.150036549707603e-06,
+      "loss": 1.5999,
+      "step": 14063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.579718589782715,
+      "learning_rate": 7.149122807017545e-06,
+      "loss": 1.5809,
+      "step": 14064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.03071928024292,
+      "learning_rate": 7.148209064327486e-06,
+      "loss": 1.275,
+      "step": 14065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7349836826324463,
+      "learning_rate": 7.1472953216374275e-06,
+      "loss": 1.872,
+      "step": 14066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7315073013305664,
+      "learning_rate": 7.146381578947369e-06,
+      "loss": 1.3439,
+      "step": 14067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8093626499176025,
+      "learning_rate": 7.14546783625731e-06,
+      "loss": 1.5692,
+      "step": 14068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4663496017456055,
+      "learning_rate": 7.144554093567252e-06,
+      "loss": 1.5513,
+      "step": 14069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.233130693435669,
+      "learning_rate": 7.143640350877194e-06,
+      "loss": 1.1986,
+      "step": 14070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6866986751556396,
+      "learning_rate": 7.142726608187135e-06,
+      "loss": 1.605,
+      "step": 14071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0207345485687256,
+      "learning_rate": 7.1418128654970766e-06,
+      "loss": 2.0137,
+      "step": 14072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.310842752456665,
+      "learning_rate": 7.140899122807018e-06,
+      "loss": 1.6429,
+      "step": 14073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1735894680023193,
+      "learning_rate": 7.139985380116959e-06,
+      "loss": 1.2052,
+      "step": 14074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.692286491394043,
+      "learning_rate": 7.139071637426901e-06,
+      "loss": 1.4486,
+      "step": 14075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8181822299957275,
+      "learning_rate": 7.138157894736842e-06,
+      "loss": 1.5994,
+      "step": 14076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6167168617248535,
+      "learning_rate": 7.137244152046784e-06,
+      "loss": 1.4661,
+      "step": 14077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.381962776184082,
+      "learning_rate": 7.136330409356726e-06,
+      "loss": 1.3305,
+      "step": 14078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2585840225219727,
+      "learning_rate": 7.135416666666667e-06,
+      "loss": 1.2915,
+      "step": 14079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.640108108520508,
+      "learning_rate": 7.134502923976608e-06,
+      "loss": 1.6629,
+      "step": 14080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7246761322021484,
+      "learning_rate": 7.13358918128655e-06,
+      "loss": 1.6459,
+      "step": 14081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.657837390899658,
+      "learning_rate": 7.132675438596491e-06,
+      "loss": 1.703,
+      "step": 14082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6108615398406982,
+      "learning_rate": 7.1317616959064325e-06,
+      "loss": 1.1364,
+      "step": 14083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5215582847595215,
+      "learning_rate": 7.130847953216375e-06,
+      "loss": 1.4886,
+      "step": 14084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.446086883544922,
+      "learning_rate": 7.129934210526316e-06,
+      "loss": 1.1519,
+      "step": 14085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9456653594970703,
+      "learning_rate": 7.1290204678362575e-06,
+      "loss": 1.9172,
+      "step": 14086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.234125852584839,
+      "learning_rate": 7.1281067251462e-06,
+      "loss": 1.0887,
+      "step": 14087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.707167625427246,
+      "learning_rate": 7.127192982456141e-06,
+      "loss": 1.7505,
+      "step": 14088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4158542156219482,
+      "learning_rate": 7.126279239766083e-06,
+      "loss": 1.3948,
+      "step": 14089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0164742469787598,
+      "learning_rate": 7.125365497076025e-06,
+      "loss": 1.9071,
+      "step": 14090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.313667058944702,
+      "learning_rate": 7.124451754385966e-06,
+      "loss": 1.4655,
+      "step": 14091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9936200380325317,
+      "learning_rate": 7.1235380116959074e-06,
+      "loss": 1.2182,
+      "step": 14092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6990668773651123,
+      "learning_rate": 7.122624269005849e-06,
+      "loss": 1.4521,
+      "step": 14093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.29186749458313,
+      "learning_rate": 7.12171052631579e-06,
+      "loss": 1.4965,
+      "step": 14094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8181984424591064,
+      "learning_rate": 7.1207967836257316e-06,
+      "loss": 1.5191,
+      "step": 14095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.93888258934021,
+      "learning_rate": 7.119883040935674e-06,
+      "loss": 1.2823,
+      "step": 14096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.618234872817993,
+      "learning_rate": 7.118969298245615e-06,
+      "loss": 1.5293,
+      "step": 14097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7110354900360107,
+      "learning_rate": 7.1180555555555565e-06,
+      "loss": 1.5033,
+      "step": 14098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.663766860961914,
+      "learning_rate": 7.117141812865498e-06,
+      "loss": 1.5551,
+      "step": 14099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4840569496154785,
+      "learning_rate": 7.116228070175439e-06,
+      "loss": 1.5888,
+      "step": 14100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.792351007461548,
+      "learning_rate": 7.115314327485381e-06,
+      "loss": 1.7181,
+      "step": 14101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.321403980255127,
+      "learning_rate": 7.114400584795322e-06,
+      "loss": 1.4381,
+      "step": 14102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5652871131896973,
+      "learning_rate": 7.113486842105264e-06,
+      "loss": 1.6271,
+      "step": 14103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7505850791931152,
+      "learning_rate": 7.112573099415206e-06,
+      "loss": 1.7001,
+      "step": 14104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.816077947616577,
+      "learning_rate": 7.111659356725147e-06,
+      "loss": 1.6364,
+      "step": 14105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.051987886428833,
+      "learning_rate": 7.110745614035088e-06,
+      "loss": 1.6144,
+      "step": 14106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4383914470672607,
+      "learning_rate": 7.10983187134503e-06,
+      "loss": 1.4829,
+      "step": 14107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.406402826309204,
+      "learning_rate": 7.108918128654971e-06,
+      "loss": 1.6806,
+      "step": 14108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.707249164581299,
+      "learning_rate": 7.1080043859649125e-06,
+      "loss": 1.4214,
+      "step": 14109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7876648902893066,
+      "learning_rate": 7.107090643274855e-06,
+      "loss": 1.6365,
+      "step": 14110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5147926807403564,
+      "learning_rate": 7.106176900584796e-06,
+      "loss": 1.5517,
+      "step": 14111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.435717821121216,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 1.5628,
+      "step": 14112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1000351905822754,
+      "learning_rate": 7.104349415204679e-06,
+      "loss": 1.5289,
+      "step": 14113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4948949813842773,
+      "learning_rate": 7.10343567251462e-06,
+      "loss": 1.287,
+      "step": 14114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.254757881164551,
+      "learning_rate": 7.102521929824562e-06,
+      "loss": 1.3649,
+      "step": 14115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.066514253616333,
+      "learning_rate": 7.101608187134503e-06,
+      "loss": 1.7532,
+      "step": 14116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.830186605453491,
+      "learning_rate": 7.100694444444445e-06,
+      "loss": 1.4977,
+      "step": 14117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.427887439727783,
+      "learning_rate": 7.0997807017543866e-06,
+      "loss": 1.5962,
+      "step": 14118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.31471848487854,
+      "learning_rate": 7.098866959064328e-06,
+      "loss": 1.3423,
+      "step": 14119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.826159715652466,
+      "learning_rate": 7.097953216374269e-06,
+      "loss": 1.7111,
+      "step": 14120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.480272054672241,
+      "learning_rate": 7.097039473684211e-06,
+      "loss": 1.5131,
+      "step": 14121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.059375286102295,
+      "learning_rate": 7.096125730994152e-06,
+      "loss": 1.6783,
+      "step": 14122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.759354829788208,
+      "learning_rate": 7.0952119883040934e-06,
+      "loss": 1.7284,
+      "step": 14123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3170456886291504,
+      "learning_rate": 7.094298245614036e-06,
+      "loss": 1.2756,
+      "step": 14124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.987011432647705,
+      "learning_rate": 7.093384502923977e-06,
+      "loss": 1.4182,
+      "step": 14125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.774940013885498,
+      "learning_rate": 7.092470760233918e-06,
+      "loss": 1.3934,
+      "step": 14126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8384547233581543,
+      "learning_rate": 7.09155701754386e-06,
+      "loss": 1.5452,
+      "step": 14127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.229053020477295,
+      "learning_rate": 7.090643274853801e-06,
+      "loss": 1.3873,
+      "step": 14128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6711578369140625,
+      "learning_rate": 7.089729532163744e-06,
+      "loss": 1.317,
+      "step": 14129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.836667537689209,
+      "learning_rate": 7.088815789473686e-06,
+      "loss": 1.5363,
+      "step": 14130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.34228515625,
+      "learning_rate": 7.087902046783627e-06,
+      "loss": 1.3155,
+      "step": 14131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4467501640319824,
+      "learning_rate": 7.086988304093568e-06,
+      "loss": 1.7193,
+      "step": 14132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6966216564178467,
+      "learning_rate": 7.08607456140351e-06,
+      "loss": 1.3272,
+      "step": 14133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.871742010116577,
+      "learning_rate": 7.085160818713451e-06,
+      "loss": 1.4566,
+      "step": 14134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0789718627929688,
+      "learning_rate": 7.0842470760233925e-06,
+      "loss": 1.6019,
+      "step": 14135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.077878475189209,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 1.5829,
+      "step": 14136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3064587116241455,
+      "learning_rate": 7.082419590643276e-06,
+      "loss": 1.1485,
+      "step": 14137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6935558319091797,
+      "learning_rate": 7.0815058479532174e-06,
+      "loss": 1.4216,
+      "step": 14138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9649198055267334,
+      "learning_rate": 7.080592105263159e-06,
+      "loss": 0.9462,
+      "step": 14139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6301522254943848,
+      "learning_rate": 7.0796783625731e-06,
+      "loss": 1.3402,
+      "step": 14140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.11368727684021,
+      "learning_rate": 7.0787646198830416e-06,
+      "loss": 1.2013,
+      "step": 14141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0993993282318115,
+      "learning_rate": 7.077850877192983e-06,
+      "loss": 1.77,
+      "step": 14142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8131844997406006,
+      "learning_rate": 7.076937134502925e-06,
+      "loss": 1.7318,
+      "step": 14143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.387688636779785,
+      "learning_rate": 7.0760233918128665e-06,
+      "loss": 1.2786,
+      "step": 14144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9918735027313232,
+      "learning_rate": 7.075109649122808e-06,
+      "loss": 1.7207,
+      "step": 14145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.911142349243164,
+      "learning_rate": 7.074195906432749e-06,
+      "loss": 1.8254,
+      "step": 14146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.615304946899414,
+      "learning_rate": 7.073282163742691e-06,
+      "loss": 1.6921,
+      "step": 14147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4799652099609375,
+      "learning_rate": 7.072368421052632e-06,
+      "loss": 1.5053,
+      "step": 14148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9384520053863525,
+      "learning_rate": 7.071454678362573e-06,
+      "loss": 1.5949,
+      "step": 14149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.360300302505493,
+      "learning_rate": 7.070540935672516e-06,
+      "loss": 1.5576,
+      "step": 14150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.571281909942627,
+      "learning_rate": 7.069627192982457e-06,
+      "loss": 1.3646,
+      "step": 14151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2489349842071533,
+      "learning_rate": 7.068713450292398e-06,
+      "loss": 1.2552,
+      "step": 14152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.674961566925049,
+      "learning_rate": 7.06779970760234e-06,
+      "loss": 1.526,
+      "step": 14153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.753972053527832,
+      "learning_rate": 7.066885964912281e-06,
+      "loss": 1.5457,
+      "step": 14154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7880795001983643,
+      "learning_rate": 7.0659722222222225e-06,
+      "loss": 1.947,
+      "step": 14155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0384204387664795,
+      "learning_rate": 7.065058479532164e-06,
+      "loss": 1.8474,
+      "step": 14156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7229743003845215,
+      "learning_rate": 7.064144736842105e-06,
+      "loss": 1.5651,
+      "step": 14157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.655272960662842,
+      "learning_rate": 7.0632309941520475e-06,
+      "loss": 1.4942,
+      "step": 14158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5661380290985107,
+      "learning_rate": 7.062317251461989e-06,
+      "loss": 1.2998,
+      "step": 14159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3389523029327393,
+      "learning_rate": 7.06140350877193e-06,
+      "loss": 1.2984,
+      "step": 14160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.936708927154541,
+      "learning_rate": 7.060489766081872e-06,
+      "loss": 1.5594,
+      "step": 14161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.214027166366577,
+      "learning_rate": 7.059576023391813e-06,
+      "loss": 1.3208,
+      "step": 14162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4622116088867188,
+      "learning_rate": 7.058662280701754e-06,
+      "loss": 1.544,
+      "step": 14163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.495795726776123,
+      "learning_rate": 7.057748538011696e-06,
+      "loss": 1.5124,
+      "step": 14164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9935953617095947,
+      "learning_rate": 7.056834795321638e-06,
+      "loss": 0.9998,
+      "step": 14165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4136202335357666,
+      "learning_rate": 7.055921052631579e-06,
+      "loss": 1.2267,
+      "step": 14166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.913946509361267,
+      "learning_rate": 7.055007309941521e-06,
+      "loss": 1.149,
+      "step": 14167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0720345973968506,
+      "learning_rate": 7.054093567251462e-06,
+      "loss": 1.3907,
+      "step": 14168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4011216163635254,
+      "learning_rate": 7.0531798245614034e-06,
+      "loss": 1.4074,
+      "step": 14169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.385349750518799,
+      "learning_rate": 7.052266081871345e-06,
+      "loss": 1.4323,
+      "step": 14170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4533941745758057,
+      "learning_rate": 7.051352339181286e-06,
+      "loss": 1.682,
+      "step": 14171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1510517597198486,
+      "learning_rate": 7.050438596491229e-06,
+      "loss": 1.3824,
+      "step": 14172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.817197561264038,
+      "learning_rate": 7.049524853801171e-06,
+      "loss": 1.5328,
+      "step": 14173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.169631242752075,
+      "learning_rate": 7.048611111111112e-06,
+      "loss": 1.2481,
+      "step": 14174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5946731567382812,
+      "learning_rate": 7.047697368421053e-06,
+      "loss": 1.85,
+      "step": 14175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.418891191482544,
+      "learning_rate": 7.046783625730995e-06,
+      "loss": 1.4022,
+      "step": 14176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.23317813873291,
+      "learning_rate": 7.045869883040937e-06,
+      "loss": 1.1952,
+      "step": 14177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3918895721435547,
+      "learning_rate": 7.044956140350878e-06,
+      "loss": 1.2836,
+      "step": 14178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9252188205718994,
+      "learning_rate": 7.04404239766082e-06,
+      "loss": 1.7638,
+      "step": 14179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9909992218017578,
+      "learning_rate": 7.043128654970761e-06,
+      "loss": 1.4171,
+      "step": 14180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3691911697387695,
+      "learning_rate": 7.0422149122807025e-06,
+      "loss": 1.3925,
+      "step": 14181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.734081506729126,
+      "learning_rate": 7.041301169590644e-06,
+      "loss": 1.7754,
+      "step": 14182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8223018646240234,
+      "learning_rate": 7.040387426900585e-06,
+      "loss": 1.541,
+      "step": 14183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6392970085144043,
+      "learning_rate": 7.0394736842105274e-06,
+      "loss": 1.7695,
+      "step": 14184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.360370635986328,
+      "learning_rate": 7.038559941520469e-06,
+      "loss": 1.3329,
+      "step": 14185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5087409019470215,
+      "learning_rate": 7.03764619883041e-06,
+      "loss": 1.37,
+      "step": 14186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6288836002349854,
+      "learning_rate": 7.0367324561403516e-06,
+      "loss": 1.6782,
+      "step": 14187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.38313627243042,
+      "learning_rate": 7.035818713450293e-06,
+      "loss": 1.4443,
+      "step": 14188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.666604995727539,
+      "learning_rate": 7.034904970760234e-06,
+      "loss": 1.2918,
+      "step": 14189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6318647861480713,
+      "learning_rate": 7.033991228070176e-06,
+      "loss": 1.772,
+      "step": 14190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1895875930786133,
+      "learning_rate": 7.033077485380118e-06,
+      "loss": 1.3325,
+      "step": 14191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.345383644104004,
+      "learning_rate": 7.032163742690059e-06,
+      "loss": 1.3824,
+      "step": 14192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.486741065979004,
+      "learning_rate": 7.031250000000001e-06,
+      "loss": 1.5298,
+      "step": 14193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.232667922973633,
+      "learning_rate": 7.030336257309942e-06,
+      "loss": 1.1811,
+      "step": 14194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6288225650787354,
+      "learning_rate": 7.029422514619883e-06,
+      "loss": 1.6987,
+      "step": 14195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8181684017181396,
+      "learning_rate": 7.028508771929825e-06,
+      "loss": 1.4263,
+      "step": 14196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.115354299545288,
+      "learning_rate": 7.027595029239766e-06,
+      "loss": 1.4182,
+      "step": 14197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5908820629119873,
+      "learning_rate": 7.026681286549708e-06,
+      "loss": 1.4576,
+      "step": 14198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6239190101623535,
+      "learning_rate": 7.02576754385965e-06,
+      "loss": 1.7613,
+      "step": 14199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8976657390594482,
+      "learning_rate": 7.024853801169591e-06,
+      "loss": 1.4047,
+      "step": 14200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.377664566040039,
+      "learning_rate": 7.0239400584795325e-06,
+      "loss": 1.3341,
+      "step": 14201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4223408699035645,
+      "learning_rate": 7.023026315789474e-06,
+      "loss": 1.4174,
+      "step": 14202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1471614837646484,
+      "learning_rate": 7.022112573099415e-06,
+      "loss": 1.2792,
+      "step": 14203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1141109466552734,
+      "learning_rate": 7.021198830409357e-06,
+      "loss": 1.468,
+      "step": 14204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.565983533859253,
+      "learning_rate": 7.020285087719299e-06,
+      "loss": 1.2632,
+      "step": 14205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4433913230895996,
+      "learning_rate": 7.01937134502924e-06,
+      "loss": 1.5303,
+      "step": 14206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.441399335861206,
+      "learning_rate": 7.018457602339182e-06,
+      "loss": 1.2434,
+      "step": 14207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5845251083374023,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 1.6895,
+      "step": 14208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4085988998413086,
+      "learning_rate": 7.016630116959064e-06,
+      "loss": 1.6497,
+      "step": 14209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4239509105682373,
+      "learning_rate": 7.015716374269006e-06,
+      "loss": 1.3641,
+      "step": 14210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3038864135742188,
+      "learning_rate": 7.014802631578947e-06,
+      "loss": 1.2721,
+      "step": 14211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.799596071243286,
+      "learning_rate": 7.013888888888889e-06,
+      "loss": 1.4343,
+      "step": 14212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9712817668914795,
+      "learning_rate": 7.012975146198831e-06,
+      "loss": 1.7649,
+      "step": 14213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.45286226272583,
+      "learning_rate": 7.012061403508772e-06,
+      "loss": 1.6569,
+      "step": 14214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.168748140335083,
+      "learning_rate": 7.011147660818714e-06,
+      "loss": 1.1568,
+      "step": 14215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3087475299835205,
+      "learning_rate": 7.010233918128656e-06,
+      "loss": 1.3674,
+      "step": 14216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5278849601745605,
+      "learning_rate": 7.009320175438598e-06,
+      "loss": 1.6258,
+      "step": 14217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6398537158966064,
+      "learning_rate": 7.008406432748539e-06,
+      "loss": 1.5819,
+      "step": 14218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9844062328338623,
+      "learning_rate": 7.007492690058481e-06,
+      "loss": 1.403,
+      "step": 14219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.361933946609497,
+      "learning_rate": 7.006578947368422e-06,
+      "loss": 1.3768,
+      "step": 14220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6732873916625977,
+      "learning_rate": 7.005665204678363e-06,
+      "loss": 1.5498,
+      "step": 14221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.080747365951538,
+      "learning_rate": 7.004751461988305e-06,
+      "loss": 1.5384,
+      "step": 14222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3259103298187256,
+      "learning_rate": 7.003837719298246e-06,
+      "loss": 1.3997,
+      "step": 14223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.097701072692871,
+      "learning_rate": 7.002923976608188e-06,
+      "loss": 1.1586,
+      "step": 14224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.744210958480835,
+      "learning_rate": 7.00201023391813e-06,
+      "loss": 1.6457,
+      "step": 14225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.558159351348877,
+      "learning_rate": 7.001096491228071e-06,
+      "loss": 1.5476,
+      "step": 14226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3331401348114014,
+      "learning_rate": 7.0001827485380125e-06,
+      "loss": 1.4562,
+      "step": 14227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4338934421539307,
+      "learning_rate": 6.999269005847954e-06,
+      "loss": 1.1762,
+      "step": 14228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2631945610046387,
+      "learning_rate": 6.998355263157895e-06,
+      "loss": 1.3486,
+      "step": 14229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5148122310638428,
+      "learning_rate": 6.997441520467837e-06,
+      "loss": 1.5475,
+      "step": 14230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8823776245117188,
+      "learning_rate": 6.996527777777779e-06,
+      "loss": 1.3871,
+      "step": 14231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.965688943862915,
+      "learning_rate": 6.99561403508772e-06,
+      "loss": 1.639,
+      "step": 14232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8302125930786133,
+      "learning_rate": 6.9947002923976616e-06,
+      "loss": 1.7202,
+      "step": 14233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.562518358230591,
+      "learning_rate": 6.993786549707603e-06,
+      "loss": 1.987,
+      "step": 14234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6561245918273926,
+      "learning_rate": 6.992872807017544e-06,
+      "loss": 1.5911,
+      "step": 14235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.47251033782959,
+      "learning_rate": 6.991959064327486e-06,
+      "loss": 1.548,
+      "step": 14236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2241880893707275,
+      "learning_rate": 6.991045321637427e-06,
+      "loss": 1.3396,
+      "step": 14237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2410645484924316,
+      "learning_rate": 6.990131578947369e-06,
+      "loss": 1.3201,
+      "step": 14238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6318020820617676,
+      "learning_rate": 6.989217836257311e-06,
+      "loss": 1.4027,
+      "step": 14239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5934717655181885,
+      "learning_rate": 6.988304093567252e-06,
+      "loss": 1.6463,
+      "step": 14240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9753451347351074,
+      "learning_rate": 6.987390350877193e-06,
+      "loss": 1.4048,
+      "step": 14241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9236228466033936,
+      "learning_rate": 6.986476608187135e-06,
+      "loss": 1.3295,
+      "step": 14242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.596569061279297,
+      "learning_rate": 6.985562865497076e-06,
+      "loss": 1.5182,
+      "step": 14243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2808475494384766,
+      "learning_rate": 6.9846491228070175e-06,
+      "loss": 1.1707,
+      "step": 14244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.021263837814331,
+      "learning_rate": 6.98373538011696e-06,
+      "loss": 1.245,
+      "step": 14245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8252055644989014,
+      "learning_rate": 6.982821637426901e-06,
+      "loss": 1.5147,
+      "step": 14246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9130566120147705,
+      "learning_rate": 6.9819078947368425e-06,
+      "loss": 1.4512,
+      "step": 14247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3035221099853516,
+      "learning_rate": 6.980994152046784e-06,
+      "loss": 1.2833,
+      "step": 14248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1986494064331055,
+      "learning_rate": 6.980080409356725e-06,
+      "loss": 1.309,
+      "step": 14249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9541776180267334,
+      "learning_rate": 6.979166666666667e-06,
+      "loss": 1.795,
+      "step": 14250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.475128650665283,
+      "learning_rate": 6.978252923976608e-06,
+      "loss": 1.2825,
+      "step": 14251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0478177070617676,
+      "learning_rate": 6.97733918128655e-06,
+      "loss": 1.7671,
+      "step": 14252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.213411331176758,
+      "learning_rate": 6.976425438596492e-06,
+      "loss": 1.1887,
+      "step": 14253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1149330139160156,
+      "learning_rate": 6.975511695906433e-06,
+      "loss": 1.1914,
+      "step": 14254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6706790924072266,
+      "learning_rate": 6.974597953216374e-06,
+      "loss": 1.6407,
+      "step": 14255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0083415508270264,
+      "learning_rate": 6.973684210526316e-06,
+      "loss": 1.094,
+      "step": 14256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1053755283355713,
+      "learning_rate": 6.972770467836257e-06,
+      "loss": 1.2981,
+      "step": 14257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4633419513702393,
+      "learning_rate": 6.9718567251462e-06,
+      "loss": 1.4439,
+      "step": 14258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6516075134277344,
+      "learning_rate": 6.9709429824561415e-06,
+      "loss": 1.7113,
+      "step": 14259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1865317821502686,
+      "learning_rate": 6.970029239766083e-06,
+      "loss": 1.4892,
+      "step": 14260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.388993740081787,
+      "learning_rate": 6.969115497076024e-06,
+      "loss": 1.2881,
+      "step": 14261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7955987453460693,
+      "learning_rate": 6.968201754385966e-06,
+      "loss": 1.5227,
+      "step": 14262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.803252696990967,
+      "learning_rate": 6.967288011695907e-06,
+      "loss": 1.6414,
+      "step": 14263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4055016040802,
+      "learning_rate": 6.966374269005849e-06,
+      "loss": 1.5242,
+      "step": 14264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.680328845977783,
+      "learning_rate": 6.965460526315791e-06,
+      "loss": 1.8923,
+      "step": 14265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.392188310623169,
+      "learning_rate": 6.964546783625732e-06,
+      "loss": 1.3297,
+      "step": 14266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.253023147583008,
+      "learning_rate": 6.963633040935673e-06,
+      "loss": 1.6343,
+      "step": 14267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4534199237823486,
+      "learning_rate": 6.962719298245615e-06,
+      "loss": 1.5458,
+      "step": 14268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1882143020629883,
+      "learning_rate": 6.961805555555556e-06,
+      "loss": 1.3486,
+      "step": 14269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6266584396362305,
+      "learning_rate": 6.9608918128654975e-06,
+      "loss": 1.3706,
+      "step": 14270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.754741668701172,
+      "learning_rate": 6.95997807017544e-06,
+      "loss": 1.7483,
+      "step": 14271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1512460708618164,
+      "learning_rate": 6.959064327485381e-06,
+      "loss": 1.3152,
+      "step": 14272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4127492904663086,
+      "learning_rate": 6.9581505847953225e-06,
+      "loss": 1.3073,
+      "step": 14273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8700509071350098,
+      "learning_rate": 6.957236842105264e-06,
+      "loss": 1.7149,
+      "step": 14274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.452636957168579,
+      "learning_rate": 6.956323099415205e-06,
+      "loss": 1.368,
+      "step": 14275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7407400608062744,
+      "learning_rate": 6.955409356725147e-06,
+      "loss": 1.5089,
+      "step": 14276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0433342456817627,
+      "learning_rate": 6.954495614035088e-06,
+      "loss": 1.0815,
+      "step": 14277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.759399652481079,
+      "learning_rate": 6.95358187134503e-06,
+      "loss": 1.4773,
+      "step": 14278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5127785205841064,
+      "learning_rate": 6.9526681286549716e-06,
+      "loss": 1.4179,
+      "step": 14279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5416510105133057,
+      "learning_rate": 6.951754385964913e-06,
+      "loss": 1.6824,
+      "step": 14280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0165719985961914,
+      "learning_rate": 6.950840643274854e-06,
+      "loss": 2.1265,
+      "step": 14281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.07348370552063,
+      "learning_rate": 6.949926900584796e-06,
+      "loss": 1.7093,
+      "step": 14282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3589460849761963,
+      "learning_rate": 6.949013157894737e-06,
+      "loss": 1.119,
+      "step": 14283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4169504642486572,
+      "learning_rate": 6.9480994152046784e-06,
+      "loss": 1.4387,
+      "step": 14284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6596486568450928,
+      "learning_rate": 6.947185672514621e-06,
+      "loss": 1.4255,
+      "step": 14285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2842321395874023,
+      "learning_rate": 6.946271929824562e-06,
+      "loss": 1.6755,
+      "step": 14286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2342419624328613,
+      "learning_rate": 6.945358187134503e-06,
+      "loss": 1.3167,
+      "step": 14287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.063844680786133,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 1.7419,
+      "step": 14288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3308169841766357,
+      "learning_rate": 6.943530701754386e-06,
+      "loss": 1.474,
+      "step": 14289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4777755737304688,
+      "learning_rate": 6.9426169590643275e-06,
+      "loss": 1.2674,
+      "step": 14290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.363253116607666,
+      "learning_rate": 6.941703216374269e-06,
+      "loss": 1.4515,
+      "step": 14291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.239971399307251,
+      "learning_rate": 6.940789473684211e-06,
+      "loss": 1.3567,
+      "step": 14292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.582057237625122,
+      "learning_rate": 6.9398757309941525e-06,
+      "loss": 1.5138,
+      "step": 14293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.119708299636841,
+      "learning_rate": 6.938961988304094e-06,
+      "loss": 1.3316,
+      "step": 14294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9621480703353882,
+      "learning_rate": 6.938048245614035e-06,
+      "loss": 0.9573,
+      "step": 14295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.967427968978882,
+      "learning_rate": 6.937134502923977e-06,
+      "loss": 1.7148,
+      "step": 14296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8141489028930664,
+      "learning_rate": 6.936220760233918e-06,
+      "loss": 1.3244,
+      "step": 14297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.222385406494141,
+      "learning_rate": 6.935307017543859e-06,
+      "loss": 1.4363,
+      "step": 14298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.053962230682373,
+      "learning_rate": 6.934393274853802e-06,
+      "loss": 1.9538,
+      "step": 14299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8591415882110596,
+      "learning_rate": 6.933479532163744e-06,
+      "loss": 1.3066,
+      "step": 14300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4985132217407227,
+      "learning_rate": 6.932565789473685e-06,
+      "loss": 1.8786,
+      "step": 14301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0867154598236084,
+      "learning_rate": 6.9316520467836266e-06,
+      "loss": 1.309,
+      "step": 14302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.575284004211426,
+      "learning_rate": 6.930738304093568e-06,
+      "loss": 1.7019,
+      "step": 14303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1854305267333984,
+      "learning_rate": 6.92982456140351e-06,
+      "loss": 1.3235,
+      "step": 14304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4628920555114746,
+      "learning_rate": 6.9289108187134515e-06,
+      "loss": 1.4085,
+      "step": 14305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9714643955230713,
+      "learning_rate": 6.927997076023393e-06,
+      "loss": 1.7055,
+      "step": 14306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6903905868530273,
+      "learning_rate": 6.927083333333334e-06,
+      "loss": 1.5903,
+      "step": 14307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3665075302124023,
+      "learning_rate": 6.926169590643276e-06,
+      "loss": 1.6055,
+      "step": 14308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3129169940948486,
+      "learning_rate": 6.925255847953217e-06,
+      "loss": 1.3571,
+      "step": 14309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0828115940093994,
+      "learning_rate": 6.924342105263158e-06,
+      "loss": 1.273,
+      "step": 14310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.214435577392578,
+      "learning_rate": 6.923428362573101e-06,
+      "loss": 1.2939,
+      "step": 14311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1073451042175293,
+      "learning_rate": 6.922514619883042e-06,
+      "loss": 1.8834,
+      "step": 14312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9177086353302,
+      "learning_rate": 6.921600877192983e-06,
+      "loss": 1.6885,
+      "step": 14313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.432342290878296,
+      "learning_rate": 6.920687134502925e-06,
+      "loss": 1.3058,
+      "step": 14314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7939350605010986,
+      "learning_rate": 6.919773391812866e-06,
+      "loss": 1.1871,
+      "step": 14315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.668445348739624,
+      "learning_rate": 6.9188596491228075e-06,
+      "loss": 1.6711,
+      "step": 14316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.525214672088623,
+      "learning_rate": 6.917945906432749e-06,
+      "loss": 1.5514,
+      "step": 14317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3224964141845703,
+      "learning_rate": 6.917032163742691e-06,
+      "loss": 1.3272,
+      "step": 14318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7804601192474365,
+      "learning_rate": 6.9161184210526325e-06,
+      "loss": 1.5516,
+      "step": 14319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0972540378570557,
+      "learning_rate": 6.915204678362574e-06,
+      "loss": 0.9703,
+      "step": 14320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8275411128997803,
+      "learning_rate": 6.914290935672515e-06,
+      "loss": 1.6797,
+      "step": 14321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.531859874725342,
+      "learning_rate": 6.913377192982457e-06,
+      "loss": 1.3549,
+      "step": 14322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3901259899139404,
+      "learning_rate": 6.912463450292398e-06,
+      "loss": 1.504,
+      "step": 14323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4116222858428955,
+      "learning_rate": 6.911549707602339e-06,
+      "loss": 1.4456,
+      "step": 14324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4306724071502686,
+      "learning_rate": 6.9106359649122816e-06,
+      "loss": 1.2902,
+      "step": 14325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.235367774963379,
+      "learning_rate": 6.909722222222223e-06,
+      "loss": 1.3393,
+      "step": 14326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.228280544281006,
+      "learning_rate": 6.908808479532164e-06,
+      "loss": 1.68,
+      "step": 14327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.458664655685425,
+      "learning_rate": 6.907894736842106e-06,
+      "loss": 1.4871,
+      "step": 14328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7824864387512207,
+      "learning_rate": 6.906980994152047e-06,
+      "loss": 1.499,
+      "step": 14329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7533562183380127,
+      "learning_rate": 6.9060672514619884e-06,
+      "loss": 1.5072,
+      "step": 14330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3732264041900635,
+      "learning_rate": 6.90515350877193e-06,
+      "loss": 1.3722,
+      "step": 14331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8103926181793213,
+      "learning_rate": 6.904239766081872e-06,
+      "loss": 1.2449,
+      "step": 14332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5854220390319824,
+      "learning_rate": 6.903326023391813e-06,
+      "loss": 1.7904,
+      "step": 14333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9664833545684814,
+      "learning_rate": 6.902412280701755e-06,
+      "loss": 1.7619,
+      "step": 14334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1775431632995605,
+      "learning_rate": 6.901498538011696e-06,
+      "loss": 1.6988,
+      "step": 14335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.49252986907959,
+      "learning_rate": 6.9005847953216375e-06,
+      "loss": 1.6027,
+      "step": 14336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.492684841156006,
+      "learning_rate": 6.899671052631579e-06,
+      "loss": 1.5141,
+      "step": 14337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5530753135681152,
+      "learning_rate": 6.89875730994152e-06,
+      "loss": 1.2604,
+      "step": 14338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9037859439849854,
+      "learning_rate": 6.8978435672514625e-06,
+      "loss": 1.5728,
+      "step": 14339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2183351516723633,
+      "learning_rate": 6.896929824561404e-06,
+      "loss": 1.2766,
+      "step": 14340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.085171937942505,
+      "learning_rate": 6.896016081871345e-06,
+      "loss": 1.3393,
+      "step": 14341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.470792293548584,
+      "learning_rate": 6.895102339181287e-06,
+      "loss": 1.4076,
+      "step": 14342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.479464292526245,
+      "learning_rate": 6.894188596491229e-06,
+      "loss": 1.445,
+      "step": 14343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.698861598968506,
+      "learning_rate": 6.893274853801171e-06,
+      "loss": 1.8984,
+      "step": 14344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0931408405303955,
+      "learning_rate": 6.8923611111111124e-06,
+      "loss": 1.5691,
+      "step": 14345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.358308792114258,
+      "learning_rate": 6.891447368421054e-06,
+      "loss": 1.3009,
+      "step": 14346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.22593355178833,
+      "learning_rate": 6.890533625730995e-06,
+      "loss": 1.1689,
+      "step": 14347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2240333557128906,
+      "learning_rate": 6.8896198830409366e-06,
+      "loss": 1.2282,
+      "step": 14348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9441120624542236,
+      "learning_rate": 6.888706140350878e-06,
+      "loss": 1.1961,
+      "step": 14349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0168585777282715,
+      "learning_rate": 6.887792397660819e-06,
+      "loss": 1.5367,
+      "step": 14350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.465693473815918,
+      "learning_rate": 6.8868786549707615e-06,
+      "loss": 1.3493,
+      "step": 14351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.264699935913086,
+      "learning_rate": 6.885964912280703e-06,
+      "loss": 1.3318,
+      "step": 14352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3338184356689453,
+      "learning_rate": 6.885051169590644e-06,
+      "loss": 1.3354,
+      "step": 14353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7828710079193115,
+      "learning_rate": 6.884137426900586e-06,
+      "loss": 1.5727,
+      "step": 14354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2027246952056885,
+      "learning_rate": 6.883223684210527e-06,
+      "loss": 1.2239,
+      "step": 14355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.413785219192505,
+      "learning_rate": 6.882309941520468e-06,
+      "loss": 1.3267,
+      "step": 14356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5195839405059814,
+      "learning_rate": 6.88139619883041e-06,
+      "loss": 1.1595,
+      "step": 14357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.793752908706665,
+      "learning_rate": 6.880482456140352e-06,
+      "loss": 1.418,
+      "step": 14358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9908134937286377,
+      "learning_rate": 6.879568713450293e-06,
+      "loss": 1.7656,
+      "step": 14359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.379831314086914,
+      "learning_rate": 6.878654970760235e-06,
+      "loss": 1.1867,
+      "step": 14360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.522286891937256,
+      "learning_rate": 6.877741228070176e-06,
+      "loss": 1.4022,
+      "step": 14361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.681826591491699,
+      "learning_rate": 6.8768274853801175e-06,
+      "loss": 1.6772,
+      "step": 14362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2391490936279297,
+      "learning_rate": 6.875913742690059e-06,
+      "loss": 1.2137,
+      "step": 14363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.972017526626587,
+      "learning_rate": 6.875e-06,
+      "loss": 1.8855,
+      "step": 14364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4210445880889893,
+      "learning_rate": 6.8740862573099425e-06,
+      "loss": 1.5532,
+      "step": 14365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.264375686645508,
+      "learning_rate": 6.873172514619884e-06,
+      "loss": 1.2967,
+      "step": 14366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.776968479156494,
+      "learning_rate": 6.872258771929825e-06,
+      "loss": 1.7375,
+      "step": 14367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6927731037139893,
+      "learning_rate": 6.871345029239767e-06,
+      "loss": 1.6039,
+      "step": 14368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3818891048431396,
+      "learning_rate": 6.870431286549708e-06,
+      "loss": 1.3034,
+      "step": 14369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.672233819961548,
+      "learning_rate": 6.869517543859649e-06,
+      "loss": 1.6644,
+      "step": 14370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3433218002319336,
+      "learning_rate": 6.868603801169591e-06,
+      "loss": 1.2924,
+      "step": 14371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5983235836029053,
+      "learning_rate": 6.867690058479533e-06,
+      "loss": 1.7468,
+      "step": 14372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0515708923339844,
+      "learning_rate": 6.866776315789474e-06,
+      "loss": 1.3292,
+      "step": 14373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5008130073547363,
+      "learning_rate": 6.865862573099416e-06,
+      "loss": 1.2692,
+      "step": 14374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.376316547393799,
+      "learning_rate": 6.864948830409357e-06,
+      "loss": 1.3819,
+      "step": 14375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9931387901306152,
+      "learning_rate": 6.8640350877192984e-06,
+      "loss": 1.9021,
+      "step": 14376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.554548740386963,
+      "learning_rate": 6.86312134502924e-06,
+      "loss": 1.49,
+      "step": 14377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1973557472229004,
+      "learning_rate": 6.862207602339181e-06,
+      "loss": 1.4441,
+      "step": 14378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.146730422973633,
+      "learning_rate": 6.861293859649123e-06,
+      "loss": 1.3129,
+      "step": 14379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.222895860671997,
+      "learning_rate": 6.860380116959065e-06,
+      "loss": 1.2263,
+      "step": 14380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3102681636810303,
+      "learning_rate": 6.859466374269006e-06,
+      "loss": 1.1306,
+      "step": 14381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.936688184738159,
+      "learning_rate": 6.8585526315789475e-06,
+      "loss": 1.7486,
+      "step": 14382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.489166736602783,
+      "learning_rate": 6.857638888888889e-06,
+      "loss": 1.4332,
+      "step": 14383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.399117946624756,
+      "learning_rate": 6.85672514619883e-06,
+      "loss": 1.1289,
+      "step": 14384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.438331127166748,
+      "learning_rate": 6.855811403508772e-06,
+      "loss": 1.4367,
+      "step": 14385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8740100860595703,
+      "learning_rate": 6.854897660818715e-06,
+      "loss": 1.7337,
+      "step": 14386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.760155439376831,
+      "learning_rate": 6.853983918128656e-06,
+      "loss": 1.7353,
+      "step": 14387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.162440061569214,
+      "learning_rate": 6.8530701754385975e-06,
+      "loss": 1.3072,
+      "step": 14388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3045408725738525,
+      "learning_rate": 6.852156432748539e-06,
+      "loss": 1.4674,
+      "step": 14389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.969203472137451,
+      "learning_rate": 6.85124269005848e-06,
+      "loss": 1.7058,
+      "step": 14390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.699582576751709,
+      "learning_rate": 6.8503289473684224e-06,
+      "loss": 1.5902,
+      "step": 14391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.189479112625122,
+      "learning_rate": 6.849415204678364e-06,
+      "loss": 1.1294,
+      "step": 14392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8038578033447266,
+      "learning_rate": 6.848501461988305e-06,
+      "loss": 1.4501,
+      "step": 14393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4608583450317383,
+      "learning_rate": 6.8475877192982466e-06,
+      "loss": 1.3299,
+      "step": 14394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9005000591278076,
+      "learning_rate": 6.846673976608188e-06,
+      "loss": 1.4405,
+      "step": 14395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6312100887298584,
+      "learning_rate": 6.845760233918129e-06,
+      "loss": 1.655,
+      "step": 14396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.828734874725342,
+      "learning_rate": 6.844846491228071e-06,
+      "loss": 1.899,
+      "step": 14397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4857027530670166,
+      "learning_rate": 6.843932748538013e-06,
+      "loss": 1.4192,
+      "step": 14398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.733189344406128,
+      "learning_rate": 6.843019005847954e-06,
+      "loss": 1.3325,
+      "step": 14399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4510886669158936,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 1.3295,
+      "step": 14400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6929781436920166,
+      "learning_rate": 6.841191520467837e-06,
+      "loss": 1.6847,
+      "step": 14401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5774075984954834,
+      "learning_rate": 6.840277777777778e-06,
+      "loss": 1.5398,
+      "step": 14402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.536055088043213,
+      "learning_rate": 6.83936403508772e-06,
+      "loss": 1.6004,
+      "step": 14403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1235556602478027,
+      "learning_rate": 6.838450292397661e-06,
+      "loss": 1.7094,
+      "step": 14404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4474499225616455,
+      "learning_rate": 6.8375365497076025e-06,
+      "loss": 1.3852,
+      "step": 14405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.325742244720459,
+      "learning_rate": 6.836622807017545e-06,
+      "loss": 1.2081,
+      "step": 14406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.378176689147949,
+      "learning_rate": 6.835709064327486e-06,
+      "loss": 1.2435,
+      "step": 14407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6109979152679443,
+      "learning_rate": 6.8347953216374275e-06,
+      "loss": 1.9021,
+      "step": 14408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2500975131988525,
+      "learning_rate": 6.833881578947369e-06,
+      "loss": 1.1729,
+      "step": 14409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5624725818634033,
+      "learning_rate": 6.83296783625731e-06,
+      "loss": 1.2528,
+      "step": 14410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9618914127349854,
+      "learning_rate": 6.832054093567252e-06,
+      "loss": 1.204,
+      "step": 14411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.855456590652466,
+      "learning_rate": 6.831140350877193e-06,
+      "loss": 1.3893,
+      "step": 14412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.267932891845703,
+      "learning_rate": 6.830226608187135e-06,
+      "loss": 1.1299,
+      "step": 14413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.110823154449463,
+      "learning_rate": 6.829312865497077e-06,
+      "loss": 1.1429,
+      "step": 14414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.911101818084717,
+      "learning_rate": 6.828399122807018e-06,
+      "loss": 1.5316,
+      "step": 14415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6383326053619385,
+      "learning_rate": 6.827485380116959e-06,
+      "loss": 1.6092,
+      "step": 14416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1403565406799316,
+      "learning_rate": 6.826571637426901e-06,
+      "loss": 1.2085,
+      "step": 14417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5136654376983643,
+      "learning_rate": 6.825657894736842e-06,
+      "loss": 1.5511,
+      "step": 14418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5984816551208496,
+      "learning_rate": 6.8247441520467835e-06,
+      "loss": 1.6103,
+      "step": 14419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2015316486358643,
+      "learning_rate": 6.823830409356726e-06,
+      "loss": 1.4745,
+      "step": 14420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.414513349533081,
+      "learning_rate": 6.822916666666667e-06,
+      "loss": 1.2854,
+      "step": 14421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.852659225463867,
+      "learning_rate": 6.8220029239766084e-06,
+      "loss": 1.4218,
+      "step": 14422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.674748659133911,
+      "learning_rate": 6.82108918128655e-06,
+      "loss": 1.5947,
+      "step": 14423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.577303409576416,
+      "learning_rate": 6.820175438596491e-06,
+      "loss": 1.7153,
+      "step": 14424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6559948921203613,
+      "learning_rate": 6.8192616959064326e-06,
+      "loss": 1.3427,
+      "step": 14425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.096658945083618,
+      "learning_rate": 6.818347953216374e-06,
+      "loss": 1.1986,
+      "step": 14426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4032411575317383,
+      "learning_rate": 6.817434210526316e-06,
+      "loss": 1.5113,
+      "step": 14427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.717979669570923,
+      "learning_rate": 6.8165204678362575e-06,
+      "loss": 1.5107,
+      "step": 14428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.103095531463623,
+      "learning_rate": 6.8156067251462e-06,
+      "loss": 1.6045,
+      "step": 14429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.24194073677063,
+      "learning_rate": 6.814692982456141e-06,
+      "loss": 1.5443,
+      "step": 14430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5454013347625732,
+      "learning_rate": 6.8137792397660825e-06,
+      "loss": 1.5131,
+      "step": 14431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.239240884780884,
+      "learning_rate": 6.812865497076025e-06,
+      "loss": 1.2336,
+      "step": 14432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6927366256713867,
+      "learning_rate": 6.811951754385966e-06,
+      "loss": 1.198,
+      "step": 14433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.069099187850952,
+      "learning_rate": 6.8110380116959075e-06,
+      "loss": 1.2312,
+      "step": 14434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.225236415863037,
+      "learning_rate": 6.810124269005849e-06,
+      "loss": 1.5248,
+      "step": 14435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6414294242858887,
+      "learning_rate": 6.80921052631579e-06,
+      "loss": 1.2774,
+      "step": 14436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6881940364837646,
+      "learning_rate": 6.808296783625732e-06,
+      "loss": 1.6937,
+      "step": 14437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2696454524993896,
+      "learning_rate": 6.807383040935673e-06,
+      "loss": 1.2641,
+      "step": 14438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.481722116470337,
+      "learning_rate": 6.806469298245615e-06,
+      "loss": 1.7129,
+      "step": 14439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6104252338409424,
+      "learning_rate": 6.8055555555555566e-06,
+      "loss": 1.3567,
+      "step": 14440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.674137830734253,
+      "learning_rate": 6.804641812865498e-06,
+      "loss": 1.3851,
+      "step": 14441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.864262342453003,
+      "learning_rate": 6.803728070175439e-06,
+      "loss": 1.956,
+      "step": 14442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7836191654205322,
+      "learning_rate": 6.802814327485381e-06,
+      "loss": 1.0109,
+      "step": 14443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5586507320404053,
+      "learning_rate": 6.801900584795322e-06,
+      "loss": 1.4859,
+      "step": 14444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1673998832702637,
+      "learning_rate": 6.8009868421052634e-06,
+      "loss": 1.4061,
+      "step": 14445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3764278888702393,
+      "learning_rate": 6.800073099415206e-06,
+      "loss": 1.776,
+      "step": 14446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.692734479904175,
+      "learning_rate": 6.799159356725147e-06,
+      "loss": 1.836,
+      "step": 14447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.80159592628479,
+      "learning_rate": 6.798245614035088e-06,
+      "loss": 1.2949,
+      "step": 14448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9039525985717773,
+      "learning_rate": 6.79733187134503e-06,
+      "loss": 1.7296,
+      "step": 14449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4172921180725098,
+      "learning_rate": 6.796418128654971e-06,
+      "loss": 1.4416,
+      "step": 14450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.084803819656372,
+      "learning_rate": 6.7955043859649125e-06,
+      "loss": 1.1068,
+      "step": 14451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.420339584350586,
+      "learning_rate": 6.794590643274854e-06,
+      "loss": 1.5858,
+      "step": 14452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.41259503364563,
+      "learning_rate": 6.793676900584796e-06,
+      "loss": 1.4166,
+      "step": 14453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3915774822235107,
+      "learning_rate": 6.7927631578947375e-06,
+      "loss": 1.3533,
+      "step": 14454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.998718500137329,
+      "learning_rate": 6.791849415204679e-06,
+      "loss": 1.9025,
+      "step": 14455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2963931560516357,
+      "learning_rate": 6.79093567251462e-06,
+      "loss": 1.342,
+      "step": 14456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6450283527374268,
+      "learning_rate": 6.790021929824562e-06,
+      "loss": 1.7666,
+      "step": 14457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.27350115776062,
+      "learning_rate": 6.789108187134503e-06,
+      "loss": 1.3367,
+      "step": 14458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0005130767822266,
+      "learning_rate": 6.788194444444444e-06,
+      "loss": 1.7598,
+      "step": 14459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8163259029388428,
+      "learning_rate": 6.787280701754387e-06,
+      "loss": 1.6764,
+      "step": 14460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0294113159179688,
+      "learning_rate": 6.786366959064328e-06,
+      "loss": 1.1652,
+      "step": 14461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0204005241394043,
+      "learning_rate": 6.785453216374269e-06,
+      "loss": 1.4065,
+      "step": 14462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5452373027801514,
+      "learning_rate": 6.784539473684211e-06,
+      "loss": 1.6158,
+      "step": 14463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.224978446960449,
+      "learning_rate": 6.783625730994152e-06,
+      "loss": 1.3812,
+      "step": 14464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2160871028900146,
+      "learning_rate": 6.7827119883040935e-06,
+      "loss": 1.4708,
+      "step": 14465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9034969806671143,
+      "learning_rate": 6.781798245614035e-06,
+      "loss": 1.5769,
+      "step": 14466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3656935691833496,
+      "learning_rate": 6.780884502923977e-06,
+      "loss": 1.3113,
+      "step": 14467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7995657920837402,
+      "learning_rate": 6.7799707602339184e-06,
+      "loss": 1.7158,
+      "step": 14468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.765649676322937,
+      "learning_rate": 6.77905701754386e-06,
+      "loss": 1.0338,
+      "step": 14469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.958346128463745,
+      "learning_rate": 6.778143274853801e-06,
+      "loss": 1.7689,
+      "step": 14470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8356285095214844,
+      "learning_rate": 6.777229532163743e-06,
+      "loss": 1.8086,
+      "step": 14471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5363147258758545,
+      "learning_rate": 6.776315789473686e-06,
+      "loss": 1.5711,
+      "step": 14472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6301565170288086,
+      "learning_rate": 6.775402046783627e-06,
+      "loss": 1.7781,
+      "step": 14473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6324284076690674,
+      "learning_rate": 6.774488304093568e-06,
+      "loss": 1.4838,
+      "step": 14474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.189835786819458,
+      "learning_rate": 6.77357456140351e-06,
+      "loss": 1.4389,
+      "step": 14475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.238455295562744,
+      "learning_rate": 6.772660818713451e-06,
+      "loss": 1.3394,
+      "step": 14476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7234697341918945,
+      "learning_rate": 6.7717470760233925e-06,
+      "loss": 1.433,
+      "step": 14477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.773897647857666,
+      "learning_rate": 6.770833333333334e-06,
+      "loss": 1.7495,
+      "step": 14478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8850066661834717,
+      "learning_rate": 6.769919590643276e-06,
+      "loss": 1.6333,
+      "step": 14479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5994725227355957,
+      "learning_rate": 6.7690058479532175e-06,
+      "loss": 1.8249,
+      "step": 14480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7137575149536133,
+      "learning_rate": 6.768092105263159e-06,
+      "loss": 1.5053,
+      "step": 14481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1296346187591553,
+      "learning_rate": 6.7671783625731e-06,
+      "loss": 1.2093,
+      "step": 14482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.650756597518921,
+      "learning_rate": 6.766264619883042e-06,
+      "loss": 1.6908,
+      "step": 14483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.497084379196167,
+      "learning_rate": 6.765350877192983e-06,
+      "loss": 1.2923,
+      "step": 14484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7154979705810547,
+      "learning_rate": 6.764437134502924e-06,
+      "loss": 1.5563,
+      "step": 14485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.797738790512085,
+      "learning_rate": 6.7635233918128666e-06,
+      "loss": 1.6472,
+      "step": 14486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.352318048477173,
+      "learning_rate": 6.762609649122808e-06,
+      "loss": 1.4711,
+      "step": 14487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.165083646774292,
+      "learning_rate": 6.761695906432749e-06,
+      "loss": 1.5441,
+      "step": 14488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.862417459487915,
+      "learning_rate": 6.760782163742691e-06,
+      "loss": 1.1741,
+      "step": 14489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.592264413833618,
+      "learning_rate": 6.759868421052632e-06,
+      "loss": 1.3316,
+      "step": 14490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.930872917175293,
+      "learning_rate": 6.7589546783625734e-06,
+      "loss": 1.5775,
+      "step": 14491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7401888370513916,
+      "learning_rate": 6.758040935672515e-06,
+      "loss": 1.7284,
+      "step": 14492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.413531541824341,
+      "learning_rate": 6.757127192982457e-06,
+      "loss": 1.5083,
+      "step": 14493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.612398862838745,
+      "learning_rate": 6.756213450292398e-06,
+      "loss": 1.7386,
+      "step": 14494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7826988697052,
+      "learning_rate": 6.75529970760234e-06,
+      "loss": 1.8113,
+      "step": 14495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.087703227996826,
+      "learning_rate": 6.754385964912281e-06,
+      "loss": 2.0471,
+      "step": 14496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8864479064941406,
+      "learning_rate": 6.7534722222222225e-06,
+      "loss": 1.8912,
+      "step": 14497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3287713527679443,
+      "learning_rate": 6.752558479532164e-06,
+      "loss": 1.508,
+      "step": 14498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2421231269836426,
+      "learning_rate": 6.751644736842105e-06,
+      "loss": 1.3068,
+      "step": 14499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2338156700134277,
+      "learning_rate": 6.7507309941520475e-06,
+      "loss": 1.9109,
+      "step": 14500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.728672981262207,
+      "learning_rate": 6.749817251461989e-06,
+      "loss": 1.625,
+      "step": 14501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7407593727111816,
+      "learning_rate": 6.74890350877193e-06,
+      "loss": 1.6317,
+      "step": 14502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1762516498565674,
+      "learning_rate": 6.747989766081872e-06,
+      "loss": 1.1315,
+      "step": 14503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5735814571380615,
+      "learning_rate": 6.747076023391813e-06,
+      "loss": 1.3308,
+      "step": 14504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2733519077301025,
+      "learning_rate": 6.746162280701754e-06,
+      "loss": 1.9486,
+      "step": 14505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.334127187728882,
+      "learning_rate": 6.745248538011696e-06,
+      "loss": 1.3719,
+      "step": 14506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3213284015655518,
+      "learning_rate": 6.744334795321638e-06,
+      "loss": 1.32,
+      "step": 14507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2447609901428223,
+      "learning_rate": 6.743421052631579e-06,
+      "loss": 1.4432,
+      "step": 14508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2282254695892334,
+      "learning_rate": 6.742507309941521e-06,
+      "loss": 1.2292,
+      "step": 14509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4790639877319336,
+      "learning_rate": 6.741593567251462e-06,
+      "loss": 1.4478,
+      "step": 14510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.102006196975708,
+      "learning_rate": 6.7406798245614035e-06,
+      "loss": 1.2899,
+      "step": 14511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2707903385162354,
+      "learning_rate": 6.739766081871345e-06,
+      "loss": 1.4566,
+      "step": 14512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8643665313720703,
+      "learning_rate": 6.738852339181286e-06,
+      "loss": 1.5681,
+      "step": 14513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.247178554534912,
+      "learning_rate": 6.737938596491229e-06,
+      "loss": 1.2686,
+      "step": 14514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.538116216659546,
+      "learning_rate": 6.737024853801171e-06,
+      "loss": 1.3704,
+      "step": 14515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5818347930908203,
+      "learning_rate": 6.736111111111112e-06,
+      "loss": 1.308,
+      "step": 14516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.750990629196167,
+      "learning_rate": 6.735197368421053e-06,
+      "loss": 1.4586,
+      "step": 14517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.999905586242676,
+      "learning_rate": 6.734283625730995e-06,
+      "loss": 1.2806,
+      "step": 14518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.431891918182373,
+      "learning_rate": 6.733369883040937e-06,
+      "loss": 1.3414,
+      "step": 14519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6970274448394775,
+      "learning_rate": 6.732456140350878e-06,
+      "loss": 1.708,
+      "step": 14520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.800008535385132,
+      "learning_rate": 6.73154239766082e-06,
+      "loss": 1.5882,
+      "step": 14521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9785239696502686,
+      "learning_rate": 6.730628654970761e-06,
+      "loss": 1.7516,
+      "step": 14522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3456921577453613,
+      "learning_rate": 6.7297149122807025e-06,
+      "loss": 1.2486,
+      "step": 14523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4818146228790283,
+      "learning_rate": 6.728801169590644e-06,
+      "loss": 1.3304,
+      "step": 14524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4768548011779785,
+      "learning_rate": 6.727887426900585e-06,
+      "loss": 1.2711,
+      "step": 14525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.6614532470703125,
+      "learning_rate": 6.7269736842105275e-06,
+      "loss": 1.3365,
+      "step": 14526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3832242488861084,
+      "learning_rate": 6.726059941520469e-06,
+      "loss": 1.4716,
+      "step": 14527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3021414279937744,
+      "learning_rate": 6.72514619883041e-06,
+      "loss": 1.1649,
+      "step": 14528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.778555154800415,
+      "learning_rate": 6.724232456140352e-06,
+      "loss": 1.9679,
+      "step": 14529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.133544921875,
+      "learning_rate": 6.723318713450293e-06,
+      "loss": 1.6772,
+      "step": 14530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.907883644104004,
+      "learning_rate": 6.722404970760234e-06,
+      "loss": 1.2056,
+      "step": 14531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.158795118331909,
+      "learning_rate": 6.721491228070176e-06,
+      "loss": 0.9786,
+      "step": 14532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.601590156555176,
+      "learning_rate": 6.720577485380118e-06,
+      "loss": 1.7768,
+      "step": 14533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3276827335357666,
+      "learning_rate": 6.719663742690059e-06,
+      "loss": 1.2033,
+      "step": 14534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7358691692352295,
+      "learning_rate": 6.718750000000001e-06,
+      "loss": 1.5037,
+      "step": 14535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5111947059631348,
+      "learning_rate": 6.717836257309942e-06,
+      "loss": 1.3899,
+      "step": 14536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5520081520080566,
+      "learning_rate": 6.7169225146198834e-06,
+      "loss": 1.5624,
+      "step": 14537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.359666109085083,
+      "learning_rate": 6.716008771929825e-06,
+      "loss": 1.4039,
+      "step": 14538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.639138698577881,
+      "learning_rate": 6.715095029239766e-06,
+      "loss": 1.1809,
+      "step": 14539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0713253021240234,
+      "learning_rate": 6.714181286549708e-06,
+      "loss": 1.9944,
+      "step": 14540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.224980115890503,
+      "learning_rate": 6.71326754385965e-06,
+      "loss": 1.8257,
+      "step": 14541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8239212036132812,
+      "learning_rate": 6.712353801169591e-06,
+      "loss": 1.599,
+      "step": 14542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4536187648773193,
+      "learning_rate": 6.7114400584795325e-06,
+      "loss": 1.3809,
+      "step": 14543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8814756870269775,
+      "learning_rate": 6.710526315789474e-06,
+      "loss": 0.9632,
+      "step": 14544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.628770112991333,
+      "learning_rate": 6.709612573099415e-06,
+      "loss": 1.7637,
+      "step": 14545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.485305070877075,
+      "learning_rate": 6.708698830409357e-06,
+      "loss": 1.544,
+      "step": 14546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.776951789855957,
+      "learning_rate": 6.707785087719299e-06,
+      "loss": 1.5404,
+      "step": 14547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7862186431884766,
+      "learning_rate": 6.70687134502924e-06,
+      "loss": 1.5541,
+      "step": 14548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7382278442382812,
+      "learning_rate": 6.705957602339182e-06,
+      "loss": 1.4754,
+      "step": 14549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.562185764312744,
+      "learning_rate": 6.705043859649123e-06,
+      "loss": 1.505,
+      "step": 14550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7318289279937744,
+      "learning_rate": 6.704130116959064e-06,
+      "loss": 1.5354,
+      "step": 14551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.450835943222046,
+      "learning_rate": 6.703216374269006e-06,
+      "loss": 1.3823,
+      "step": 14552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.898919105529785,
+      "learning_rate": 6.702302631578947e-06,
+      "loss": 1.4523,
+      "step": 14553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6017203330993652,
+      "learning_rate": 6.701388888888889e-06,
+      "loss": 1.4725,
+      "step": 14554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2096848487854004,
+      "learning_rate": 6.700475146198831e-06,
+      "loss": 2.0338,
+      "step": 14555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2680463790893555,
+      "learning_rate": 6.699561403508772e-06,
+      "loss": 1.0642,
+      "step": 14556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6214489936828613,
+      "learning_rate": 6.698647660818714e-06,
+      "loss": 1.504,
+      "step": 14557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5333240032196045,
+      "learning_rate": 6.697733918128656e-06,
+      "loss": 1.1758,
+      "step": 14558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.787813186645508,
+      "learning_rate": 6.696820175438598e-06,
+      "loss": 1.5183,
+      "step": 14559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1318368911743164,
+      "learning_rate": 6.695906432748539e-06,
+      "loss": 1.1342,
+      "step": 14560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4593255519866943,
+      "learning_rate": 6.694992690058481e-06,
+      "loss": 1.4883,
+      "step": 14561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5294973850250244,
+      "learning_rate": 6.694078947368422e-06,
+      "loss": 1.2995,
+      "step": 14562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5358335971832275,
+      "learning_rate": 6.693165204678363e-06,
+      "loss": 1.729,
+      "step": 14563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8885853290557861,
+      "learning_rate": 6.692251461988305e-06,
+      "loss": 1.1468,
+      "step": 14564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.996337652206421,
+      "learning_rate": 6.691337719298246e-06,
+      "loss": 1.6234,
+      "step": 14565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9159464836120605,
+      "learning_rate": 6.690423976608188e-06,
+      "loss": 1.3864,
+      "step": 14566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.244762420654297,
+      "learning_rate": 6.68951023391813e-06,
+      "loss": 1.4444,
+      "step": 14567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9790215492248535,
+      "learning_rate": 6.688596491228071e-06,
+      "loss": 1.4674,
+      "step": 14568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.353055953979492,
+      "learning_rate": 6.6876827485380125e-06,
+      "loss": 1.5135,
+      "step": 14569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0485291481018066,
+      "learning_rate": 6.686769005847954e-06,
+      "loss": 1.1295,
+      "step": 14570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.96700119972229,
+      "learning_rate": 6.685855263157895e-06,
+      "loss": 1.2796,
+      "step": 14571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.321364164352417,
+      "learning_rate": 6.684941520467837e-06,
+      "loss": 1.2324,
+      "step": 14572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0347073078155518,
+      "learning_rate": 6.684027777777779e-06,
+      "loss": 1.9981,
+      "step": 14573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.598501682281494,
+      "learning_rate": 6.68311403508772e-06,
+      "loss": 1.4092,
+      "step": 14574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.689351797103882,
+      "learning_rate": 6.682200292397662e-06,
+      "loss": 1.7252,
+      "step": 14575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4965808391571045,
+      "learning_rate": 6.681286549707603e-06,
+      "loss": 1.6476,
+      "step": 14576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1018190383911133,
+      "learning_rate": 6.680372807017544e-06,
+      "loss": 1.4603,
+      "step": 14577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.731502056121826,
+      "learning_rate": 6.679459064327486e-06,
+      "loss": 1.5302,
+      "step": 14578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.966339588165283,
+      "learning_rate": 6.678545321637427e-06,
+      "loss": 1.6404,
+      "step": 14579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.481682538986206,
+      "learning_rate": 6.677631578947369e-06,
+      "loss": 1.3894,
+      "step": 14580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5939512252807617,
+      "learning_rate": 6.676717836257311e-06,
+      "loss": 1.5299,
+      "step": 14581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6108145713806152,
+      "learning_rate": 6.675804093567252e-06,
+      "loss": 1.9648,
+      "step": 14582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.560673713684082,
+      "learning_rate": 6.6748903508771934e-06,
+      "loss": 1.574,
+      "step": 14583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.66291880607605,
+      "learning_rate": 6.673976608187135e-06,
+      "loss": 1.3273,
+      "step": 14584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7846596240997314,
+      "learning_rate": 6.673062865497076e-06,
+      "loss": 1.5082,
+      "step": 14585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4257099628448486,
+      "learning_rate": 6.6721491228070176e-06,
+      "loss": 1.3928,
+      "step": 14586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2780165672302246,
+      "learning_rate": 6.67123538011696e-06,
+      "loss": 1.365,
+      "step": 14587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2712783813476562,
+      "learning_rate": 6.670321637426901e-06,
+      "loss": 1.482,
+      "step": 14588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.299107074737549,
+      "learning_rate": 6.6694078947368425e-06,
+      "loss": 1.4639,
+      "step": 14589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2984700202941895,
+      "learning_rate": 6.668494152046784e-06,
+      "loss": 1.3963,
+      "step": 14590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5604536533355713,
+      "learning_rate": 6.667580409356725e-06,
+      "loss": 1.3874,
+      "step": 14591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.52105975151062,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.663,
+      "step": 14592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.557978391647339,
+      "learning_rate": 6.665752923976608e-06,
+      "loss": 1.5311,
+      "step": 14593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.139596939086914,
+      "learning_rate": 6.66483918128655e-06,
+      "loss": 1.3016,
+      "step": 14594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6882615089416504,
+      "learning_rate": 6.663925438596492e-06,
+      "loss": 1.524,
+      "step": 14595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.224091053009033,
+      "learning_rate": 6.663011695906433e-06,
+      "loss": 1.4246,
+      "step": 14596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6351475715637207,
+      "learning_rate": 6.662097953216374e-06,
+      "loss": 1.4307,
+      "step": 14597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6225433349609375,
+      "learning_rate": 6.661184210526316e-06,
+      "loss": 1.635,
+      "step": 14598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5376415252685547,
+      "learning_rate": 6.660270467836257e-06,
+      "loss": 1.5031,
+      "step": 14599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.431133508682251,
+      "learning_rate": 6.6593567251462e-06,
+      "loss": 1.3402,
+      "step": 14600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3879082202911377,
+      "learning_rate": 6.6584429824561416e-06,
+      "loss": 1.3313,
+      "step": 14601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9699599742889404,
+      "learning_rate": 6.657529239766083e-06,
+      "loss": 1.5123,
+      "step": 14602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9543778896331787,
+      "learning_rate": 6.656615497076024e-06,
+      "loss": 1.234,
+      "step": 14603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.64943528175354,
+      "learning_rate": 6.655701754385966e-06,
+      "loss": 1.5681,
+      "step": 14604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7110629081726074,
+      "learning_rate": 6.654788011695907e-06,
+      "loss": 1.593,
+      "step": 14605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.001469850540161,
+      "learning_rate": 6.653874269005849e-06,
+      "loss": 1.0221,
+      "step": 14606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3919124603271484,
+      "learning_rate": 6.652960526315791e-06,
+      "loss": 1.4246,
+      "step": 14607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.436846971511841,
+      "learning_rate": 6.652046783625732e-06,
+      "loss": 1.5142,
+      "step": 14608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.333897113800049,
+      "learning_rate": 6.651133040935673e-06,
+      "loss": 1.3025,
+      "step": 14609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.682896137237549,
+      "learning_rate": 6.650219298245615e-06,
+      "loss": 1.4112,
+      "step": 14610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.235452175140381,
+      "learning_rate": 6.649305555555556e-06,
+      "loss": 1.0762,
+      "step": 14611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6746623516082764,
+      "learning_rate": 6.6483918128654975e-06,
+      "loss": 1.4253,
+      "step": 14612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8083226680755615,
+      "learning_rate": 6.64747807017544e-06,
+      "loss": 1.6439,
+      "step": 14613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.504862070083618,
+      "learning_rate": 6.646564327485381e-06,
+      "loss": 1.3981,
+      "step": 14614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.770726203918457,
+      "learning_rate": 6.6456505847953225e-06,
+      "loss": 1.5742,
+      "step": 14615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.325779438018799,
+      "learning_rate": 6.644736842105264e-06,
+      "loss": 1.5673,
+      "step": 14616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.497856616973877,
+      "learning_rate": 6.643823099415205e-06,
+      "loss": 1.8084,
+      "step": 14617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.55122971534729,
+      "learning_rate": 6.642909356725147e-06,
+      "loss": 1.567,
+      "step": 14618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.24566388130188,
+      "learning_rate": 6.641995614035088e-06,
+      "loss": 1.3498,
+      "step": 14619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9842017889022827,
+      "learning_rate": 6.64108187134503e-06,
+      "loss": 1.3038,
+      "step": 14620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.326970338821411,
+      "learning_rate": 6.640168128654972e-06,
+      "loss": 1.2501,
+      "step": 14621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.826838731765747,
+      "learning_rate": 6.639254385964913e-06,
+      "loss": 1.3195,
+      "step": 14622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.407775640487671,
+      "learning_rate": 6.638340643274854e-06,
+      "loss": 1.7127,
+      "step": 14623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.075657606124878,
+      "learning_rate": 6.637426900584796e-06,
+      "loss": 1.1767,
+      "step": 14624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5636844635009766,
+      "learning_rate": 6.636513157894737e-06,
+      "loss": 1.4725,
+      "step": 14625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.33847713470459,
+      "learning_rate": 6.6355994152046785e-06,
+      "loss": 1.4087,
+      "step": 14626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.554288387298584,
+      "learning_rate": 6.63468567251462e-06,
+      "loss": 1.6264,
+      "step": 14627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5750925540924072,
+      "learning_rate": 6.633771929824562e-06,
+      "loss": 1.565,
+      "step": 14628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9210588932037354,
+      "learning_rate": 6.6328581871345034e-06,
+      "loss": 1.4594,
+      "step": 14629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8304567337036133,
+      "learning_rate": 6.631944444444445e-06,
+      "loss": 1.4316,
+      "step": 14630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.234487533569336,
+      "learning_rate": 6.631030701754386e-06,
+      "loss": 1.1099,
+      "step": 14631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.528120994567871,
+      "learning_rate": 6.6301169590643276e-06,
+      "loss": 1.5065,
+      "step": 14632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7096006870269775,
+      "learning_rate": 6.629203216374269e-06,
+      "loss": 1.6993,
+      "step": 14633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2757463455200195,
+      "learning_rate": 6.62828947368421e-06,
+      "loss": 1.4584,
+      "step": 14634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.583914041519165,
+      "learning_rate": 6.6273757309941525e-06,
+      "loss": 1.5951,
+      "step": 14635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.522073745727539,
+      "learning_rate": 6.626461988304094e-06,
+      "loss": 1.4108,
+      "step": 14636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9724318981170654,
+      "learning_rate": 6.625548245614035e-06,
+      "loss": 1.6324,
+      "step": 14637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1518802642822266,
+      "learning_rate": 6.624634502923977e-06,
+      "loss": 1.3344,
+      "step": 14638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2246975898742676,
+      "learning_rate": 6.623720760233918e-06,
+      "loss": 1.4698,
+      "step": 14639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8475515842437744,
+      "learning_rate": 6.622807017543859e-06,
+      "loss": 1.461,
+      "step": 14640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.703963041305542,
+      "learning_rate": 6.621893274853801e-06,
+      "loss": 1.7266,
+      "step": 14641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.080679178237915,
+      "learning_rate": 6.620979532163744e-06,
+      "loss": 1.3016,
+      "step": 14642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5939502716064453,
+      "learning_rate": 6.620065789473685e-06,
+      "loss": 1.5354,
+      "step": 14643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.951781988143921,
+      "learning_rate": 6.619152046783627e-06,
+      "loss": 1.7202,
+      "step": 14644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2881131172180176,
+      "learning_rate": 6.618238304093568e-06,
+      "loss": 1.538,
+      "step": 14645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3448448181152344,
+      "learning_rate": 6.61732456140351e-06,
+      "loss": 1.2906,
+      "step": 14646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.648446798324585,
+      "learning_rate": 6.6164108187134516e-06,
+      "loss": 1.5187,
+      "step": 14647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.605637550354004,
+      "learning_rate": 6.615497076023393e-06,
+      "loss": 1.4587,
+      "step": 14648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7230236530303955,
+      "learning_rate": 6.614583333333334e-06,
+      "loss": 1.4529,
+      "step": 14649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7504236698150635,
+      "learning_rate": 6.613669590643276e-06,
+      "loss": 1.5074,
+      "step": 14650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3613927364349365,
+      "learning_rate": 6.612755847953217e-06,
+      "loss": 1.9984,
+      "step": 14651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3603758811950684,
+      "learning_rate": 6.6118421052631584e-06,
+      "loss": 1.4861,
+      "step": 14652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546557903289795,
+      "learning_rate": 6.6109283625731e-06,
+      "loss": 1.5514,
+      "step": 14653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.584338903427124,
+      "learning_rate": 6.610014619883042e-06,
+      "loss": 1.5281,
+      "step": 14654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3838329315185547,
+      "learning_rate": 6.609100877192983e-06,
+      "loss": 1.2453,
+      "step": 14655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.372224807739258,
+      "learning_rate": 6.608187134502925e-06,
+      "loss": 1.7656,
+      "step": 14656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.588088274002075,
+      "learning_rate": 6.607273391812866e-06,
+      "loss": 1.6959,
+      "step": 14657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3881072998046875,
+      "learning_rate": 6.6063596491228075e-06,
+      "loss": 1.7433,
+      "step": 14658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.692646026611328,
+      "learning_rate": 6.605445906432749e-06,
+      "loss": 1.6085,
+      "step": 14659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1650071144104004,
+      "learning_rate": 6.60453216374269e-06,
+      "loss": 1.3723,
+      "step": 14660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7984697818756104,
+      "learning_rate": 6.6036184210526325e-06,
+      "loss": 1.6708,
+      "step": 14661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6446056365966797,
+      "learning_rate": 6.602704678362574e-06,
+      "loss": 1.184,
+      "step": 14662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5791306495666504,
+      "learning_rate": 6.601790935672515e-06,
+      "loss": 1.7067,
+      "step": 14663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0636250972747803,
+      "learning_rate": 6.600877192982457e-06,
+      "loss": 1.3479,
+      "step": 14664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.750701427459717,
+      "learning_rate": 6.599963450292398e-06,
+      "loss": 1.2745,
+      "step": 14665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5929532051086426,
+      "learning_rate": 6.599049707602339e-06,
+      "loss": 1.2394,
+      "step": 14666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5174288749694824,
+      "learning_rate": 6.598135964912281e-06,
+      "loss": 1.6614,
+      "step": 14667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4938089847564697,
+      "learning_rate": 6.597222222222223e-06,
+      "loss": 1.3643,
+      "step": 14668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.243741989135742,
+      "learning_rate": 6.596308479532164e-06,
+      "loss": 1.1378,
+      "step": 14669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5456318855285645,
+      "learning_rate": 6.595394736842106e-06,
+      "loss": 1.427,
+      "step": 14670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.25329852104187,
+      "learning_rate": 6.594480994152047e-06,
+      "loss": 1.271,
+      "step": 14671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2895541191101074,
+      "learning_rate": 6.5935672514619885e-06,
+      "loss": 1.5271,
+      "step": 14672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.113647937774658,
+      "learning_rate": 6.59265350877193e-06,
+      "loss": 1.3336,
+      "step": 14673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.672117233276367,
+      "learning_rate": 6.591739766081871e-06,
+      "loss": 1.5905,
+      "step": 14674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.072071075439453,
+      "learning_rate": 6.5908260233918134e-06,
+      "loss": 1.3623,
+      "step": 14675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9761745929718018,
+      "learning_rate": 6.589912280701755e-06,
+      "loss": 1.2111,
+      "step": 14676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.265591621398926,
+      "learning_rate": 6.588998538011696e-06,
+      "loss": 1.2878,
+      "step": 14677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.62027645111084,
+      "learning_rate": 6.5880847953216376e-06,
+      "loss": 1.6935,
+      "step": 14678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.665950059890747,
+      "learning_rate": 6.587171052631579e-06,
+      "loss": 1.6176,
+      "step": 14679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5864102840423584,
+      "learning_rate": 6.58625730994152e-06,
+      "loss": 1.6443,
+      "step": 14680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.740614414215088,
+      "learning_rate": 6.585343567251462e-06,
+      "loss": 1.5597,
+      "step": 14681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1057488918304443,
+      "learning_rate": 6.584429824561404e-06,
+      "loss": 1.6449,
+      "step": 14682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.835710287094116,
+      "learning_rate": 6.583516081871345e-06,
+      "loss": 1.8362,
+      "step": 14683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.559828519821167,
+      "learning_rate": 6.582602339181287e-06,
+      "loss": 1.3205,
+      "step": 14684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.185617208480835,
+      "learning_rate": 6.581688596491229e-06,
+      "loss": 1.1001,
+      "step": 14685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6642262935638428,
+      "learning_rate": 6.58077485380117e-06,
+      "loss": 1.5992,
+      "step": 14686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.438235282897949,
+      "learning_rate": 6.5798611111111125e-06,
+      "loss": 1.3279,
+      "step": 14687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.008518934249878,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 1.6406,
+      "step": 14688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2648582458496094,
+      "learning_rate": 6.578033625730995e-06,
+      "loss": 1.3858,
+      "step": 14689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.577807903289795,
+      "learning_rate": 6.577119883040937e-06,
+      "loss": 1.2144,
+      "step": 14690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.430809497833252,
+      "learning_rate": 6.576206140350878e-06,
+      "loss": 1.2314,
+      "step": 14691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.581596851348877,
+      "learning_rate": 6.575292397660819e-06,
+      "loss": 1.3913,
+      "step": 14692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.318208694458008,
+      "learning_rate": 6.574378654970761e-06,
+      "loss": 1.3637,
+      "step": 14693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.626650810241699,
+      "learning_rate": 6.573464912280703e-06,
+      "loss": 1.5752,
+      "step": 14694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.193416118621826,
+      "learning_rate": 6.572551169590644e-06,
+      "loss": 1.2555,
+      "step": 14695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.206271171569824,
+      "learning_rate": 6.571637426900586e-06,
+      "loss": 1.2914,
+      "step": 14696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3757262229919434,
+      "learning_rate": 6.570723684210527e-06,
+      "loss": 1.4978,
+      "step": 14697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.618363857269287,
+      "learning_rate": 6.5698099415204684e-06,
+      "loss": 1.2262,
+      "step": 14698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5831618309020996,
+      "learning_rate": 6.56889619883041e-06,
+      "loss": 1.5885,
+      "step": 14699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2189557552337646,
+      "learning_rate": 6.567982456140351e-06,
+      "loss": 1.2042,
+      "step": 14700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9269967079162598,
+      "learning_rate": 6.567068713450293e-06,
+      "loss": 1.7771,
+      "step": 14701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.704719066619873,
+      "learning_rate": 6.566154970760235e-06,
+      "loss": 1.7631,
+      "step": 14702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.170607566833496,
+      "learning_rate": 6.565241228070176e-06,
+      "loss": 1.3096,
+      "step": 14703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2076900005340576,
+      "learning_rate": 6.5643274853801175e-06,
+      "loss": 1.3123,
+      "step": 14704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.197917938232422,
+      "learning_rate": 6.563413742690059e-06,
+      "loss": 1.1909,
+      "step": 14705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1297543048858643,
+      "learning_rate": 6.5625e-06,
+      "loss": 1.1455,
+      "step": 14706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.209944009780884,
+      "learning_rate": 6.561586257309942e-06,
+      "loss": 1.2949,
+      "step": 14707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.606718063354492,
+      "learning_rate": 6.560672514619884e-06,
+      "loss": 1.5439,
+      "step": 14708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.69006609916687,
+      "learning_rate": 6.559758771929825e-06,
+      "loss": 1.5157,
+      "step": 14709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.14924955368042,
+      "learning_rate": 6.558845029239767e-06,
+      "loss": 1.3602,
+      "step": 14710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.513651132583618,
+      "learning_rate": 6.557931286549708e-06,
+      "loss": 1.5399,
+      "step": 14711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.442373037338257,
+      "learning_rate": 6.557017543859649e-06,
+      "loss": 1.3399,
+      "step": 14712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4571399688720703,
+      "learning_rate": 6.556103801169591e-06,
+      "loss": 1.2577,
+      "step": 14713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4034292697906494,
+      "learning_rate": 6.555190058479532e-06,
+      "loss": 1.2983,
+      "step": 14714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3101589679718018,
+      "learning_rate": 6.554276315789474e-06,
+      "loss": 1.5569,
+      "step": 14715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4824435710906982,
+      "learning_rate": 6.553362573099416e-06,
+      "loss": 1.5946,
+      "step": 14716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.678734064102173,
+      "learning_rate": 6.552448830409357e-06,
+      "loss": 1.6285,
+      "step": 14717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.533987283706665,
+      "learning_rate": 6.5515350877192985e-06,
+      "loss": 1.5258,
+      "step": 14718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5653765201568604,
+      "learning_rate": 6.55062134502924e-06,
+      "loss": 1.4567,
+      "step": 14719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.887328624725342,
+      "learning_rate": 6.549707602339181e-06,
+      "loss": 1.3686,
+      "step": 14720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2785890102386475,
+      "learning_rate": 6.548793859649123e-06,
+      "loss": 1.6239,
+      "step": 14721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.239189624786377,
+      "learning_rate": 6.547880116959065e-06,
+      "loss": 1.5042,
+      "step": 14722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3144311904907227,
+      "learning_rate": 6.546966374269006e-06,
+      "loss": 1.4735,
+      "step": 14723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9111084938049316,
+      "learning_rate": 6.5460526315789476e-06,
+      "loss": 1.597,
+      "step": 14724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0697009563446045,
+      "learning_rate": 6.545138888888889e-06,
+      "loss": 1.7045,
+      "step": 14725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.419598340988159,
+      "learning_rate": 6.54422514619883e-06,
+      "loss": 1.5107,
+      "step": 14726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.398088216781616,
+      "learning_rate": 6.543311403508772e-06,
+      "loss": 1.3387,
+      "step": 14727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7663979530334473,
+      "learning_rate": 6.542397660818715e-06,
+      "loss": 1.6247,
+      "step": 14728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6211984157562256,
+      "learning_rate": 6.541483918128656e-06,
+      "loss": 1.3741,
+      "step": 14729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2587931156158447,
+      "learning_rate": 6.5405701754385975e-06,
+      "loss": 1.5022,
+      "step": 14730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.350458860397339,
+      "learning_rate": 6.539656432748539e-06,
+      "loss": 1.3254,
+      "step": 14731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.153547763824463,
+      "learning_rate": 6.53874269005848e-06,
+      "loss": 1.2809,
+      "step": 14732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2643368244171143,
+      "learning_rate": 6.537828947368422e-06,
+      "loss": 1.4458,
+      "step": 14733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4162092208862305,
+      "learning_rate": 6.536915204678364e-06,
+      "loss": 1.2331,
+      "step": 14734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3287317752838135,
+      "learning_rate": 6.536001461988305e-06,
+      "loss": 1.1695,
+      "step": 14735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.201052188873291,
+      "learning_rate": 6.535087719298247e-06,
+      "loss": 1.2183,
+      "step": 14736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5545852184295654,
+      "learning_rate": 6.534173976608188e-06,
+      "loss": 1.7352,
+      "step": 14737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4518539905548096,
+      "learning_rate": 6.533260233918129e-06,
+      "loss": 1.1604,
+      "step": 14738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0220866203308105,
+      "learning_rate": 6.532346491228071e-06,
+      "loss": 2.0296,
+      "step": 14739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7904934883117676,
+      "learning_rate": 6.531432748538012e-06,
+      "loss": 1.4471,
+      "step": 14740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.033947229385376,
+      "learning_rate": 6.530519005847954e-06,
+      "loss": 1.3565,
+      "step": 14741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4376556873321533,
+      "learning_rate": 6.529605263157896e-06,
+      "loss": 1.2597,
+      "step": 14742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.65674090385437,
+      "learning_rate": 6.528691520467837e-06,
+      "loss": 1.5594,
+      "step": 14743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.470872163772583,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 1.3177,
+      "step": 14744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.482943534851074,
+      "learning_rate": 6.52686403508772e-06,
+      "loss": 1.6839,
+      "step": 14745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.46781849861145,
+      "learning_rate": 6.525950292397661e-06,
+      "loss": 1.2762,
+      "step": 14746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7335495948791504,
+      "learning_rate": 6.5250365497076026e-06,
+      "loss": 1.5324,
+      "step": 14747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4234328269958496,
+      "learning_rate": 6.524122807017545e-06,
+      "loss": 1.4624,
+      "step": 14748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8028130531311035,
+      "learning_rate": 6.523209064327486e-06,
+      "loss": 1.9016,
+      "step": 14749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3364779949188232,
+      "learning_rate": 6.5222953216374275e-06,
+      "loss": 1.4016,
+      "step": 14750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6423797607421875,
+      "learning_rate": 6.521381578947369e-06,
+      "loss": 1.6418,
+      "step": 14751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4815802574157715,
+      "learning_rate": 6.52046783625731e-06,
+      "loss": 1.4243,
+      "step": 14752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.229400634765625,
+      "learning_rate": 6.519554093567252e-06,
+      "loss": 1.2982,
+      "step": 14753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1455414295196533,
+      "learning_rate": 6.518640350877193e-06,
+      "loss": 1.2219,
+      "step": 14754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3902342319488525,
+      "learning_rate": 6.517726608187135e-06,
+      "loss": 1.4154,
+      "step": 14755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6651108264923096,
+      "learning_rate": 6.516812865497077e-06,
+      "loss": 1.5266,
+      "step": 14756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.278902053833008,
+      "learning_rate": 6.515899122807018e-06,
+      "loss": 1.307,
+      "step": 14757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.243476629257202,
+      "learning_rate": 6.514985380116959e-06,
+      "loss": 1.2885,
+      "step": 14758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.410517930984497,
+      "learning_rate": 6.514071637426901e-06,
+      "loss": 1.4055,
+      "step": 14759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5917139053344727,
+      "learning_rate": 6.513157894736842e-06,
+      "loss": 1.3836,
+      "step": 14760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8685262203216553,
+      "learning_rate": 6.5122441520467835e-06,
+      "loss": 1.6062,
+      "step": 14761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4654126167297363,
+      "learning_rate": 6.511330409356726e-06,
+      "loss": 1.6174,
+      "step": 14762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5817036628723145,
+      "learning_rate": 6.510416666666667e-06,
+      "loss": 1.468,
+      "step": 14763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5464370250701904,
+      "learning_rate": 6.5095029239766085e-06,
+      "loss": 1.415,
+      "step": 14764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.26904559135437,
+      "learning_rate": 6.50858918128655e-06,
+      "loss": 1.2638,
+      "step": 14765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.672995090484619,
+      "learning_rate": 6.507675438596491e-06,
+      "loss": 1.4457,
+      "step": 14766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9065043926239014,
+      "learning_rate": 6.506761695906433e-06,
+      "loss": 2.0512,
+      "step": 14767
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5887322425842285,
+      "learning_rate": 6.505847953216374e-06,
+      "loss": 1.6624,
+      "step": 14768
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.580928325653076,
+      "learning_rate": 6.504934210526316e-06,
+      "loss": 1.5226,
+      "step": 14769
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.859581708908081,
+      "learning_rate": 6.5040204678362576e-06,
+      "loss": 1.4172,
+      "step": 14770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6688079833984375,
+      "learning_rate": 6.5031067251462e-06,
+      "loss": 1.3573,
+      "step": 14771
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.053286552429199,
+      "learning_rate": 6.502192982456141e-06,
+      "loss": 1.3068,
+      "step": 14772
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3996734619140625,
+      "learning_rate": 6.5012792397660825e-06,
+      "loss": 1.347,
+      "step": 14773
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.320467710494995,
+      "learning_rate": 6.500365497076025e-06,
+      "loss": 1.2755,
+      "step": 14774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.473264694213867,
+      "learning_rate": 6.499451754385966e-06,
+      "loss": 1.4346,
+      "step": 14775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.524794578552246,
+      "learning_rate": 6.4985380116959075e-06,
+      "loss": 1.4477,
+      "step": 14776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.956052541732788,
+      "learning_rate": 6.497624269005849e-06,
+      "loss": 1.5208,
+      "step": 14777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7180237770080566,
+      "learning_rate": 6.49671052631579e-06,
+      "loss": 1.2577,
+      "step": 14778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.16202974319458,
+      "learning_rate": 6.495796783625732e-06,
+      "loss": 1.2238,
+      "step": 14779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8551230430603027,
+      "learning_rate": 6.494883040935673e-06,
+      "loss": 1.5966,
+      "step": 14780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3351335525512695,
+      "learning_rate": 6.493969298245615e-06,
+      "loss": 1.3705,
+      "step": 14781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.177454948425293,
+      "learning_rate": 6.493055555555557e-06,
+      "loss": 1.301,
+      "step": 14782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4930574893951416,
+      "learning_rate": 6.492141812865498e-06,
+      "loss": 1.5448,
+      "step": 14783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1580910682678223,
+      "learning_rate": 6.491228070175439e-06,
+      "loss": 1.8429,
+      "step": 14784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.617110252380371,
+      "learning_rate": 6.490314327485381e-06,
+      "loss": 1.549,
+      "step": 14785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9798810482025146,
+      "learning_rate": 6.489400584795322e-06,
+      "loss": 1.7494,
+      "step": 14786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.649153232574463,
+      "learning_rate": 6.4884868421052635e-06,
+      "loss": 1.3604,
+      "step": 14787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1040501594543457,
+      "learning_rate": 6.487573099415206e-06,
+      "loss": 1.6692,
+      "step": 14788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.339411973953247,
+      "learning_rate": 6.486659356725147e-06,
+      "loss": 1.2257,
+      "step": 14789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.17930269241333,
+      "learning_rate": 6.4857456140350884e-06,
+      "loss": 1.6501,
+      "step": 14790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.656200647354126,
+      "learning_rate": 6.48483187134503e-06,
+      "loss": 1.705,
+      "step": 14791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.125105857849121,
+      "learning_rate": 6.483918128654971e-06,
+      "loss": 1.6265,
+      "step": 14792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7629661560058594,
+      "learning_rate": 6.4830043859649126e-06,
+      "loss": 1.7288,
+      "step": 14793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2263286113739014,
+      "learning_rate": 6.482090643274854e-06,
+      "loss": 1.3699,
+      "step": 14794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.707425117492676,
+      "learning_rate": 6.481176900584796e-06,
+      "loss": 1.5855,
+      "step": 14795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5766048431396484,
+      "learning_rate": 6.4802631578947375e-06,
+      "loss": 1.556,
+      "step": 14796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2389354705810547,
+      "learning_rate": 6.479349415204679e-06,
+      "loss": 1.1347,
+      "step": 14797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.579630136489868,
+      "learning_rate": 6.47843567251462e-06,
+      "loss": 1.6822,
+      "step": 14798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.299008369445801,
+      "learning_rate": 6.477521929824562e-06,
+      "loss": 1.3031,
+      "step": 14799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.241121530532837,
+      "learning_rate": 6.476608187134503e-06,
+      "loss": 1.2516,
+      "step": 14800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.378028154373169,
+      "learning_rate": 6.475694444444444e-06,
+      "loss": 1.3537,
+      "step": 14801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.669821262359619,
+      "learning_rate": 6.474780701754387e-06,
+      "loss": 1.5509,
+      "step": 14802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5248093605041504,
+      "learning_rate": 6.473866959064328e-06,
+      "loss": 1.4534,
+      "step": 14803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.141580581665039,
+      "learning_rate": 6.472953216374269e-06,
+      "loss": 1.8738,
+      "step": 14804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9643335342407227,
+      "learning_rate": 6.472039473684211e-06,
+      "loss": 1.6388,
+      "step": 14805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.568187713623047,
+      "learning_rate": 6.471125730994152e-06,
+      "loss": 0.9013,
+      "step": 14806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4392800331115723,
+      "learning_rate": 6.4702119883040935e-06,
+      "loss": 1.2377,
+      "step": 14807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1271591186523438,
+      "learning_rate": 6.469298245614035e-06,
+      "loss": 1.0689,
+      "step": 14808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5915966033935547,
+      "learning_rate": 6.468384502923977e-06,
+      "loss": 1.3378,
+      "step": 14809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.651681423187256,
+      "learning_rate": 6.4674707602339185e-06,
+      "loss": 1.4783,
+      "step": 14810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2661380767822266,
+      "learning_rate": 6.46655701754386e-06,
+      "loss": 1.0731,
+      "step": 14811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7930519580841064,
+      "learning_rate": 6.465643274853801e-06,
+      "loss": 1.6088,
+      "step": 14812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6765137910842896,
+      "learning_rate": 6.4647295321637434e-06,
+      "loss": 0.9827,
+      "step": 14813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4675965309143066,
+      "learning_rate": 6.463815789473686e-06,
+      "loss": 1.5877,
+      "step": 14814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7582547664642334,
+      "learning_rate": 6.462902046783627e-06,
+      "loss": 1.6357,
+      "step": 14815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3706743717193604,
+      "learning_rate": 6.461988304093568e-06,
+      "loss": 1.9322,
+      "step": 14816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.380220890045166,
+      "learning_rate": 6.46107456140351e-06,
+      "loss": 1.1987,
+      "step": 14817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2759761810302734,
+      "learning_rate": 6.460160818713451e-06,
+      "loss": 1.4322,
+      "step": 14818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.184121608734131,
+      "learning_rate": 6.4592470760233925e-06,
+      "loss": 1.2429,
+      "step": 14819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.492797374725342,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 1.7134,
+      "step": 14820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.327212333679199,
+      "learning_rate": 6.457419590643276e-06,
+      "loss": 1.2597,
+      "step": 14821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.512922763824463,
+      "learning_rate": 6.4565058479532175e-06,
+      "loss": 1.4511,
+      "step": 14822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.335397720336914,
+      "learning_rate": 6.455592105263159e-06,
+      "loss": 1.2891,
+      "step": 14823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4894585609436035,
+      "learning_rate": 6.4546783625731e-06,
+      "loss": 1.4765,
+      "step": 14824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3073346614837646,
+      "learning_rate": 6.453764619883042e-06,
+      "loss": 1.2115,
+      "step": 14825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5652334690093994,
+      "learning_rate": 6.452850877192983e-06,
+      "loss": 1.72,
+      "step": 14826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8301761150360107,
+      "learning_rate": 6.451937134502924e-06,
+      "loss": 1.4395,
+      "step": 14827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.60931658744812,
+      "learning_rate": 6.451023391812867e-06,
+      "loss": 1.5355,
+      "step": 14828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.304656982421875,
+      "learning_rate": 6.450109649122808e-06,
+      "loss": 1.1404,
+      "step": 14829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.419736862182617,
+      "learning_rate": 6.449195906432749e-06,
+      "loss": 1.3081,
+      "step": 14830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.132498264312744,
+      "learning_rate": 6.448282163742691e-06,
+      "loss": 2.1071,
+      "step": 14831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5937352180480957,
+      "learning_rate": 6.447368421052632e-06,
+      "loss": 1.4732,
+      "step": 14832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4046075344085693,
+      "learning_rate": 6.4464546783625735e-06,
+      "loss": 1.3534,
+      "step": 14833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.672675848007202,
+      "learning_rate": 6.445540935672515e-06,
+      "loss": 1.425,
+      "step": 14834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7541065216064453,
+      "learning_rate": 6.444627192982457e-06,
+      "loss": 1.4227,
+      "step": 14835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25282883644104,
+      "learning_rate": 6.4437134502923984e-06,
+      "loss": 1.4051,
+      "step": 14836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.35932993888855,
+      "learning_rate": 6.44279970760234e-06,
+      "loss": 1.3876,
+      "step": 14837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6909892559051514,
+      "learning_rate": 6.441885964912281e-06,
+      "loss": 1.7104,
+      "step": 14838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.559037446975708,
+      "learning_rate": 6.4409722222222226e-06,
+      "loss": 1.9293,
+      "step": 14839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1677658557891846,
+      "learning_rate": 6.440058479532164e-06,
+      "loss": 0.9852,
+      "step": 14840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.225674629211426,
+      "learning_rate": 6.439144736842105e-06,
+      "loss": 1.3325,
+      "step": 14841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.843073844909668,
+      "learning_rate": 6.4382309941520475e-06,
+      "loss": 1.9218,
+      "step": 14842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4550325870513916,
+      "learning_rate": 6.437317251461989e-06,
+      "loss": 2.3216,
+      "step": 14843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.804795026779175,
+      "learning_rate": 6.43640350877193e-06,
+      "loss": 1.6273,
+      "step": 14844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.081867218017578,
+      "learning_rate": 6.435489766081872e-06,
+      "loss": 1.0985,
+      "step": 14845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1715316772460938,
+      "learning_rate": 6.434576023391813e-06,
+      "loss": 1.1657,
+      "step": 14846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484968662261963,
+      "learning_rate": 6.433662280701754e-06,
+      "loss": 1.4804,
+      "step": 14847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.996054172515869,
+      "learning_rate": 6.432748538011696e-06,
+      "loss": 1.4649,
+      "step": 14848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.965574264526367,
+      "learning_rate": 6.431834795321638e-06,
+      "loss": 1.8119,
+      "step": 14849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0814521312713623,
+      "learning_rate": 6.430921052631579e-06,
+      "loss": 1.6205,
+      "step": 14850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5639548301696777,
+      "learning_rate": 6.430007309941521e-06,
+      "loss": 1.1639,
+      "step": 14851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9559619426727295,
+      "learning_rate": 6.429093567251462e-06,
+      "loss": 1.2425,
+      "step": 14852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3014960289001465,
+      "learning_rate": 6.4281798245614035e-06,
+      "loss": 1.3774,
+      "step": 14853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1041643619537354,
+      "learning_rate": 6.427266081871345e-06,
+      "loss": 1.6428,
+      "step": 14854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7176566123962402,
+      "learning_rate": 6.426352339181286e-06,
+      "loss": 1.6324,
+      "step": 14855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8226020336151123,
+      "learning_rate": 6.425438596491229e-06,
+      "loss": 1.6688,
+      "step": 14856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9708346128463745,
+      "learning_rate": 6.424524853801171e-06,
+      "loss": 1.1388,
+      "step": 14857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.547945499420166,
+      "learning_rate": 6.423611111111112e-06,
+      "loss": 1.4976,
+      "step": 14858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3726234436035156,
+      "learning_rate": 6.4226973684210534e-06,
+      "loss": 1.2183,
+      "step": 14859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7439825534820557,
+      "learning_rate": 6.421783625730995e-06,
+      "loss": 1.7522,
+      "step": 14860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5343148708343506,
+      "learning_rate": 6.420869883040937e-06,
+      "loss": 1.5375,
+      "step": 14861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.76472806930542,
+      "learning_rate": 6.419956140350878e-06,
+      "loss": 1.485,
+      "step": 14862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5176475048065186,
+      "learning_rate": 6.41904239766082e-06,
+      "loss": 1.1222,
+      "step": 14863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.517595052719116,
+      "learning_rate": 6.418128654970761e-06,
+      "loss": 1.4186,
+      "step": 14864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3003761768341064,
+      "learning_rate": 6.4172149122807025e-06,
+      "loss": 1.6396,
+      "step": 14865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6101672649383545,
+      "learning_rate": 6.416301169590644e-06,
+      "loss": 1.4762,
+      "step": 14866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9152538776397705,
+      "learning_rate": 6.415387426900585e-06,
+      "loss": 1.4119,
+      "step": 14867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4598727226257324,
+      "learning_rate": 6.4144736842105275e-06,
+      "loss": 1.4242,
+      "step": 14868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5587828159332275,
+      "learning_rate": 6.413559941520469e-06,
+      "loss": 1.6007,
+      "step": 14869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4203290939331055,
+      "learning_rate": 6.41264619883041e-06,
+      "loss": 1.4623,
+      "step": 14870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.369349956512451,
+      "learning_rate": 6.411732456140352e-06,
+      "loss": 1.6407,
+      "step": 14871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0861904621124268,
+      "learning_rate": 6.410818713450293e-06,
+      "loss": 0.945,
+      "step": 14872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4047648906707764,
+      "learning_rate": 6.409904970760234e-06,
+      "loss": 1.387,
+      "step": 14873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1631920337677,
+      "learning_rate": 6.408991228070176e-06,
+      "loss": 1.3559,
+      "step": 14874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4646246433258057,
+      "learning_rate": 6.408077485380117e-06,
+      "loss": 1.3597,
+      "step": 14875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6836869716644287,
+      "learning_rate": 6.407163742690059e-06,
+      "loss": 1.5319,
+      "step": 14876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0267157554626465,
+      "learning_rate": 6.406250000000001e-06,
+      "loss": 1.5268,
+      "step": 14877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2819013595581055,
+      "learning_rate": 6.405336257309942e-06,
+      "loss": 1.5218,
+      "step": 14878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5610368251800537,
+      "learning_rate": 6.4044225146198835e-06,
+      "loss": 0.9459,
+      "step": 14879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6033577919006348,
+      "learning_rate": 6.403508771929825e-06,
+      "loss": 1.4651,
+      "step": 14880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.367262840270996,
+      "learning_rate": 6.402595029239766e-06,
+      "loss": 1.2211,
+      "step": 14881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8840134143829346,
+      "learning_rate": 6.401681286549708e-06,
+      "loss": 1.3216,
+      "step": 14882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7297708988189697,
+      "learning_rate": 6.40076754385965e-06,
+      "loss": 1.4167,
+      "step": 14883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0739431381225586,
+      "learning_rate": 6.399853801169591e-06,
+      "loss": 1.4892,
+      "step": 14884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.538888454437256,
+      "learning_rate": 6.3989400584795326e-06,
+      "loss": 1.1434,
+      "step": 14885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.978635311126709,
+      "learning_rate": 6.398026315789474e-06,
+      "loss": 1.5227,
+      "step": 14886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5810954570770264,
+      "learning_rate": 6.397112573099415e-06,
+      "loss": 1.5279,
+      "step": 14887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.61307430267334,
+      "learning_rate": 6.396198830409357e-06,
+      "loss": 1.4269,
+      "step": 14888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9425175189971924,
+      "learning_rate": 6.395285087719298e-06,
+      "loss": 1.6249,
+      "step": 14889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5285933017730713,
+      "learning_rate": 6.39437134502924e-06,
+      "loss": 1.632,
+      "step": 14890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0962910652160645,
+      "learning_rate": 6.393457602339182e-06,
+      "loss": 1.2908,
+      "step": 14891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5460171699523926,
+      "learning_rate": 6.392543859649123e-06,
+      "loss": 1.6943,
+      "step": 14892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.204991579055786,
+      "learning_rate": 6.391630116959064e-06,
+      "loss": 1.1567,
+      "step": 14893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7894437313079834,
+      "learning_rate": 6.390716374269006e-06,
+      "loss": 1.5684,
+      "step": 14894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.179232358932495,
+      "learning_rate": 6.389802631578947e-06,
+      "loss": 1.122,
+      "step": 14895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1238291263580322,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 1.2096,
+      "step": 14896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.25533127784729,
+      "learning_rate": 6.387975146198831e-06,
+      "loss": 1.1679,
+      "step": 14897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.658259630203247,
+      "learning_rate": 6.387061403508772e-06,
+      "loss": 1.6496,
+      "step": 14898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5939717292785645,
+      "learning_rate": 6.386147660818714e-06,
+      "loss": 1.4267,
+      "step": 14899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.564182758331299,
+      "learning_rate": 6.385233918128656e-06,
+      "loss": 1.3665,
+      "step": 14900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.278498411178589,
+      "learning_rate": 6.384320175438597e-06,
+      "loss": 1.3549,
+      "step": 14901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.625964879989624,
+      "learning_rate": 6.383406432748539e-06,
+      "loss": 1.7217,
+      "step": 14902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2768681049346924,
+      "learning_rate": 6.382492690058481e-06,
+      "loss": 1.4583,
+      "step": 14903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4726288318634033,
+      "learning_rate": 6.381578947368422e-06,
+      "loss": 1.4193,
+      "step": 14904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3959217071533203,
+      "learning_rate": 6.3806652046783634e-06,
+      "loss": 1.5367,
+      "step": 14905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5861318111419678,
+      "learning_rate": 6.379751461988305e-06,
+      "loss": 1.3946,
+      "step": 14906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.373009204864502,
+      "learning_rate": 6.378837719298246e-06,
+      "loss": 1.4928,
+      "step": 14907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6480910778045654,
+      "learning_rate": 6.3779239766081876e-06,
+      "loss": 1.4466,
+      "step": 14908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.500281572341919,
+      "learning_rate": 6.37701023391813e-06,
+      "loss": 1.272,
+      "step": 14909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2966651916503906,
+      "learning_rate": 6.376096491228071e-06,
+      "loss": 1.311,
+      "step": 14910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.861049175262451,
+      "learning_rate": 6.3751827485380125e-06,
+      "loss": 1.9468,
+      "step": 14911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5076303482055664,
+      "learning_rate": 6.374269005847954e-06,
+      "loss": 1.2317,
+      "step": 14912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0627009868621826,
+      "learning_rate": 6.373355263157895e-06,
+      "loss": 1.917,
+      "step": 14913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3397269248962402,
+      "learning_rate": 6.372441520467837e-06,
+      "loss": 1.4987,
+      "step": 14914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9980099201202393,
+      "learning_rate": 6.371527777777778e-06,
+      "loss": 1.6669,
+      "step": 14915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4534189701080322,
+      "learning_rate": 6.37061403508772e-06,
+      "loss": 1.3684,
+      "step": 14916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5050063133239746,
+      "learning_rate": 6.369700292397662e-06,
+      "loss": 1.2001,
+      "step": 14917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2052390575408936,
+      "learning_rate": 6.368786549707603e-06,
+      "loss": 1.1977,
+      "step": 14918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7047781944274902,
+      "learning_rate": 6.367872807017544e-06,
+      "loss": 1.6573,
+      "step": 14919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1374855041503906,
+      "learning_rate": 6.366959064327486e-06,
+      "loss": 1.3153,
+      "step": 14920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2360119819641113,
+      "learning_rate": 6.366045321637427e-06,
+      "loss": 1.8209,
+      "step": 14921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4394805431365967,
+      "learning_rate": 6.3651315789473685e-06,
+      "loss": 1.5766,
+      "step": 14922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6423499584198,
+      "learning_rate": 6.364217836257311e-06,
+      "loss": 1.2437,
+      "step": 14923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9732718467712402,
+      "learning_rate": 6.363304093567252e-06,
+      "loss": 1.5825,
+      "step": 14924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4273812770843506,
+      "learning_rate": 6.3623903508771935e-06,
+      "loss": 1.6341,
+      "step": 14925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.060114622116089,
+      "learning_rate": 6.361476608187135e-06,
+      "loss": 1.5731,
+      "step": 14926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5964066982269287,
+      "learning_rate": 6.360562865497076e-06,
+      "loss": 1.4908,
+      "step": 14927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5710339546203613,
+      "learning_rate": 6.359649122807018e-06,
+      "loss": 1.401,
+      "step": 14928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.610321044921875,
+      "learning_rate": 6.358735380116959e-06,
+      "loss": 1.3925,
+      "step": 14929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0905818939208984,
+      "learning_rate": 6.357821637426901e-06,
+      "loss": 1.0655,
+      "step": 14930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0424957275390625,
+      "learning_rate": 6.3569078947368426e-06,
+      "loss": 1.1868,
+      "step": 14931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2885618209838867,
+      "learning_rate": 6.355994152046784e-06,
+      "loss": 1.2828,
+      "step": 14932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6266250610351562,
+      "learning_rate": 6.355080409356725e-06,
+      "loss": 1.6448,
+      "step": 14933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9818005561828613,
+      "learning_rate": 6.354166666666667e-06,
+      "loss": 1.8634,
+      "step": 14934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2263264656066895,
+      "learning_rate": 6.353252923976608e-06,
+      "loss": 1.3529,
+      "step": 14935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6855545043945312,
+      "learning_rate": 6.3523391812865494e-06,
+      "loss": 1.5353,
+      "step": 14936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6071128845214844,
+      "learning_rate": 6.351425438596492e-06,
+      "loss": 1.6645,
+      "step": 14937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3486456871032715,
+      "learning_rate": 6.350511695906433e-06,
+      "loss": 1.2721,
+      "step": 14938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.532195568084717,
+      "learning_rate": 6.349597953216374e-06,
+      "loss": 1.4924,
+      "step": 14939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3482139110565186,
+      "learning_rate": 6.348684210526316e-06,
+      "loss": 1.4102,
+      "step": 14940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2258176803588867,
+      "learning_rate": 6.347770467836257e-06,
+      "loss": 1.4388,
+      "step": 14941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.274869918823242,
+      "learning_rate": 6.3468567251462e-06,
+      "loss": 2.1871,
+      "step": 14942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.175652265548706,
+      "learning_rate": 6.345942982456142e-06,
+      "loss": 1.724,
+      "step": 14943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5892510414123535,
+      "learning_rate": 6.345029239766083e-06,
+      "loss": 1.5141,
+      "step": 14944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.410367727279663,
+      "learning_rate": 6.344115497076024e-06,
+      "loss": 1.6578,
+      "step": 14945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4047861099243164,
+      "learning_rate": 6.343201754385966e-06,
+      "loss": 1.3537,
+      "step": 14946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.424140453338623,
+      "learning_rate": 6.342288011695907e-06,
+      "loss": 1.3787,
+      "step": 14947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7492804527282715,
+      "learning_rate": 6.3413742690058485e-06,
+      "loss": 1.8487,
+      "step": 14948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2578518390655518,
+      "learning_rate": 6.340460526315791e-06,
+      "loss": 1.3639,
+      "step": 14949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7701852321624756,
+      "learning_rate": 6.339546783625732e-06,
+      "loss": 1.3855,
+      "step": 14950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4360735416412354,
+      "learning_rate": 6.3386330409356734e-06,
+      "loss": 1.3537,
+      "step": 14951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.745656728744507,
+      "learning_rate": 6.337719298245615e-06,
+      "loss": 1.4484,
+      "step": 14952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.283991813659668,
+      "learning_rate": 6.336805555555556e-06,
+      "loss": 1.2792,
+      "step": 14953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6429545879364014,
+      "learning_rate": 6.3358918128654976e-06,
+      "loss": 1.562,
+      "step": 14954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7862040996551514,
+      "learning_rate": 6.334978070175439e-06,
+      "loss": 1.6224,
+      "step": 14955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.901090383529663,
+      "learning_rate": 6.334064327485381e-06,
+      "loss": 1.6752,
+      "step": 14956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.471461057662964,
+      "learning_rate": 6.3331505847953225e-06,
+      "loss": 1.3809,
+      "step": 14957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8878321647644043,
+      "learning_rate": 6.332236842105264e-06,
+      "loss": 1.2638,
+      "step": 14958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6761412620544434,
+      "learning_rate": 6.331323099415205e-06,
+      "loss": 1.5915,
+      "step": 14959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.331511974334717,
+      "learning_rate": 6.330409356725147e-06,
+      "loss": 1.3698,
+      "step": 14960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7855818271636963,
+      "learning_rate": 6.329495614035088e-06,
+      "loss": 1.5402,
+      "step": 14961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3137400150299072,
+      "learning_rate": 6.328581871345029e-06,
+      "loss": 1.5687,
+      "step": 14962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7028727531433105,
+      "learning_rate": 6.327668128654972e-06,
+      "loss": 1.2842,
+      "step": 14963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.281538248062134,
+      "learning_rate": 6.326754385964913e-06,
+      "loss": 1.3478,
+      "step": 14964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8106367588043213,
+      "learning_rate": 6.325840643274854e-06,
+      "loss": 1.431,
+      "step": 14965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.606497287750244,
+      "learning_rate": 6.324926900584796e-06,
+      "loss": 1.2912,
+      "step": 14966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7716481685638428,
+      "learning_rate": 6.324013157894737e-06,
+      "loss": 1.5598,
+      "step": 14967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6712541580200195,
+      "learning_rate": 6.3230994152046785e-06,
+      "loss": 1.5343,
+      "step": 14968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7510602474212646,
+      "learning_rate": 6.32218567251462e-06,
+      "loss": 1.4447,
+      "step": 14969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3176462650299072,
+      "learning_rate": 6.321271929824562e-06,
+      "loss": 1.4844,
+      "step": 14970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3688950538635254,
+      "learning_rate": 6.3203581871345035e-06,
+      "loss": 1.5104,
+      "step": 14971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.393275499343872,
+      "learning_rate": 6.319444444444445e-06,
+      "loss": 1.369,
+      "step": 14972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9287221431732178,
+      "learning_rate": 6.318530701754386e-06,
+      "loss": 1.3559,
+      "step": 14973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6570467948913574,
+      "learning_rate": 6.317616959064328e-06,
+      "loss": 1.4453,
+      "step": 14974
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1669254302978516,
+      "learning_rate": 6.316703216374269e-06,
+      "loss": 1.6759,
+      "step": 14975
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1389777660369873,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 1.5855,
+      "step": 14976
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5527570247650146,
+      "learning_rate": 6.3148757309941526e-06,
+      "loss": 1.4087,
+      "step": 14977
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5912163257598877,
+      "learning_rate": 6.313961988304094e-06,
+      "loss": 1.6525,
+      "step": 14978
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.715265989303589,
+      "learning_rate": 6.313048245614035e-06,
+      "loss": 1.3164,
+      "step": 14979
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3845131397247314,
+      "learning_rate": 6.312134502923977e-06,
+      "loss": 1.3145,
+      "step": 14980
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8824353218078613,
+      "learning_rate": 6.311220760233918e-06,
+      "loss": 1.396,
+      "step": 14981
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.009730100631714,
+      "learning_rate": 6.310307017543859e-06,
+      "loss": 1.0103,
+      "step": 14982
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.620161294937134,
+      "learning_rate": 6.309393274853801e-06,
+      "loss": 1.2478,
+      "step": 14983
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.606736660003662,
+      "learning_rate": 6.308479532163744e-06,
+      "loss": 1.5892,
+      "step": 14984
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.573601484298706,
+      "learning_rate": 6.307565789473685e-06,
+      "loss": 1.643,
+      "step": 14985
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.445582151412964,
+      "learning_rate": 6.306652046783627e-06,
+      "loss": 1.3893,
+      "step": 14986
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1808531284332275,
+      "learning_rate": 6.305738304093568e-06,
+      "loss": 1.2972,
+      "step": 14987
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8811182975769043,
+      "learning_rate": 6.304824561403509e-06,
+      "loss": 1.4877,
+      "step": 14988
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8916641473770142,
+      "learning_rate": 6.303910818713452e-06,
+      "loss": 1.1393,
+      "step": 14989
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.475933313369751,
+      "learning_rate": 6.302997076023393e-06,
+      "loss": 1.227,
+      "step": 14990
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.70815372467041,
+      "learning_rate": 6.302083333333334e-06,
+      "loss": 1.5591,
+      "step": 14991
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1693246364593506,
+      "learning_rate": 6.301169590643276e-06,
+      "loss": 1.2956,
+      "step": 14992
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7158865928649902,
+      "learning_rate": 6.300255847953217e-06,
+      "loss": 1.8241,
+      "step": 14993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.787076473236084,
+      "learning_rate": 6.2993421052631585e-06,
+      "loss": 1.2377,
+      "step": 14994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2419087886810303,
+      "learning_rate": 6.2984283625731e-06,
+      "loss": 1.5511,
+      "step": 14995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7712254524230957,
+      "learning_rate": 6.297514619883042e-06,
+      "loss": 1.4948,
+      "step": 14996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3110616207122803,
+      "learning_rate": 6.2966008771929834e-06,
+      "loss": 1.4105,
+      "step": 14997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3498106002807617,
+      "learning_rate": 6.295687134502925e-06,
+      "loss": 1.3779,
+      "step": 14998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.841359853744507,
+      "learning_rate": 6.294773391812866e-06,
+      "loss": 1.6795,
+      "step": 14999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6604621410369873,
+      "learning_rate": 6.2938596491228076e-06,
+      "loss": 1.656,
+      "step": 15000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3006250858306885,
+      "learning_rate": 6.292945906432749e-06,
+      "loss": 1.4632,
+      "step": 15001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.884897470474243,
+      "learning_rate": 6.29203216374269e-06,
+      "loss": 1.5168,
+      "step": 15002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.71151065826416,
+      "learning_rate": 6.2911184210526325e-06,
+      "loss": 1.6753,
+      "step": 15003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4468300342559814,
+      "learning_rate": 6.290204678362574e-06,
+      "loss": 1.1618,
+      "step": 15004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.889723300933838,
+      "learning_rate": 6.289290935672515e-06,
+      "loss": 1.4663,
+      "step": 15005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4291458129882812,
+      "learning_rate": 6.288377192982457e-06,
+      "loss": 1.3647,
+      "step": 15006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.748818874359131,
+      "learning_rate": 6.287463450292398e-06,
+      "loss": 1.5295,
+      "step": 15007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.617046594619751,
+      "learning_rate": 6.286549707602339e-06,
+      "loss": 1.6523,
+      "step": 15008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7383615970611572,
+      "learning_rate": 6.285635964912281e-06,
+      "loss": 1.5568,
+      "step": 15009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5518693923950195,
+      "learning_rate": 6.284722222222223e-06,
+      "loss": 1.2758,
+      "step": 15010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9149489402770996,
+      "learning_rate": 6.283808479532164e-06,
+      "loss": 1.1533,
+      "step": 15011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8796823024749756,
+      "learning_rate": 6.282894736842106e-06,
+      "loss": 1.5626,
+      "step": 15012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.865206003189087,
+      "learning_rate": 6.281980994152047e-06,
+      "loss": 1.5365,
+      "step": 15013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5507349967956543,
+      "learning_rate": 6.2810672514619885e-06,
+      "loss": 1.3946,
+      "step": 15014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.578519582748413,
+      "learning_rate": 6.28015350877193e-06,
+      "loss": 1.3918,
+      "step": 15015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.553736686706543,
+      "learning_rate": 6.279239766081871e-06,
+      "loss": 1.4654,
+      "step": 15016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.31925892829895,
+      "learning_rate": 6.2783260233918135e-06,
+      "loss": 1.1942,
+      "step": 15017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.21032452583313,
+      "learning_rate": 6.277412280701755e-06,
+      "loss": 2.1378,
+      "step": 15018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.127016305923462,
+      "learning_rate": 6.276498538011696e-06,
+      "loss": 1.2817,
+      "step": 15019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4021387100219727,
+      "learning_rate": 6.275584795321638e-06,
+      "loss": 1.34,
+      "step": 15020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1246981620788574,
+      "learning_rate": 6.274671052631579e-06,
+      "loss": 1.1203,
+      "step": 15021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.672778844833374,
+      "learning_rate": 6.27375730994152e-06,
+      "loss": 1.2761,
+      "step": 15022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2759125232696533,
+      "learning_rate": 6.272843567251462e-06,
+      "loss": 1.1433,
+      "step": 15023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5307328701019287,
+      "learning_rate": 6.271929824561404e-06,
+      "loss": 1.361,
+      "step": 15024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.206777572631836,
+      "learning_rate": 6.271016081871345e-06,
+      "loss": 1.386,
+      "step": 15025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5506787300109863,
+      "learning_rate": 6.270102339181287e-06,
+      "loss": 1.5087,
+      "step": 15026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.596224069595337,
+      "learning_rate": 6.269188596491229e-06,
+      "loss": 1.3671,
+      "step": 15027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.644301176071167,
+      "learning_rate": 6.26827485380117e-06,
+      "loss": 1.5602,
+      "step": 15028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.950380325317383,
+      "learning_rate": 6.2673611111111125e-06,
+      "loss": 1.2367,
+      "step": 15029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.29725980758667,
+      "learning_rate": 6.266447368421054e-06,
+      "loss": 1.219,
+      "step": 15030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6352953910827637,
+      "learning_rate": 6.265533625730995e-06,
+      "loss": 1.2993,
+      "step": 15031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.668670892715454,
+      "learning_rate": 6.264619883040937e-06,
+      "loss": 1.4169,
+      "step": 15032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4253246784210205,
+      "learning_rate": 6.263706140350878e-06,
+      "loss": 1.441,
+      "step": 15033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5359420776367188,
+      "learning_rate": 6.262792397660819e-06,
+      "loss": 1.1809,
+      "step": 15034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1810500621795654,
+      "learning_rate": 6.261878654970761e-06,
+      "loss": 1.0357,
+      "step": 15035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6372785568237305,
+      "learning_rate": 6.260964912280703e-06,
+      "loss": 1.4184,
+      "step": 15036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2923316955566406,
+      "learning_rate": 6.260051169590644e-06,
+      "loss": 1.3208,
+      "step": 15037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3760433197021484,
+      "learning_rate": 6.259137426900586e-06,
+      "loss": 1.2982,
+      "step": 15038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2674403190612793,
+      "learning_rate": 6.258223684210527e-06,
+      "loss": 1.339,
+      "step": 15039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5941083431243896,
+      "learning_rate": 6.2573099415204685e-06,
+      "loss": 1.3841,
+      "step": 15040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5231785774230957,
+      "learning_rate": 6.25639619883041e-06,
+      "loss": 1.3415,
+      "step": 15041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5336673259735107,
+      "learning_rate": 6.255482456140351e-06,
+      "loss": 1.487,
+      "step": 15042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.785709857940674,
+      "learning_rate": 6.2545687134502934e-06,
+      "loss": 1.6934,
+      "step": 15043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9084396362304688,
+      "learning_rate": 6.253654970760235e-06,
+      "loss": 1.7555,
+      "step": 15044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9623031616210938,
+      "learning_rate": 6.252741228070176e-06,
+      "loss": 1.3769,
+      "step": 15045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8915512561798096,
+      "learning_rate": 6.2518274853801176e-06,
+      "loss": 1.4445,
+      "step": 15046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4982855319976807,
+      "learning_rate": 6.250913742690059e-06,
+      "loss": 1.4165,
+      "step": 15047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3235015869140625,
+      "learning_rate": 6.25e-06,
+      "loss": 1.3908,
+      "step": 15048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7289531230926514,
+      "learning_rate": 6.249086257309942e-06,
+      "loss": 1.6112,
+      "step": 15049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.387528657913208,
+      "learning_rate": 6.248172514619884e-06,
+      "loss": 1.3661,
+      "step": 15050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3408899307250977,
+      "learning_rate": 6.247258771929825e-06,
+      "loss": 1.4022,
+      "step": 15051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.49120831489563,
+      "learning_rate": 6.246345029239767e-06,
+      "loss": 1.4604,
+      "step": 15052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.199145555496216,
+      "learning_rate": 6.245431286549708e-06,
+      "loss": 1.3621,
+      "step": 15053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.175347328186035,
+      "learning_rate": 6.244517543859649e-06,
+      "loss": 1.3941,
+      "step": 15054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.499281883239746,
+      "learning_rate": 6.243603801169591e-06,
+      "loss": 1.5442,
+      "step": 15055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9508376121520996,
+      "learning_rate": 6.242690058479532e-06,
+      "loss": 1.6573,
+      "step": 15056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.713338851928711,
+      "learning_rate": 6.241776315789474e-06,
+      "loss": 1.752,
+      "step": 15057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5369656085968018,
+      "learning_rate": 6.240862573099416e-06,
+      "loss": 1.3224,
+      "step": 15058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.468996524810791,
+      "learning_rate": 6.239948830409357e-06,
+      "loss": 1.5236,
+      "step": 15059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5421416759490967,
+      "learning_rate": 6.2390350877192985e-06,
+      "loss": 1.6664,
+      "step": 15060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.967615842819214,
+      "learning_rate": 6.23812134502924e-06,
+      "loss": 1.5938,
+      "step": 15061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1743381023406982,
+      "learning_rate": 6.237207602339181e-06,
+      "loss": 1.1763,
+      "step": 15062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.257478713989258,
+      "learning_rate": 6.236293859649123e-06,
+      "loss": 1.5595,
+      "step": 15063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5114591121673584,
+      "learning_rate": 6.235380116959065e-06,
+      "loss": 1.5222,
+      "step": 15064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7048802375793457,
+      "learning_rate": 6.234466374269006e-06,
+      "loss": 1.4521,
+      "step": 15065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.961033344268799,
+      "learning_rate": 6.233552631578948e-06,
+      "loss": 2.0219,
+      "step": 15066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0771050453186035,
+      "learning_rate": 6.232638888888889e-06,
+      "loss": 1.1968,
+      "step": 15067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.919140100479126,
+      "learning_rate": 6.23172514619883e-06,
+      "loss": 1.6837,
+      "step": 15068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.460364580154419,
+      "learning_rate": 6.230811403508772e-06,
+      "loss": 1.7049,
+      "step": 15069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7510321140289307,
+      "learning_rate": 6.229897660818715e-06,
+      "loss": 1.4852,
+      "step": 15070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.995993137359619,
+      "learning_rate": 6.228983918128656e-06,
+      "loss": 1.8173,
+      "step": 15071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6715354919433594,
+      "learning_rate": 6.2280701754385975e-06,
+      "loss": 1.4572,
+      "step": 15072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3007869720458984,
+      "learning_rate": 6.227156432748539e-06,
+      "loss": 1.6049,
+      "step": 15073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1980366706848145,
+      "learning_rate": 6.22624269005848e-06,
+      "loss": 2.2199,
+      "step": 15074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0463578701019287,
+      "learning_rate": 6.225328947368422e-06,
+      "loss": 1.2221,
+      "step": 15075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2387325763702393,
+      "learning_rate": 6.224415204678364e-06,
+      "loss": 1.6212,
+      "step": 15076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8691208362579346,
+      "learning_rate": 6.223501461988305e-06,
+      "loss": 1.4749,
+      "step": 15077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.48364520072937,
+      "learning_rate": 6.222587719298247e-06,
+      "loss": 1.4406,
+      "step": 15078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3797779083251953,
+      "learning_rate": 6.221673976608188e-06,
+      "loss": 1.6425,
+      "step": 15079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.319448232650757,
+      "learning_rate": 6.220760233918129e-06,
+      "loss": 1.153,
+      "step": 15080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.065817356109619,
+      "learning_rate": 6.219846491228071e-06,
+      "loss": 1.9503,
+      "step": 15081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2816030979156494,
+      "learning_rate": 6.218932748538012e-06,
+      "loss": 1.3327,
+      "step": 15082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6245341300964355,
+      "learning_rate": 6.218019005847954e-06,
+      "loss": 1.3184,
+      "step": 15083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.477583408355713,
+      "learning_rate": 6.217105263157896e-06,
+      "loss": 1.5457,
+      "step": 15084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.283992290496826,
+      "learning_rate": 6.216191520467837e-06,
+      "loss": 1.2872,
+      "step": 15085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6510722637176514,
+      "learning_rate": 6.2152777777777785e-06,
+      "loss": 1.7911,
+      "step": 15086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5084903240203857,
+      "learning_rate": 6.21436403508772e-06,
+      "loss": 1.6047,
+      "step": 15087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.314377546310425,
+      "learning_rate": 6.213450292397661e-06,
+      "loss": 1.3474,
+      "step": 15088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.411240816116333,
+      "learning_rate": 6.212536549707603e-06,
+      "loss": 1.1806,
+      "step": 15089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2943360805511475,
+      "learning_rate": 6.211622807017545e-06,
+      "loss": 1.2103,
+      "step": 15090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4642066955566406,
+      "learning_rate": 6.210709064327486e-06,
+      "loss": 1.5129,
+      "step": 15091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.404891014099121,
+      "learning_rate": 6.2097953216374276e-06,
+      "loss": 1.5123,
+      "step": 15092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.393505811691284,
+      "learning_rate": 6.208881578947369e-06,
+      "loss": 1.4488,
+      "step": 15093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2442076206207275,
+      "learning_rate": 6.20796783625731e-06,
+      "loss": 1.4813,
+      "step": 15094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1724255084991455,
+      "learning_rate": 6.207054093567252e-06,
+      "loss": 1.5294,
+      "step": 15095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.647413730621338,
+      "learning_rate": 6.206140350877193e-06,
+      "loss": 1.7022,
+      "step": 15096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4767513275146484,
+      "learning_rate": 6.205226608187135e-06,
+      "loss": 1.6927,
+      "step": 15097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8772847652435303,
+      "learning_rate": 6.204312865497077e-06,
+      "loss": 1.5037,
+      "step": 15098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.711916446685791,
+      "learning_rate": 6.203399122807018e-06,
+      "loss": 1.5188,
+      "step": 15099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7890660762786865,
+      "learning_rate": 6.202485380116959e-06,
+      "loss": 1.4234,
+      "step": 15100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.341179847717285,
+      "learning_rate": 6.201571637426901e-06,
+      "loss": 1.3961,
+      "step": 15101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.29996657371521,
+      "learning_rate": 6.200657894736842e-06,
+      "loss": 1.3737,
+      "step": 15102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.053246259689331,
+      "learning_rate": 6.1997441520467835e-06,
+      "loss": 1.2248,
+      "step": 15103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3412280082702637,
+      "learning_rate": 6.198830409356725e-06,
+      "loss": 1.355,
+      "step": 15104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3006367683410645,
+      "learning_rate": 6.197916666666667e-06,
+      "loss": 1.2794,
+      "step": 15105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6877408027648926,
+      "learning_rate": 6.1970029239766085e-06,
+      "loss": 1.5945,
+      "step": 15106
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2220938205718994,
+      "learning_rate": 6.19608918128655e-06,
+      "loss": 1.4108,
+      "step": 15107
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3621883392333984,
+      "learning_rate": 6.195175438596491e-06,
+      "loss": 1.5858,
+      "step": 15108
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.548739194869995,
+      "learning_rate": 6.194261695906433e-06,
+      "loss": 1.2294,
+      "step": 15109
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6473538875579834,
+      "learning_rate": 6.193347953216374e-06,
+      "loss": 1.692,
+      "step": 15110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.233182668685913,
+      "learning_rate": 6.192434210526315e-06,
+      "loss": 1.5701,
+      "step": 15111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6396737098693848,
+      "learning_rate": 6.191520467836258e-06,
+      "loss": 1.4693,
+      "step": 15112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7876811027526855,
+      "learning_rate": 6.1906067251462e-06,
+      "loss": 1.8569,
+      "step": 15113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7263500690460205,
+      "learning_rate": 6.189692982456141e-06,
+      "loss": 1.6757,
+      "step": 15114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.386984348297119,
+      "learning_rate": 6.1887792397660826e-06,
+      "loss": 1.5079,
+      "step": 15115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6780810356140137,
+      "learning_rate": 6.187865497076025e-06,
+      "loss": 1.6196,
+      "step": 15116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9986257553100586,
+      "learning_rate": 6.186951754385966e-06,
+      "loss": 1.4096,
+      "step": 15117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.763636589050293,
+      "learning_rate": 6.1860380116959075e-06,
+      "loss": 1.6111,
+      "step": 15118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6206963062286377,
+      "learning_rate": 6.185124269005849e-06,
+      "loss": 1.3936,
+      "step": 15119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9420744180679321,
+      "learning_rate": 6.18421052631579e-06,
+      "loss": 1.0794,
+      "step": 15120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4075839519500732,
+      "learning_rate": 6.183296783625732e-06,
+      "loss": 1.2707,
+      "step": 15121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6525607109069824,
+      "learning_rate": 6.182383040935673e-06,
+      "loss": 1.5465,
+      "step": 15122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6123745441436768,
+      "learning_rate": 6.181469298245614e-06,
+      "loss": 1.4081,
+      "step": 15123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9140195846557617,
+      "learning_rate": 6.180555555555557e-06,
+      "loss": 1.806,
+      "step": 15124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.106250286102295,
+      "learning_rate": 6.179641812865498e-06,
+      "loss": 1.4843,
+      "step": 15125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.490673780441284,
+      "learning_rate": 6.178728070175439e-06,
+      "loss": 1.55,
+      "step": 15126
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4834916591644287,
+      "learning_rate": 6.177814327485381e-06,
+      "loss": 1.5461,
+      "step": 15127
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6745901107788086,
+      "learning_rate": 6.176900584795322e-06,
+      "loss": 1.593,
+      "step": 15128
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.653780937194824,
+      "learning_rate": 6.1759868421052635e-06,
+      "loss": 1.6868,
+      "step": 15129
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.366220474243164,
+      "learning_rate": 6.175073099415205e-06,
+      "loss": 1.1909,
+      "step": 15130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4841175079345703,
+      "learning_rate": 6.174159356725147e-06,
+      "loss": 1.5387,
+      "step": 15131
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3102433681488037,
+      "learning_rate": 6.1732456140350885e-06,
+      "loss": 1.3695,
+      "step": 15132
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.376904249191284,
+      "learning_rate": 6.17233187134503e-06,
+      "loss": 1.3609,
+      "step": 15133
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.128293991088867,
+      "learning_rate": 6.171418128654971e-06,
+      "loss": 1.2265,
+      "step": 15134
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3370919227600098,
+      "learning_rate": 6.170504385964913e-06,
+      "loss": 1.4815,
+      "step": 15135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7716503143310547,
+      "learning_rate": 6.169590643274854e-06,
+      "loss": 1.6111,
+      "step": 15136
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.47249436378479,
+      "learning_rate": 6.168676900584795e-06,
+      "loss": 1.4511,
+      "step": 15137
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.015920400619507,
+      "learning_rate": 6.1677631578947376e-06,
+      "loss": 1.7593,
+      "step": 15138
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.864020586013794,
+      "learning_rate": 6.166849415204679e-06,
+      "loss": 1.3365,
+      "step": 15139
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.255232334136963,
+      "learning_rate": 6.16593567251462e-06,
+      "loss": 1.4969,
+      "step": 15140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6763787269592285,
+      "learning_rate": 6.165021929824562e-06,
+      "loss": 1.4157,
+      "step": 15141
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9546167850494385,
+      "learning_rate": 6.164108187134503e-06,
+      "loss": 1.578,
+      "step": 15142
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3640894889831543,
+      "learning_rate": 6.163194444444444e-06,
+      "loss": 1.3886,
+      "step": 15143
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5508534908294678,
+      "learning_rate": 6.162280701754386e-06,
+      "loss": 1.4191,
+      "step": 15144
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5589988231658936,
+      "learning_rate": 6.161366959064328e-06,
+      "loss": 0.9758,
+      "step": 15145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.450432062149048,
+      "learning_rate": 6.160453216374269e-06,
+      "loss": 1.3746,
+      "step": 15146
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3254806995391846,
+      "learning_rate": 6.159539473684211e-06,
+      "loss": 1.259,
+      "step": 15147
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.476775646209717,
+      "learning_rate": 6.158625730994152e-06,
+      "loss": 1.2562,
+      "step": 15148
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.16395902633667,
+      "learning_rate": 6.1577119883040935e-06,
+      "loss": 1.6295,
+      "step": 15149
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0833921432495117,
+      "learning_rate": 6.156798245614035e-06,
+      "loss": 1.64,
+      "step": 15150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.605802297592163,
+      "learning_rate": 6.155884502923976e-06,
+      "loss": 1.3848,
+      "step": 15151
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0129690170288086,
+      "learning_rate": 6.1549707602339185e-06,
+      "loss": 1.7484,
+      "step": 15152
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3746559619903564,
+      "learning_rate": 6.15405701754386e-06,
+      "loss": 1.4951,
+      "step": 15153
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4710137844085693,
+      "learning_rate": 6.153143274853801e-06,
+      "loss": 1.2545,
+      "step": 15154
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6822569370269775,
+      "learning_rate": 6.1522295321637435e-06,
+      "loss": 1.5993,
+      "step": 15155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.118496894836426,
+      "learning_rate": 6.151315789473685e-06,
+      "loss": 1.4042,
+      "step": 15156
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9949228763580322,
+      "learning_rate": 6.150402046783627e-06,
+      "loss": 1.6951,
+      "step": 15157
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1929876804351807,
+      "learning_rate": 6.1494883040935684e-06,
+      "loss": 1.4015,
+      "step": 15158
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6348001956939697,
+      "learning_rate": 6.14857456140351e-06,
+      "loss": 1.5247,
+      "step": 15159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0159342288970947,
+      "learning_rate": 6.147660818713451e-06,
+      "loss": 1.839,
+      "step": 15160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3454701900482178,
+      "learning_rate": 6.1467470760233925e-06,
+      "loss": 1.192,
+      "step": 15161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2192065715789795,
+      "learning_rate": 6.145833333333334e-06,
+      "loss": 1.96,
+      "step": 15162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3540966510772705,
+      "learning_rate": 6.144919590643275e-06,
+      "loss": 1.6796,
+      "step": 15163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0306365489959717,
+      "learning_rate": 6.1440058479532175e-06,
+      "loss": 1.656,
+      "step": 15164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1656124591827393,
+      "learning_rate": 6.143092105263159e-06,
+      "loss": 1.3589,
+      "step": 15165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7342000007629395,
+      "learning_rate": 6.1421783625731e-06,
+      "loss": 1.7199,
+      "step": 15166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4115471839904785,
+      "learning_rate": 6.141264619883042e-06,
+      "loss": 1.5046,
+      "step": 15167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1884453296661377,
+      "learning_rate": 6.140350877192983e-06,
+      "loss": 1.4429,
+      "step": 15168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.422851085662842,
+      "learning_rate": 6.139437134502924e-06,
+      "loss": 1.2832,
+      "step": 15169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5013580322265625,
+      "learning_rate": 6.138523391812866e-06,
+      "loss": 1.5724,
+      "step": 15170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.080911159515381,
+      "learning_rate": 6.137609649122808e-06,
+      "loss": 1.9713,
+      "step": 15171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8419289588928223,
+      "learning_rate": 6.136695906432749e-06,
+      "loss": 1.492,
+      "step": 15172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8775527477264404,
+      "learning_rate": 6.135782163742691e-06,
+      "loss": 1.7885,
+      "step": 15173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.597963571548462,
+      "learning_rate": 6.134868421052632e-06,
+      "loss": 1.4096,
+      "step": 15174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6162993907928467,
+      "learning_rate": 6.1339546783625735e-06,
+      "loss": 1.6948,
+      "step": 15175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.655015707015991,
+      "learning_rate": 6.133040935672515e-06,
+      "loss": 1.565,
+      "step": 15176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.281890392303467,
+      "learning_rate": 6.132127192982456e-06,
+      "loss": 1.7206,
+      "step": 15177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1447553634643555,
+      "learning_rate": 6.1312134502923985e-06,
+      "loss": 1.3659,
+      "step": 15178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2043440341949463,
+      "learning_rate": 6.13029970760234e-06,
+      "loss": 1.6288,
+      "step": 15179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.276285171508789,
+      "learning_rate": 6.129385964912281e-06,
+      "loss": 1.4475,
+      "step": 15180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6141889095306396,
+      "learning_rate": 6.128472222222223e-06,
+      "loss": 1.6309,
+      "step": 15181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.156766653060913,
+      "learning_rate": 6.127558479532164e-06,
+      "loss": 1.1218,
+      "step": 15182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4678759574890137,
+      "learning_rate": 6.126644736842105e-06,
+      "loss": 1.4873,
+      "step": 15183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.843590021133423,
+      "learning_rate": 6.125730994152047e-06,
+      "loss": 1.4169,
+      "step": 15184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2072768211364746,
+      "learning_rate": 6.124817251461989e-06,
+      "loss": 1.4268,
+      "step": 15185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1236660480499268,
+      "learning_rate": 6.12390350877193e-06,
+      "loss": 1.0613,
+      "step": 15186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.415367603302002,
+      "learning_rate": 6.122989766081872e-06,
+      "loss": 1.6288,
+      "step": 15187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6072864532470703,
+      "learning_rate": 6.122076023391813e-06,
+      "loss": 1.3568,
+      "step": 15188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.273054838180542,
+      "learning_rate": 6.121162280701754e-06,
+      "loss": 1.4426,
+      "step": 15189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.82413387298584,
+      "learning_rate": 6.120248538011696e-06,
+      "loss": 1.6207,
+      "step": 15190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.134679079055786,
+      "learning_rate": 6.119334795321637e-06,
+      "loss": 1.1167,
+      "step": 15191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8809967041015625,
+      "learning_rate": 6.118421052631579e-06,
+      "loss": 1.4035,
+      "step": 15192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3750884532928467,
+      "learning_rate": 6.117507309941521e-06,
+      "loss": 1.5601,
+      "step": 15193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6388349533081055,
+      "learning_rate": 6.116593567251462e-06,
+      "loss": 1.4744,
+      "step": 15194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5908291339874268,
+      "learning_rate": 6.1156798245614035e-06,
+      "loss": 1.5897,
+      "step": 15195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.905355930328369,
+      "learning_rate": 6.114766081871345e-06,
+      "loss": 1.6552,
+      "step": 15196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.749910354614258,
+      "learning_rate": 6.113852339181286e-06,
+      "loss": 1.5041,
+      "step": 15197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.027973175048828,
+      "learning_rate": 6.112938596491229e-06,
+      "loss": 1.8167,
+      "step": 15198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8659095764160156,
+      "learning_rate": 6.112024853801171e-06,
+      "loss": 1.397,
+      "step": 15199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.302314043045044,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 1.6448,
+      "step": 15200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.066948890686035,
+      "learning_rate": 6.1101973684210535e-06,
+      "loss": 1.2871,
+      "step": 15201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5553462505340576,
+      "learning_rate": 6.109283625730995e-06,
+      "loss": 1.468,
+      "step": 15202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0870165824890137,
+      "learning_rate": 6.108369883040936e-06,
+      "loss": 1.3321,
+      "step": 15203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.424065351486206,
+      "learning_rate": 6.1074561403508784e-06,
+      "loss": 1.3137,
+      "step": 15204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.111654758453369,
+      "learning_rate": 6.10654239766082e-06,
+      "loss": 1.4905,
+      "step": 15205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2785189151763916,
+      "learning_rate": 6.105628654970761e-06,
+      "loss": 1.3076,
+      "step": 15206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1629133224487305,
+      "learning_rate": 6.1047149122807025e-06,
+      "loss": 1.293,
+      "step": 15207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3353772163391113,
+      "learning_rate": 6.103801169590644e-06,
+      "loss": 1.2558,
+      "step": 15208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6892096996307373,
+      "learning_rate": 6.102887426900585e-06,
+      "loss": 1.2652,
+      "step": 15209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7493526935577393,
+      "learning_rate": 6.101973684210527e-06,
+      "loss": 1.8076,
+      "step": 15210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3329761028289795,
+      "learning_rate": 6.101059941520469e-06,
+      "loss": 1.3127,
+      "step": 15211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3727223873138428,
+      "learning_rate": 6.10014619883041e-06,
+      "loss": 1.3708,
+      "step": 15212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.471684694290161,
+      "learning_rate": 6.099232456140352e-06,
+      "loss": 1.7123,
+      "step": 15213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.355534076690674,
+      "learning_rate": 6.098318713450293e-06,
+      "loss": 1.4207,
+      "step": 15214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.80354905128479,
+      "learning_rate": 6.097404970760234e-06,
+      "loss": 1.4855,
+      "step": 15215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6865766048431396,
+      "learning_rate": 6.096491228070176e-06,
+      "loss": 1.4637,
+      "step": 15216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5475809574127197,
+      "learning_rate": 6.095577485380117e-06,
+      "loss": 1.4259,
+      "step": 15217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3699421882629395,
+      "learning_rate": 6.094663742690059e-06,
+      "loss": 1.5635,
+      "step": 15218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8752825260162354,
+      "learning_rate": 6.093750000000001e-06,
+      "loss": 1.4437,
+      "step": 15219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.819131851196289,
+      "learning_rate": 6.092836257309942e-06,
+      "loss": 1.8152,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.237536668777466,
+      "learning_rate": 6.0919225146198835e-06,
+      "loss": 1.2658,
+      "step": 15221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4059829711914062,
+      "learning_rate": 6.091008771929825e-06,
+      "loss": 1.5347,
+      "step": 15222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.913658618927002,
+      "learning_rate": 6.090095029239766e-06,
+      "loss": 1.0754,
+      "step": 15223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4123096466064453,
+      "learning_rate": 6.089181286549708e-06,
+      "loss": 1.3663,
+      "step": 15224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4970779418945312,
+      "learning_rate": 6.08826754385965e-06,
+      "loss": 1.1524,
+      "step": 15225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.302121162414551,
+      "learning_rate": 6.087353801169591e-06,
+      "loss": 1.4442,
+      "step": 15226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.812283992767334,
+      "learning_rate": 6.086440058479533e-06,
+      "loss": 1.7729,
+      "step": 15227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9841995239257812,
+      "learning_rate": 6.085526315789474e-06,
+      "loss": 1.6123,
+      "step": 15228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5526790618896484,
+      "learning_rate": 6.084612573099415e-06,
+      "loss": 1.4261,
+      "step": 15229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5854554176330566,
+      "learning_rate": 6.083698830409357e-06,
+      "loss": 1.4288,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5829362869262695,
+      "learning_rate": 6.082785087719298e-06,
+      "loss": 1.6949,
+      "step": 15231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.822788715362549,
+      "learning_rate": 6.08187134502924e-06,
+      "loss": 1.521,
+      "step": 15232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3379666805267334,
+      "learning_rate": 6.080957602339182e-06,
+      "loss": 1.4434,
+      "step": 15233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5536365509033203,
+      "learning_rate": 6.080043859649123e-06,
+      "loss": 1.4622,
+      "step": 15234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1000638008117676,
+      "learning_rate": 6.079130116959064e-06,
+      "loss": 1.0129,
+      "step": 15235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8406295776367188,
+      "learning_rate": 6.078216374269006e-06,
+      "loss": 1.679,
+      "step": 15236
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.112130880355835,
+      "learning_rate": 6.077302631578947e-06,
+      "loss": 1.25,
+      "step": 15237
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.145860195159912,
+      "learning_rate": 6.0763888888888885e-06,
+      "loss": 1.1819,
+      "step": 15238
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.656956195831299,
+      "learning_rate": 6.075475146198831e-06,
+      "loss": 1.5482,
+      "step": 15239
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2295048236846924,
+      "learning_rate": 6.074561403508772e-06,
+      "loss": 2.1075,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.884855270385742,
+      "learning_rate": 6.073647660818714e-06,
+      "loss": 1.7785,
+      "step": 15241
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.086620330810547,
+      "learning_rate": 6.072733918128656e-06,
+      "loss": 1.2343,
+      "step": 15242
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3643879890441895,
+      "learning_rate": 6.071820175438597e-06,
+      "loss": 1.2158,
+      "step": 15243
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.323418140411377,
+      "learning_rate": 6.070906432748539e-06,
+      "loss": 1.904,
+      "step": 15244
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1122450828552246,
+      "learning_rate": 6.069992690058481e-06,
+      "loss": 1.1363,
+      "step": 15245
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.013434410095215,
+      "learning_rate": 6.069078947368422e-06,
+      "loss": 1.2213,
+      "step": 15246
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6974198818206787,
+      "learning_rate": 6.0681652046783635e-06,
+      "loss": 1.3564,
+      "step": 15247
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6285910606384277,
+      "learning_rate": 6.067251461988305e-06,
+      "loss": 1.7776,
+      "step": 15248
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6711337566375732,
+      "learning_rate": 6.066337719298246e-06,
+      "loss": 1.6876,
+      "step": 15249
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.597349166870117,
+      "learning_rate": 6.065423976608188e-06,
+      "loss": 1.7349,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8333141803741455,
+      "learning_rate": 6.06451023391813e-06,
+      "loss": 1.8601,
+      "step": 15251
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4007415771484375,
+      "learning_rate": 6.063596491228071e-06,
+      "loss": 1.1734,
+      "step": 15252
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.857372999191284,
+      "learning_rate": 6.0626827485380125e-06,
+      "loss": 1.7476,
+      "step": 15253
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.61226224899292,
+      "learning_rate": 6.061769005847954e-06,
+      "loss": 1.7508,
+      "step": 15254
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1691181659698486,
+      "learning_rate": 6.060855263157895e-06,
+      "loss": 1.5229,
+      "step": 15255
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7727253437042236,
+      "learning_rate": 6.059941520467837e-06,
+      "loss": 1.9311,
+      "step": 15256
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0343093872070312,
+      "learning_rate": 6.059027777777778e-06,
+      "loss": 1.3063,
+      "step": 15257
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3993217945098877,
+      "learning_rate": 6.05811403508772e-06,
+      "loss": 1.4608,
+      "step": 15258
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2713112831115723,
+      "learning_rate": 6.057200292397662e-06,
+      "loss": 1.2389,
+      "step": 15259
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.904120683670044,
+      "learning_rate": 6.056286549707603e-06,
+      "loss": 1.713,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.876492738723755,
+      "learning_rate": 6.055372807017544e-06,
+      "loss": 1.5251,
+      "step": 15261
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0387210845947266,
+      "learning_rate": 6.054459064327486e-06,
+      "loss": 1.1774,
+      "step": 15262
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4990336894989014,
+      "learning_rate": 6.053545321637427e-06,
+      "loss": 1.3239,
+      "step": 15263
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.091902732849121,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 1.3399,
+      "step": 15264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5400192737579346,
+      "learning_rate": 6.051717836257311e-06,
+      "loss": 1.5087,
+      "step": 15265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7292122840881348,
+      "learning_rate": 6.050804093567252e-06,
+      "loss": 1.529,
+      "step": 15266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1822290420532227,
+      "learning_rate": 6.0498903508771935e-06,
+      "loss": 1.2723,
+      "step": 15267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4656949043273926,
+      "learning_rate": 6.048976608187135e-06,
+      "loss": 1.3124,
+      "step": 15268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4721479415893555,
+      "learning_rate": 6.048062865497076e-06,
+      "loss": 1.4144,
+      "step": 15269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5254464149475098,
+      "learning_rate": 6.047149122807018e-06,
+      "loss": 1.2406,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6769073009490967,
+      "learning_rate": 6.046235380116959e-06,
+      "loss": 1.4315,
+      "step": 15271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0456039905548096,
+      "learning_rate": 6.045321637426901e-06,
+      "loss": 1.61,
+      "step": 15272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3248298168182373,
+      "learning_rate": 6.044407894736843e-06,
+      "loss": 1.4799,
+      "step": 15273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3481056690216064,
+      "learning_rate": 6.043494152046784e-06,
+      "loss": 1.3212,
+      "step": 15274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.636896848678589,
+      "learning_rate": 6.042580409356725e-06,
+      "loss": 1.4084,
+      "step": 15275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.078174352645874,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 1.6136,
+      "step": 15276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.822471857070923,
+      "learning_rate": 6.040752923976608e-06,
+      "loss": 1.6858,
+      "step": 15277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4726173877716064,
+      "learning_rate": 6.0398391812865495e-06,
+      "loss": 1.2854,
+      "step": 15278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.45353364944458,
+      "learning_rate": 6.038925438596492e-06,
+      "loss": 1.5614,
+      "step": 15279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.43247389793396,
+      "learning_rate": 6.038011695906433e-06,
+      "loss": 1.2435,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3969664573669434,
+      "learning_rate": 6.037097953216374e-06,
+      "loss": 1.135,
+      "step": 15281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.712205171585083,
+      "learning_rate": 6.036184210526316e-06,
+      "loss": 1.7869,
+      "step": 15282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7664954662323,
+      "learning_rate": 6.035270467836257e-06,
+      "loss": 1.157,
+      "step": 15283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2407588958740234,
+      "learning_rate": 6.0343567251462e-06,
+      "loss": 1.2174,
+      "step": 15284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1329355239868164,
+      "learning_rate": 6.033442982456142e-06,
+      "loss": 1.5962,
+      "step": 15285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7653064727783203,
+      "learning_rate": 6.032529239766083e-06,
+      "loss": 1.7304,
+      "step": 15286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2327945232391357,
+      "learning_rate": 6.031615497076024e-06,
+      "loss": 1.2961,
+      "step": 15287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2569668292999268,
+      "learning_rate": 6.030701754385966e-06,
+      "loss": 1.4699,
+      "step": 15288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6664512157440186,
+      "learning_rate": 6.029788011695907e-06,
+      "loss": 1.6169,
+      "step": 15289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.770864248275757,
+      "learning_rate": 6.0288742690058485e-06,
+      "loss": 1.5695,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3997597694396973,
+      "learning_rate": 6.027960526315791e-06,
+      "loss": 1.5663,
+      "step": 15291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4061572551727295,
+      "learning_rate": 6.027046783625732e-06,
+      "loss": 1.2611,
+      "step": 15292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0525176525115967,
+      "learning_rate": 6.0261330409356735e-06,
+      "loss": 1.5432,
+      "step": 15293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3702714443206787,
+      "learning_rate": 6.025219298245615e-06,
+      "loss": 1.2815,
+      "step": 15294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.99851655960083,
+      "learning_rate": 6.024305555555556e-06,
+      "loss": 1.1947,
+      "step": 15295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6843109130859375,
+      "learning_rate": 6.023391812865498e-06,
+      "loss": 1.274,
+      "step": 15296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.178236246109009,
+      "learning_rate": 6.022478070175439e-06,
+      "loss": 1.1328,
+      "step": 15297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.82767391204834,
+      "learning_rate": 6.021564327485381e-06,
+      "loss": 1.5325,
+      "step": 15298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.179374933242798,
+      "learning_rate": 6.0206505847953225e-06,
+      "loss": 1.2033,
+      "step": 15299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1657168865203857,
+      "learning_rate": 6.019736842105264e-06,
+      "loss": 1.2749,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2100348472595215,
+      "learning_rate": 6.018823099415205e-06,
+      "loss": 1.2046,
+      "step": 15301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7646026611328125,
+      "learning_rate": 6.017909356725147e-06,
+      "loss": 1.358,
+      "step": 15302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0957539081573486,
+      "learning_rate": 6.016995614035088e-06,
+      "loss": 1.5865,
+      "step": 15303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1667938232421875,
+      "learning_rate": 6.016081871345029e-06,
+      "loss": 1.9259,
+      "step": 15304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9971792697906494,
+      "learning_rate": 6.015168128654972e-06,
+      "loss": 1.6597,
+      "step": 15305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.218316078186035,
+      "learning_rate": 6.014254385964913e-06,
+      "loss": 1.2759,
+      "step": 15306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1226115226745605,
+      "learning_rate": 6.013340643274854e-06,
+      "loss": 1.1967,
+      "step": 15307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.693117141723633,
+      "learning_rate": 6.012426900584796e-06,
+      "loss": 1.4713,
+      "step": 15308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9728811979293823,
+      "learning_rate": 6.011513157894737e-06,
+      "loss": 1.0214,
+      "step": 15309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.345655918121338,
+      "learning_rate": 6.0105994152046785e-06,
+      "loss": 1.4125,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.775994300842285,
+      "learning_rate": 6.00968567251462e-06,
+      "loss": 1.4096,
+      "step": 15311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.514378547668457,
+      "learning_rate": 6.008771929824562e-06,
+      "loss": 1.4712,
+      "step": 15312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4169259071350098,
+      "learning_rate": 6.0078581871345035e-06,
+      "loss": 1.4231,
+      "step": 15313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.321859836578369,
+      "learning_rate": 6.006944444444445e-06,
+      "loss": 1.2959,
+      "step": 15314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.784708261489868,
+      "learning_rate": 6.006030701754386e-06,
+      "loss": 1.6882,
+      "step": 15315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7425482273101807,
+      "learning_rate": 6.005116959064328e-06,
+      "loss": 1.4976,
+      "step": 15316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4545791149139404,
+      "learning_rate": 6.004203216374269e-06,
+      "loss": 1.5339,
+      "step": 15317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.528146266937256,
+      "learning_rate": 6.00328947368421e-06,
+      "loss": 1.3388,
+      "step": 15318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.856001853942871,
+      "learning_rate": 6.002375730994153e-06,
+      "loss": 1.7098,
+      "step": 15319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.992326498031616,
+      "learning_rate": 6.001461988304094e-06,
+      "loss": 1.6956,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.749175786972046,
+      "learning_rate": 6.000548245614035e-06,
+      "loss": 1.6714,
+      "step": 15321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7177047729492188,
+      "learning_rate": 5.999634502923977e-06,
+      "loss": 1.5815,
+      "step": 15322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.850982666015625,
+      "learning_rate": 5.998720760233918e-06,
+      "loss": 1.8596,
+      "step": 15323
+    },
+    {
+      "epoch": 0.7,
+      "eval_train_bleu": 0.23249995714809302,
+      "eval_train_loss": 2.1017439365386963,
+      "eval_train_runtime": 23.8384,
+      "eval_train_samples_per_second": 41.991,
+      "eval_train_steps_per_second": 1.342,
+      "step": 15323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.763394355773926,
+      "learning_rate": 5.9978070175438595e-06,
+      "loss": 1.7564,
+      "step": 15324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4294869899749756,
+      "learning_rate": 5.996893274853801e-06,
+      "loss": 1.3746,
+      "step": 15325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8170113563537598,
+      "learning_rate": 5.995979532163744e-06,
+      "loss": 1.3248,
+      "step": 15326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.35785174369812,
+      "learning_rate": 5.995065789473685e-06,
+      "loss": 1.218,
+      "step": 15327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8131093978881836,
+      "learning_rate": 5.994152046783627e-06,
+      "loss": 1.6626,
+      "step": 15328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6731674671173096,
+      "learning_rate": 5.993238304093568e-06,
+      "loss": 2.116,
+      "step": 15329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4972214698791504,
+      "learning_rate": 5.992324561403509e-06,
+      "loss": 1.4004,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9457385540008545,
+      "learning_rate": 5.991410818713452e-06,
+      "loss": 1.4063,
+      "step": 15331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2996225357055664,
+      "learning_rate": 5.990497076023393e-06,
+      "loss": 1.3241,
+      "step": 15332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4126124382019043,
+      "learning_rate": 5.989583333333334e-06,
+      "loss": 1.3828,
+      "step": 15333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.326861619949341,
+      "learning_rate": 5.988669590643276e-06,
+      "loss": 1.2321,
+      "step": 15334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0873074531555176,
+      "learning_rate": 5.987755847953217e-06,
+      "loss": 1.9253,
+      "step": 15335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.44486927986145,
+      "learning_rate": 5.9868421052631585e-06,
+      "loss": 1.3745,
+      "step": 15336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9469285011291504,
+      "learning_rate": 5.9859283625731e-06,
+      "loss": 1.6944,
+      "step": 15337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.829610586166382,
+      "learning_rate": 5.985014619883042e-06,
+      "loss": 1.562,
+      "step": 15338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.549475908279419,
+      "learning_rate": 5.9841008771929835e-06,
+      "loss": 1.5094,
+      "step": 15339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8296315670013428,
+      "learning_rate": 5.983187134502925e-06,
+      "loss": 1.7133,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2186813354492188,
+      "learning_rate": 5.982273391812866e-06,
+      "loss": 1.2636,
+      "step": 15341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4478113651275635,
+      "learning_rate": 5.981359649122808e-06,
+      "loss": 1.3575,
+      "step": 15342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1546201705932617,
+      "learning_rate": 5.980445906432749e-06,
+      "loss": 1.3218,
+      "step": 15343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5354440212249756,
+      "learning_rate": 5.97953216374269e-06,
+      "loss": 1.463,
+      "step": 15344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.752847671508789,
+      "learning_rate": 5.9786184210526325e-06,
+      "loss": 1.5239,
+      "step": 15345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4066197872161865,
+      "learning_rate": 5.977704678362574e-06,
+      "loss": 1.2365,
+      "step": 15346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.913111686706543,
+      "learning_rate": 5.976790935672515e-06,
+      "loss": 1.6931,
+      "step": 15347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.693432331085205,
+      "learning_rate": 5.975877192982457e-06,
+      "loss": 1.5011,
+      "step": 15348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.308091402053833,
+      "learning_rate": 5.974963450292398e-06,
+      "loss": 1.3336,
+      "step": 15349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2564187049865723,
+      "learning_rate": 5.974049707602339e-06,
+      "loss": 1.8231,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8804864883422852,
+      "learning_rate": 5.973135964912281e-06,
+      "loss": 0.9958,
+      "step": 15351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.519911527633667,
+      "learning_rate": 5.972222222222222e-06,
+      "loss": 1.5267,
+      "step": 15352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0341668128967285,
+      "learning_rate": 5.971308479532164e-06,
+      "loss": 1.2555,
+      "step": 15353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.18438982963562,
+      "learning_rate": 5.970394736842106e-06,
+      "loss": 1.1118,
+      "step": 15354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.89500093460083,
+      "learning_rate": 5.969480994152047e-06,
+      "loss": 1.5788,
+      "step": 15355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4147427082061768,
+      "learning_rate": 5.9685672514619885e-06,
+      "loss": 1.2941,
+      "step": 15356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3014976978302,
+      "learning_rate": 5.96765350877193e-06,
+      "loss": 1.5218,
+      "step": 15357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9801721572875977,
+      "learning_rate": 5.966739766081871e-06,
+      "loss": 1.6748,
+      "step": 15358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7337424755096436,
+      "learning_rate": 5.965826023391813e-06,
+      "loss": 1.6288,
+      "step": 15359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1755247116088867,
+      "learning_rate": 5.964912280701755e-06,
+      "loss": 1.361,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.484379291534424,
+      "learning_rate": 5.963998538011696e-06,
+      "loss": 1.4968,
+      "step": 15361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5845634937286377,
+      "learning_rate": 5.963084795321638e-06,
+      "loss": 1.6333,
+      "step": 15362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7019925117492676,
+      "learning_rate": 5.962171052631579e-06,
+      "loss": 1.3168,
+      "step": 15363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.177244186401367,
+      "learning_rate": 5.96125730994152e-06,
+      "loss": 1.603,
+      "step": 15364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5286340713500977,
+      "learning_rate": 5.960343567251462e-06,
+      "loss": 1.5948,
+      "step": 15365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2002384662628174,
+      "learning_rate": 5.959429824561403e-06,
+      "loss": 1.4106,
+      "step": 15366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.09297776222229,
+      "learning_rate": 5.958516081871345e-06,
+      "loss": 1.7466,
+      "step": 15367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6347694396972656,
+      "learning_rate": 5.957602339181287e-06,
+      "loss": 1.8095,
+      "step": 15368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.10695481300354,
+      "learning_rate": 5.956688596491229e-06,
+      "loss": 1.2327,
+      "step": 15369
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4761765003204346,
+      "learning_rate": 5.95577485380117e-06,
+      "loss": 1.6028,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.590715169906616,
+      "learning_rate": 5.954861111111112e-06,
+      "loss": 1.5604,
+      "step": 15371
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8269782066345215,
+      "learning_rate": 5.953947368421054e-06,
+      "loss": 1.3441,
+      "step": 15372
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0891497135162354,
+      "learning_rate": 5.953033625730995e-06,
+      "loss": 1.3842,
+      "step": 15373
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5087454319000244,
+      "learning_rate": 5.952119883040937e-06,
+      "loss": 1.6291,
+      "step": 15374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4507241249084473,
+      "learning_rate": 5.951206140350878e-06,
+      "loss": 1.5284,
+      "step": 15375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7168023586273193,
+      "learning_rate": 5.950292397660819e-06,
+      "loss": 1.6921,
+      "step": 15376
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.243001699447632,
+      "learning_rate": 5.949378654970761e-06,
+      "loss": 1.419,
+      "step": 15377
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7823996543884277,
+      "learning_rate": 5.948464912280702e-06,
+      "loss": 1.7713,
+      "step": 15378
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3973190784454346,
+      "learning_rate": 5.947551169590644e-06,
+      "loss": 1.6067,
+      "step": 15379
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.874666690826416,
+      "learning_rate": 5.946637426900586e-06,
+      "loss": 1.6162,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.919877529144287,
+      "learning_rate": 5.945723684210527e-06,
+      "loss": 1.3867,
+      "step": 15381
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3327419757843018,
+      "learning_rate": 5.9448099415204685e-06,
+      "loss": 1.2919,
+      "step": 15382
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1266283988952637,
+      "learning_rate": 5.94389619883041e-06,
+      "loss": 1.3076,
+      "step": 15383
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3625636100769043,
+      "learning_rate": 5.942982456140351e-06,
+      "loss": 1.5109,
+      "step": 15384
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.812061071395874,
+      "learning_rate": 5.942068713450293e-06,
+      "loss": 1.5068,
+      "step": 15385
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4341511726379395,
+      "learning_rate": 5.941154970760235e-06,
+      "loss": 1.5426,
+      "step": 15386
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.681288003921509,
+      "learning_rate": 5.940241228070176e-06,
+      "loss": 1.574,
+      "step": 15387
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.227365016937256,
+      "learning_rate": 5.939327485380118e-06,
+      "loss": 1.2424,
+      "step": 15388
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7581114768981934,
+      "learning_rate": 5.938413742690059e-06,
+      "loss": 1.5698,
+      "step": 15389
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6428699493408203,
+      "learning_rate": 5.9375e-06,
+      "loss": 1.8385,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734494209289551,
+      "learning_rate": 5.936586257309942e-06,
+      "loss": 1.4607,
+      "step": 15391
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5638506412506104,
+      "learning_rate": 5.935672514619883e-06,
+      "loss": 1.5424,
+      "step": 15392
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.534994125366211,
+      "learning_rate": 5.934758771929825e-06,
+      "loss": 1.2804,
+      "step": 15393
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2547659873962402,
+      "learning_rate": 5.933845029239767e-06,
+      "loss": 1.3073,
+      "step": 15394
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3565871715545654,
+      "learning_rate": 5.932931286549708e-06,
+      "loss": 1.03,
+      "step": 15395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7405028343200684,
+      "learning_rate": 5.932017543859649e-06,
+      "loss": 1.6335,
+      "step": 15396
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.145777702331543,
+      "learning_rate": 5.931103801169591e-06,
+      "loss": 1.4069,
+      "step": 15397
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4890856742858887,
+      "learning_rate": 5.930190058479532e-06,
+      "loss": 1.304,
+      "step": 15398
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5833520889282227,
+      "learning_rate": 5.9292763157894735e-06,
+      "loss": 1.4517,
+      "step": 15399
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.17179274559021,
+      "learning_rate": 5.928362573099416e-06,
+      "loss": 1.1249,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6843740940093994,
+      "learning_rate": 5.927448830409357e-06,
+      "loss": 1.2794,
+      "step": 15401
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.459247350692749,
+      "learning_rate": 5.9265350877192985e-06,
+      "loss": 1.4147,
+      "step": 15402
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3346352577209473,
+      "learning_rate": 5.92562134502924e-06,
+      "loss": 1.2184,
+      "step": 15403
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.457446813583374,
+      "learning_rate": 5.924707602339181e-06,
+      "loss": 1.5436,
+      "step": 15404
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.686678886413574,
+      "learning_rate": 5.923793859649123e-06,
+      "loss": 1.6897,
+      "step": 15405
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3555283546447754,
+      "learning_rate": 5.922880116959064e-06,
+      "loss": 1.637,
+      "step": 15406
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3955955505371094,
+      "learning_rate": 5.921966374269006e-06,
+      "loss": 1.7621,
+      "step": 15407
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.740429162979126,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 1.4928,
+      "step": 15408
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.666795253753662,
+      "learning_rate": 5.920138888888889e-06,
+      "loss": 1.3031,
+      "step": 15409
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.767702341079712,
+      "learning_rate": 5.91922514619883e-06,
+      "loss": 1.5243,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.289412498474121,
+      "learning_rate": 5.918311403508772e-06,
+      "loss": 1.4272,
+      "step": 15411
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4500575065612793,
+      "learning_rate": 5.917397660818715e-06,
+      "loss": 1.4437,
+      "step": 15412
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3889830112457275,
+      "learning_rate": 5.916483918128656e-06,
+      "loss": 1.5409,
+      "step": 15413
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8754169940948486,
+      "learning_rate": 5.9155701754385975e-06,
+      "loss": 1.9299,
+      "step": 15414
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8979556560516357,
+      "learning_rate": 5.914656432748539e-06,
+      "loss": 1.6972,
+      "step": 15415
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8073134422302246,
+      "learning_rate": 5.91374269005848e-06,
+      "loss": 1.7478,
+      "step": 15416
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.80757212638855,
+      "learning_rate": 5.912828947368422e-06,
+      "loss": 1.9859,
+      "step": 15417
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0758285522460938,
+      "learning_rate": 5.911915204678363e-06,
+      "loss": 1.6263,
+      "step": 15418
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3638112545013428,
+      "learning_rate": 5.911001461988305e-06,
+      "loss": 1.5017,
+      "step": 15419
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8217406272888184,
+      "learning_rate": 5.910087719298247e-06,
+      "loss": 1.5572,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5861763954162598,
+      "learning_rate": 5.909173976608188e-06,
+      "loss": 1.6982,
+      "step": 15421
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0434770584106445,
+      "learning_rate": 5.908260233918129e-06,
+      "loss": 1.6933,
+      "step": 15422
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7666406631469727,
+      "learning_rate": 5.907346491228071e-06,
+      "loss": 1.78,
+      "step": 15423
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8180878162384033,
+      "learning_rate": 5.906432748538012e-06,
+      "loss": 1.6937,
+      "step": 15424
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.476633310317993,
+      "learning_rate": 5.9055190058479535e-06,
+      "loss": 1.5266,
+      "step": 15425
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.055264472961426,
+      "learning_rate": 5.904605263157896e-06,
+      "loss": 1.2719,
+      "step": 15426
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4110279083251953,
+      "learning_rate": 5.903691520467837e-06,
+      "loss": 1.4913,
+      "step": 15427
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1478404998779297,
+      "learning_rate": 5.9027777777777785e-06,
+      "loss": 1.0917,
+      "step": 15428
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.237325429916382,
+      "learning_rate": 5.90186403508772e-06,
+      "loss": 1.3746,
+      "step": 15429
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5260937213897705,
+      "learning_rate": 5.900950292397661e-06,
+      "loss": 1.1721,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.503873586654663,
+      "learning_rate": 5.900036549707603e-06,
+      "loss": 1.5233,
+      "step": 15431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4436092376708984,
+      "learning_rate": 5.899122807017544e-06,
+      "loss": 1.5612,
+      "step": 15432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.132498025894165,
+      "learning_rate": 5.898209064327486e-06,
+      "loss": 1.2733,
+      "step": 15433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.501899480819702,
+      "learning_rate": 5.897295321637428e-06,
+      "loss": 1.532,
+      "step": 15434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2087595462799072,
+      "learning_rate": 5.896381578947369e-06,
+      "loss": 1.1279,
+      "step": 15435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.709907293319702,
+      "learning_rate": 5.89546783625731e-06,
+      "loss": 1.5051,
+      "step": 15436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3038547039031982,
+      "learning_rate": 5.894554093567252e-06,
+      "loss": 1.5636,
+      "step": 15437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.659114360809326,
+      "learning_rate": 5.893640350877193e-06,
+      "loss": 1.3268,
+      "step": 15438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2473196983337402,
+      "learning_rate": 5.8927266081871344e-06,
+      "loss": 1.3492,
+      "step": 15439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5893993377685547,
+      "learning_rate": 5.891812865497077e-06,
+      "loss": 1.113,
+      "step": 15440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5407423973083496,
+      "learning_rate": 5.890899122807018e-06,
+      "loss": 1.6077,
+      "step": 15441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03361177444458,
+      "learning_rate": 5.889985380116959e-06,
+      "loss": 1.2188,
+      "step": 15442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7424299716949463,
+      "learning_rate": 5.889071637426901e-06,
+      "loss": 1.5868,
+      "step": 15443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.774132013320923,
+      "learning_rate": 5.888157894736842e-06,
+      "loss": 1.3523,
+      "step": 15444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.407935380935669,
+      "learning_rate": 5.8872441520467835e-06,
+      "loss": 1.3657,
+      "step": 15445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.051387310028076,
+      "learning_rate": 5.886330409356725e-06,
+      "loss": 1.5307,
+      "step": 15446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.326537609100342,
+      "learning_rate": 5.885416666666667e-06,
+      "loss": 1.4636,
+      "step": 15447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7457075119018555,
+      "learning_rate": 5.8845029239766085e-06,
+      "loss": 1.6895,
+      "step": 15448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.218634843826294,
+      "learning_rate": 5.88358918128655e-06,
+      "loss": 1.3391,
+      "step": 15449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.588913679122925,
+      "learning_rate": 5.882675438596491e-06,
+      "loss": 1.6911,
+      "step": 15450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.124570608139038,
+      "learning_rate": 5.881761695906433e-06,
+      "loss": 1.02,
+      "step": 15451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.21704363822937,
+      "learning_rate": 5.880847953216374e-06,
+      "loss": 1.3556,
+      "step": 15452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2239632606506348,
+      "learning_rate": 5.879934210526315e-06,
+      "loss": 1.249,
+      "step": 15453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.182011842727661,
+      "learning_rate": 5.879020467836258e-06,
+      "loss": 1.2107,
+      "step": 15454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.856452226638794,
+      "learning_rate": 5.8781067251462e-06,
+      "loss": 1.0816,
+      "step": 15455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.316364049911499,
+      "learning_rate": 5.877192982456141e-06,
+      "loss": 1.4955,
+      "step": 15456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3305654525756836,
+      "learning_rate": 5.876279239766083e-06,
+      "loss": 1.453,
+      "step": 15457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.360447406768799,
+      "learning_rate": 5.875365497076024e-06,
+      "loss": 1.7044,
+      "step": 15458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5637097358703613,
+      "learning_rate": 5.874451754385966e-06,
+      "loss": 1.6873,
+      "step": 15459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6427855491638184,
+      "learning_rate": 5.8735380116959075e-06,
+      "loss": 1.2621,
+      "step": 15460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4091415405273438,
+      "learning_rate": 5.872624269005849e-06,
+      "loss": 1.5478,
+      "step": 15461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.268524169921875,
+      "learning_rate": 5.87171052631579e-06,
+      "loss": 1.2666,
+      "step": 15462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5006706714630127,
+      "learning_rate": 5.870796783625732e-06,
+      "loss": 1.7398,
+      "step": 15463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.861619472503662,
+      "learning_rate": 5.869883040935673e-06,
+      "loss": 1.8259,
+      "step": 15464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.351318359375,
+      "learning_rate": 5.868969298245614e-06,
+      "loss": 1.4004,
+      "step": 15465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.561833143234253,
+      "learning_rate": 5.868055555555557e-06,
+      "loss": 1.3547,
+      "step": 15466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7426700592041016,
+      "learning_rate": 5.867141812865498e-06,
+      "loss": 1.4284,
+      "step": 15467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4504995346069336,
+      "learning_rate": 5.866228070175439e-06,
+      "loss": 1.4375,
+      "step": 15468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6803064346313477,
+      "learning_rate": 5.865314327485381e-06,
+      "loss": 1.503,
+      "step": 15469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2401578426361084,
+      "learning_rate": 5.864400584795322e-06,
+      "loss": 1.1241,
+      "step": 15470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1297543048858643,
+      "learning_rate": 5.8634868421052635e-06,
+      "loss": 1.293,
+      "step": 15471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.889970064163208,
+      "learning_rate": 5.862573099415205e-06,
+      "loss": 1.2873,
+      "step": 15472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2322938442230225,
+      "learning_rate": 5.861659356725147e-06,
+      "loss": 1.3663,
+      "step": 15473
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0281500816345215,
+      "learning_rate": 5.8607456140350885e-06,
+      "loss": 1.3144,
+      "step": 15474
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.557373523712158,
+      "learning_rate": 5.85983187134503e-06,
+      "loss": 1.6507,
+      "step": 15475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.946794033050537,
+      "learning_rate": 5.858918128654971e-06,
+      "loss": 1.7053,
+      "step": 15476
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2987606525421143,
+      "learning_rate": 5.858004385964913e-06,
+      "loss": 2.011,
+      "step": 15477
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5971412658691406,
+      "learning_rate": 5.857090643274854e-06,
+      "loss": 1.5942,
+      "step": 15478
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1036722660064697,
+      "learning_rate": 5.856176900584795e-06,
+      "loss": 1.3557,
+      "step": 15479
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.877089023590088,
+      "learning_rate": 5.855263157894738e-06,
+      "loss": 1.2834,
+      "step": 15480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8412837982177734,
+      "learning_rate": 5.854349415204679e-06,
+      "loss": 1.7427,
+      "step": 15481
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7637076377868652,
+      "learning_rate": 5.85343567251462e-06,
+      "loss": 1.5069,
+      "step": 15482
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6744494438171387,
+      "learning_rate": 5.852521929824562e-06,
+      "loss": 1.5707,
+      "step": 15483
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.355146884918213,
+      "learning_rate": 5.851608187134503e-06,
+      "loss": 1.2184,
+      "step": 15484
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2788891792297363,
+      "learning_rate": 5.8506944444444444e-06,
+      "loss": 1.3024,
+      "step": 15485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3853518962860107,
+      "learning_rate": 5.849780701754386e-06,
+      "loss": 1.5106,
+      "step": 15486
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4510703086853027,
+      "learning_rate": 5.848866959064328e-06,
+      "loss": 1.6127,
+      "step": 15487
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6573290824890137,
+      "learning_rate": 5.847953216374269e-06,
+      "loss": 1.2913,
+      "step": 15488
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.137040615081787,
+      "learning_rate": 5.847039473684211e-06,
+      "loss": 1.4317,
+      "step": 15489
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.401806354522705,
+      "learning_rate": 5.846125730994152e-06,
+      "loss": 1.1881,
+      "step": 15490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.075798273086548,
+      "learning_rate": 5.8452119883040935e-06,
+      "loss": 1.135,
+      "step": 15491
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2268383502960205,
+      "learning_rate": 5.844298245614035e-06,
+      "loss": 1.3157,
+      "step": 15492
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.423617124557495,
+      "learning_rate": 5.843384502923976e-06,
+      "loss": 1.423,
+      "step": 15493
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4039156436920166,
+      "learning_rate": 5.8424707602339185e-06,
+      "loss": 1.4973,
+      "step": 15494
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6440484523773193,
+      "learning_rate": 5.84155701754386e-06,
+      "loss": 1.3245,
+      "step": 15495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5769805908203125,
+      "learning_rate": 5.840643274853801e-06,
+      "loss": 1.3751,
+      "step": 15496
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.243051052093506,
+      "learning_rate": 5.8397295321637435e-06,
+      "loss": 1.2633,
+      "step": 15497
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4472875595092773,
+      "learning_rate": 5.838815789473685e-06,
+      "loss": 1.1072,
+      "step": 15498
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7261507511138916,
+      "learning_rate": 5.837902046783627e-06,
+      "loss": 1.6069,
+      "step": 15499
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.441988468170166,
+      "learning_rate": 5.8369883040935685e-06,
+      "loss": 1.3636,
+      "step": 15500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2099802494049072,
+      "learning_rate": 5.83607456140351e-06,
+      "loss": 1.3313,
+      "step": 15501
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.667855978012085,
+      "learning_rate": 5.835160818713451e-06,
+      "loss": 1.6936,
+      "step": 15502
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.328735589981079,
+      "learning_rate": 5.834247076023393e-06,
+      "loss": 1.406,
+      "step": 15503
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5728299617767334,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.7531,
+      "step": 15504
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.102553129196167,
+      "learning_rate": 5.832419590643275e-06,
+      "loss": 1.0451,
+      "step": 15505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.007822036743164,
+      "learning_rate": 5.8315058479532175e-06,
+      "loss": 1.1925,
+      "step": 15506
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4989876747131348,
+      "learning_rate": 5.830592105263159e-06,
+      "loss": 1.3011,
+      "step": 15507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3884739875793457,
+      "learning_rate": 5.8296783625731e-06,
+      "loss": 1.2699,
+      "step": 15508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.809441566467285,
+      "learning_rate": 5.828764619883042e-06,
+      "loss": 1.7555,
+      "step": 15509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6067569255828857,
+      "learning_rate": 5.827850877192983e-06,
+      "loss": 1.7336,
+      "step": 15510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.587433099746704,
+      "learning_rate": 5.826937134502924e-06,
+      "loss": 1.6616,
+      "step": 15511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.971771717071533,
+      "learning_rate": 5.826023391812866e-06,
+      "loss": 1.677,
+      "step": 15512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5934066772460938,
+      "learning_rate": 5.825109649122808e-06,
+      "loss": 1.2611,
+      "step": 15513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4869306087493896,
+      "learning_rate": 5.824195906432749e-06,
+      "loss": 1.392,
+      "step": 15514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3321964740753174,
+      "learning_rate": 5.823282163742691e-06,
+      "loss": 1.4676,
+      "step": 15515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5097150802612305,
+      "learning_rate": 5.822368421052632e-06,
+      "loss": 1.2295,
+      "step": 15516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.637269973754883,
+      "learning_rate": 5.8214546783625735e-06,
+      "loss": 1.4921,
+      "step": 15517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8051116466522217,
+      "learning_rate": 5.820540935672515e-06,
+      "loss": 1.5677,
+      "step": 15518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9185309410095215,
+      "learning_rate": 5.819627192982456e-06,
+      "loss": 1.6217,
+      "step": 15519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6876277923583984,
+      "learning_rate": 5.8187134502923985e-06,
+      "loss": 1.7084,
+      "step": 15520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6733810901641846,
+      "learning_rate": 5.81779970760234e-06,
+      "loss": 1.6394,
+      "step": 15521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9033992290496826,
+      "learning_rate": 5.816885964912281e-06,
+      "loss": 2.1268,
+      "step": 15522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.399153232574463,
+      "learning_rate": 5.815972222222223e-06,
+      "loss": 1.2303,
+      "step": 15523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4718236923217773,
+      "learning_rate": 5.815058479532164e-06,
+      "loss": 1.2666,
+      "step": 15524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.269496440887451,
+      "learning_rate": 5.814144736842105e-06,
+      "loss": 1.2933,
+      "step": 15525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4046480655670166,
+      "learning_rate": 5.813230994152047e-06,
+      "loss": 1.7928,
+      "step": 15526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.729627847671509,
+      "learning_rate": 5.812317251461989e-06,
+      "loss": 1.4984,
+      "step": 15527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.414243221282959,
+      "learning_rate": 5.81140350877193e-06,
+      "loss": 1.5503,
+      "step": 15528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.862616777420044,
+      "learning_rate": 5.810489766081872e-06,
+      "loss": 1.6186,
+      "step": 15529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8119049072265625,
+      "learning_rate": 5.809576023391813e-06,
+      "loss": 1.7183,
+      "step": 15530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.705716848373413,
+      "learning_rate": 5.8086622807017544e-06,
+      "loss": 1.518,
+      "step": 15531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1550440788269043,
+      "learning_rate": 5.807748538011696e-06,
+      "loss": 1.3862,
+      "step": 15532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.335862636566162,
+      "learning_rate": 5.806834795321637e-06,
+      "loss": 1.3619,
+      "step": 15533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7629809379577637,
+      "learning_rate": 5.805921052631579e-06,
+      "loss": 1.4818,
+      "step": 15534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.589555263519287,
+      "learning_rate": 5.805007309941521e-06,
+      "loss": 1.441,
+      "step": 15535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.357363700866699,
+      "learning_rate": 5.804093567251462e-06,
+      "loss": 1.308,
+      "step": 15536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9145259857177734,
+      "learning_rate": 5.8031798245614035e-06,
+      "loss": 1.7681,
+      "step": 15537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8628218173980713,
+      "learning_rate": 5.802266081871345e-06,
+      "loss": 1.6975,
+      "step": 15538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5523672103881836,
+      "learning_rate": 5.801352339181286e-06,
+      "loss": 1.6262,
+      "step": 15539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9893770217895508,
+      "learning_rate": 5.800438596491229e-06,
+      "loss": 1.4369,
+      "step": 15540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9835326671600342,
+      "learning_rate": 5.799524853801171e-06,
+      "loss": 1.1721,
+      "step": 15541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5536441802978516,
+      "learning_rate": 5.798611111111112e-06,
+      "loss": 1.4121,
+      "step": 15542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.794995069503784,
+      "learning_rate": 5.7976973684210535e-06,
+      "loss": 1.2931,
+      "step": 15543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.838685989379883,
+      "learning_rate": 5.796783625730995e-06,
+      "loss": 1.5583,
+      "step": 15544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4017324447631836,
+      "learning_rate": 5.795869883040936e-06,
+      "loss": 1.5731,
+      "step": 15545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.239185094833374,
+      "learning_rate": 5.7949561403508785e-06,
+      "loss": 1.1996,
+      "step": 15546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.42834734916687,
+      "learning_rate": 5.79404239766082e-06,
+      "loss": 1.3652,
+      "step": 15547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.753208637237549,
+      "learning_rate": 5.793128654970761e-06,
+      "loss": 1.6443,
+      "step": 15548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.585676670074463,
+      "learning_rate": 5.792214912280703e-06,
+      "loss": 1.7921,
+      "step": 15549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4613687992095947,
+      "learning_rate": 5.791301169590644e-06,
+      "loss": 1.5678,
+      "step": 15550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.874084711074829,
+      "learning_rate": 5.790387426900585e-06,
+      "loss": 1.6585,
+      "step": 15551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.41215443611145,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 1.1302,
+      "step": 15552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.35725474357605,
+      "learning_rate": 5.788559941520469e-06,
+      "loss": 1.3187,
+      "step": 15553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.694889783859253,
+      "learning_rate": 5.78764619883041e-06,
+      "loss": 1.5668,
+      "step": 15554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5083439350128174,
+      "learning_rate": 5.786732456140352e-06,
+      "loss": 1.3947,
+      "step": 15555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.237678289413452,
+      "learning_rate": 5.785818713450293e-06,
+      "loss": 1.3026,
+      "step": 15556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9060139656066895,
+      "learning_rate": 5.784904970760234e-06,
+      "loss": 1.51,
+      "step": 15557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6605911254882812,
+      "learning_rate": 5.783991228070176e-06,
+      "loss": 1.1122,
+      "step": 15558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.293595314025879,
+      "learning_rate": 5.783077485380117e-06,
+      "loss": 2.1246,
+      "step": 15559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9394922256469727,
+      "learning_rate": 5.782163742690059e-06,
+      "loss": 1.6696,
+      "step": 15560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9482309818267822,
+      "learning_rate": 5.781250000000001e-06,
+      "loss": 1.5534,
+      "step": 15561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.267305612564087,
+      "learning_rate": 5.780336257309942e-06,
+      "loss": 1.4207,
+      "step": 15562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4173569679260254,
+      "learning_rate": 5.7794225146198835e-06,
+      "loss": 1.4318,
+      "step": 15563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2522084712982178,
+      "learning_rate": 5.778508771929825e-06,
+      "loss": 1.2923,
+      "step": 15564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2673275470733643,
+      "learning_rate": 5.777595029239766e-06,
+      "loss": 0.9548,
+      "step": 15565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.329303741455078,
+      "learning_rate": 5.776681286549708e-06,
+      "loss": 1.4065,
+      "step": 15566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4589672088623047,
+      "learning_rate": 5.77576754385965e-06,
+      "loss": 1.3328,
+      "step": 15567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2402830123901367,
+      "learning_rate": 5.774853801169591e-06,
+      "loss": 1.2659,
+      "step": 15568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7911107540130615,
+      "learning_rate": 5.773940058479533e-06,
+      "loss": 1.7541,
+      "step": 15569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.406543493270874,
+      "learning_rate": 5.773026315789474e-06,
+      "loss": 1.7831,
+      "step": 15570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5402109622955322,
+      "learning_rate": 5.772112573099415e-06,
+      "loss": 1.3665,
+      "step": 15571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.369027614593506,
+      "learning_rate": 5.771198830409357e-06,
+      "loss": 1.3063,
+      "step": 15572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6908440589904785,
+      "learning_rate": 5.770285087719298e-06,
+      "loss": 1.3761,
+      "step": 15573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.317375659942627,
+      "learning_rate": 5.7693713450292395e-06,
+      "loss": 1.4497,
+      "step": 15574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.729562520980835,
+      "learning_rate": 5.768457602339182e-06,
+      "loss": 1.6232,
+      "step": 15575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.045910120010376,
+      "learning_rate": 5.767543859649123e-06,
+      "loss": 1.1167,
+      "step": 15576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8961371183395386,
+      "learning_rate": 5.7666301169590644e-06,
+      "loss": 1.0653,
+      "step": 15577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.275953769683838,
+      "learning_rate": 5.765716374269006e-06,
+      "loss": 1.5883,
+      "step": 15578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.770569086074829,
+      "learning_rate": 5.764802631578947e-06,
+      "loss": 1.8896,
+      "step": 15579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3968300819396973,
+      "learning_rate": 5.7638888888888886e-06,
+      "loss": 1.2032,
+      "step": 15580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.666222095489502,
+      "learning_rate": 5.76297514619883e-06,
+      "loss": 1.628,
+      "step": 15581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2953476905822754,
+      "learning_rate": 5.762061403508772e-06,
+      "loss": 1.1942,
+      "step": 15582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6593830585479736,
+      "learning_rate": 5.761147660818714e-06,
+      "loss": 1.5068,
+      "step": 15583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.069941997528076,
+      "learning_rate": 5.760233918128656e-06,
+      "loss": 2.008,
+      "step": 15584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.45192551612854,
+      "learning_rate": 5.759320175438597e-06,
+      "loss": 1.2074,
+      "step": 15585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.205716609954834,
+      "learning_rate": 5.758406432748539e-06,
+      "loss": 1.2449,
+      "step": 15586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3408682346343994,
+      "learning_rate": 5.757492690058481e-06,
+      "loss": 1.3249,
+      "step": 15587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5266029834747314,
+      "learning_rate": 5.756578947368422e-06,
+      "loss": 1.5929,
+      "step": 15588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.228416681289673,
+      "learning_rate": 5.7556652046783635e-06,
+      "loss": 1.6812,
+      "step": 15589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.196148157119751,
+      "learning_rate": 5.754751461988305e-06,
+      "loss": 1.2366,
+      "step": 15590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.832719326019287,
+      "learning_rate": 5.753837719298246e-06,
+      "loss": 1.468,
+      "step": 15591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.954101324081421,
+      "learning_rate": 5.752923976608188e-06,
+      "loss": 1.3787,
+      "step": 15592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4805948734283447,
+      "learning_rate": 5.75201023391813e-06,
+      "loss": 1.9929,
+      "step": 15593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2884247303009033,
+      "learning_rate": 5.751096491228071e-06,
+      "loss": 1.2336,
+      "step": 15594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8201395273208618,
+      "learning_rate": 5.750182748538013e-06,
+      "loss": 0.971,
+      "step": 15595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.575382947921753,
+      "learning_rate": 5.749269005847954e-06,
+      "loss": 1.3452,
+      "step": 15596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4603683948516846,
+      "learning_rate": 5.748355263157895e-06,
+      "loss": 1.4511,
+      "step": 15597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.200573205947876,
+      "learning_rate": 5.747441520467837e-06,
+      "loss": 1.2676,
+      "step": 15598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0755608081817627,
+      "learning_rate": 5.746527777777778e-06,
+      "loss": 1.0578,
+      "step": 15599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8251757621765137,
+      "learning_rate": 5.7456140350877194e-06,
+      "loss": 1.8072,
+      "step": 15600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3685004711151123,
+      "learning_rate": 5.744700292397662e-06,
+      "loss": 1.3923,
+      "step": 15601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.509058952331543,
+      "learning_rate": 5.743786549707603e-06,
+      "loss": 1.5362,
+      "step": 15602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5480494499206543,
+      "learning_rate": 5.742872807017544e-06,
+      "loss": 1.3806,
+      "step": 15603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.509284734725952,
+      "learning_rate": 5.741959064327486e-06,
+      "loss": 1.4915,
+      "step": 15604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1185190677642822,
+      "learning_rate": 5.741045321637427e-06,
+      "loss": 1.1734,
+      "step": 15605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.229398727416992,
+      "learning_rate": 5.7401315789473685e-06,
+      "loss": 1.3367,
+      "step": 15606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9529008865356445,
+      "learning_rate": 5.73921783625731e-06,
+      "loss": 1.419,
+      "step": 15607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3998525142669678,
+      "learning_rate": 5.738304093567252e-06,
+      "loss": 1.3343,
+      "step": 15608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6254782676696777,
+      "learning_rate": 5.7373903508771935e-06,
+      "loss": 1.5656,
+      "step": 15609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0564675331115723,
+      "learning_rate": 5.736476608187135e-06,
+      "loss": 1.2782,
+      "step": 15610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6037745475769043,
+      "learning_rate": 5.735562865497076e-06,
+      "loss": 1.2685,
+      "step": 15611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8726890087127686,
+      "learning_rate": 5.734649122807018e-06,
+      "loss": 1.398,
+      "step": 15612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1833279132843018,
+      "learning_rate": 5.733735380116959e-06,
+      "loss": 1.3556,
+      "step": 15613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5546107292175293,
+      "learning_rate": 5.7328216374269e-06,
+      "loss": 1.4804,
+      "step": 15614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5900800228118896,
+      "learning_rate": 5.731907894736843e-06,
+      "loss": 1.6306,
+      "step": 15615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.572788953781128,
+      "learning_rate": 5.730994152046784e-06,
+      "loss": 1.7114,
+      "step": 15616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.476400852203369,
+      "learning_rate": 5.730080409356725e-06,
+      "loss": 1.3217,
+      "step": 15617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.335620641708374,
+      "learning_rate": 5.729166666666667e-06,
+      "loss": 1.405,
+      "step": 15618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4962475299835205,
+      "learning_rate": 5.728252923976608e-06,
+      "loss": 1.3213,
+      "step": 15619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.362150192260742,
+      "learning_rate": 5.7273391812865495e-06,
+      "loss": 1.2132,
+      "step": 15620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1328046321868896,
+      "learning_rate": 5.726425438596491e-06,
+      "loss": 1.2516,
+      "step": 15621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5674545764923096,
+      "learning_rate": 5.725511695906433e-06,
+      "loss": 1.5056,
+      "step": 15622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4931273460388184,
+      "learning_rate": 5.7245979532163744e-06,
+      "loss": 1.8217,
+      "step": 15623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.156507730484009,
+      "learning_rate": 5.723684210526316e-06,
+      "loss": 1.3264,
+      "step": 15624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.697887659072876,
+      "learning_rate": 5.722770467836257e-06,
+      "loss": 1.3447,
+      "step": 15625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.577821731567383,
+      "learning_rate": 5.721856725146199e-06,
+      "loss": 1.4712,
+      "step": 15626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2957444190979004,
+      "learning_rate": 5.720942982456142e-06,
+      "loss": 1.4675,
+      "step": 15627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.063277244567871,
+      "learning_rate": 5.720029239766083e-06,
+      "loss": 1.2626,
+      "step": 15628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.101304531097412,
+      "learning_rate": 5.719115497076024e-06,
+      "loss": 1.2565,
+      "step": 15629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7321372032165527,
+      "learning_rate": 5.718201754385966e-06,
+      "loss": 1.4476,
+      "step": 15630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.637269973754883,
+      "learning_rate": 5.717288011695907e-06,
+      "loss": 1.7213,
+      "step": 15631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.412616014480591,
+      "learning_rate": 5.7163742690058485e-06,
+      "loss": 1.273,
+      "step": 15632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.400758981704712,
+      "learning_rate": 5.71546052631579e-06,
+      "loss": 1.5203,
+      "step": 15633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4204447269439697,
+      "learning_rate": 5.714546783625732e-06,
+      "loss": 1.7114,
+      "step": 15634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1976170539855957,
+      "learning_rate": 5.7136330409356735e-06,
+      "loss": 1.3866,
+      "step": 15635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.129403591156006,
+      "learning_rate": 5.712719298245615e-06,
+      "loss": 1.6487,
+      "step": 15636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.101043939590454,
+      "learning_rate": 5.711805555555556e-06,
+      "loss": 1.6894,
+      "step": 15637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6248066425323486,
+      "learning_rate": 5.710891812865498e-06,
+      "loss": 1.6044,
+      "step": 15638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9166219234466553,
+      "learning_rate": 5.709978070175439e-06,
+      "loss": 1.7168,
+      "step": 15639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5218207836151123,
+      "learning_rate": 5.70906432748538e-06,
+      "loss": 1.6272,
+      "step": 15640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0862977504730225,
+      "learning_rate": 5.7081505847953226e-06,
+      "loss": 1.3962,
+      "step": 15641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.846285343170166,
+      "learning_rate": 5.707236842105264e-06,
+      "loss": 1.8077,
+      "step": 15642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.48063588142395,
+      "learning_rate": 5.706323099415205e-06,
+      "loss": 1.4276,
+      "step": 15643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2461276054382324,
+      "learning_rate": 5.705409356725147e-06,
+      "loss": 1.3485,
+      "step": 15644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3608334064483643,
+      "learning_rate": 5.704495614035088e-06,
+      "loss": 1.3984,
+      "step": 15645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.866521120071411,
+      "learning_rate": 5.7035818713450294e-06,
+      "loss": 1.702,
+      "step": 15646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.850353479385376,
+      "learning_rate": 5.702668128654971e-06,
+      "loss": 1.721,
+      "step": 15647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4078917503356934,
+      "learning_rate": 5.701754385964913e-06,
+      "loss": 1.4428,
+      "step": 15648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7530598640441895,
+      "learning_rate": 5.700840643274854e-06,
+      "loss": 1.667,
+      "step": 15649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.353376626968384,
+      "learning_rate": 5.699926900584796e-06,
+      "loss": 1.5011,
+      "step": 15650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.685601234436035,
+      "learning_rate": 5.699013157894737e-06,
+      "loss": 1.9558,
+      "step": 15651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6210744380950928,
+      "learning_rate": 5.6980994152046785e-06,
+      "loss": 1.4276,
+      "step": 15652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5227675437927246,
+      "learning_rate": 5.69718567251462e-06,
+      "loss": 1.2806,
+      "step": 15653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.500277042388916,
+      "learning_rate": 5.696271929824561e-06,
+      "loss": 1.6304,
+      "step": 15654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.578183889389038,
+      "learning_rate": 5.6953581871345035e-06,
+      "loss": 1.4419,
+      "step": 15655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8370308876037598,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 1.6458,
+      "step": 15656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8556722402572632,
+      "learning_rate": 5.693530701754386e-06,
+      "loss": 1.1862,
+      "step": 15657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5275959968566895,
+      "learning_rate": 5.692616959064328e-06,
+      "loss": 1.2211,
+      "step": 15658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.619316577911377,
+      "learning_rate": 5.691703216374269e-06,
+      "loss": 1.6441,
+      "step": 15659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.537039279937744,
+      "learning_rate": 5.69078947368421e-06,
+      "loss": 1.5763,
+      "step": 15660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.627256155014038,
+      "learning_rate": 5.689875730994152e-06,
+      "loss": 1.6424,
+      "step": 15661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.918273448944092,
+      "learning_rate": 5.688961988304094e-06,
+      "loss": 1.7953,
+      "step": 15662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5744872093200684,
+      "learning_rate": 5.688048245614035e-06,
+      "loss": 1.471,
+      "step": 15663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.311579942703247,
+      "learning_rate": 5.687134502923977e-06,
+      "loss": 1.3564,
+      "step": 15664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0227608680725098,
+      "learning_rate": 5.686220760233918e-06,
+      "loss": 1.2456,
+      "step": 15665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4045066833496094,
+      "learning_rate": 5.6853070175438595e-06,
+      "loss": 1.4656,
+      "step": 15666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.380510091781616,
+      "learning_rate": 5.684393274853801e-06,
+      "loss": 1.5513,
+      "step": 15667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.536719799041748,
+      "learning_rate": 5.683479532163744e-06,
+      "loss": 1.3499,
+      "step": 15668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5472805500030518,
+      "learning_rate": 5.682565789473685e-06,
+      "loss": 1.2191,
+      "step": 15669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1875741481781006,
+      "learning_rate": 5.681652046783627e-06,
+      "loss": 1.0286,
+      "step": 15670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.392909288406372,
+      "learning_rate": 5.680738304093568e-06,
+      "loss": 1.2926,
+      "step": 15671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.401737928390503,
+      "learning_rate": 5.679824561403509e-06,
+      "loss": 1.479,
+      "step": 15672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6960020065307617,
+      "learning_rate": 5.678910818713451e-06,
+      "loss": 1.6327,
+      "step": 15673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3513784408569336,
+      "learning_rate": 5.677997076023393e-06,
+      "loss": 1.4882,
+      "step": 15674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.196528196334839,
+      "learning_rate": 5.677083333333334e-06,
+      "loss": 1.3305,
+      "step": 15675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4567854404449463,
+      "learning_rate": 5.676169590643276e-06,
+      "loss": 1.5547,
+      "step": 15676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2159976959228516,
+      "learning_rate": 5.675255847953217e-06,
+      "loss": 1.5444,
+      "step": 15677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.452043294906616,
+      "learning_rate": 5.6743421052631585e-06,
+      "loss": 1.2844,
+      "step": 15678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4076976776123047,
+      "learning_rate": 5.6734283625731e-06,
+      "loss": 1.5063,
+      "step": 15679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1887857913970947,
+      "learning_rate": 5.672514619883041e-06,
+      "loss": 1.3012,
+      "step": 15680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.813493251800537,
+      "learning_rate": 5.6716008771929835e-06,
+      "loss": 1.6661,
+      "step": 15681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7810066938400269,
+      "learning_rate": 5.670687134502925e-06,
+      "loss": 0.9318,
+      "step": 15682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4346230030059814,
+      "learning_rate": 5.669773391812866e-06,
+      "loss": 1.5189,
+      "step": 15683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.336806058883667,
+      "learning_rate": 5.668859649122808e-06,
+      "loss": 1.3059,
+      "step": 15684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4651083946228027,
+      "learning_rate": 5.667945906432749e-06,
+      "loss": 1.4047,
+      "step": 15685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6976187229156494,
+      "learning_rate": 5.66703216374269e-06,
+      "loss": 1.4528,
+      "step": 15686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.760476589202881,
+      "learning_rate": 5.666118421052632e-06,
+      "loss": 1.494,
+      "step": 15687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8072996139526367,
+      "learning_rate": 5.665204678362574e-06,
+      "loss": 1.133,
+      "step": 15688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5428977012634277,
+      "learning_rate": 5.664290935672515e-06,
+      "loss": 1.5369,
+      "step": 15689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9128477573394775,
+      "learning_rate": 5.663377192982457e-06,
+      "loss": 1.3646,
+      "step": 15690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.582656145095825,
+      "learning_rate": 5.662463450292398e-06,
+      "loss": 1.3592,
+      "step": 15691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6645290851593018,
+      "learning_rate": 5.6615497076023394e-06,
+      "loss": 1.6436,
+      "step": 15692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.863553524017334,
+      "learning_rate": 5.660635964912281e-06,
+      "loss": 1.9546,
+      "step": 15693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6796975135803223,
+      "learning_rate": 5.659722222222222e-06,
+      "loss": 1.6611,
+      "step": 15694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.427746057510376,
+      "learning_rate": 5.658808479532164e-06,
+      "loss": 1.3513,
+      "step": 15695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.906589984893799,
+      "learning_rate": 5.657894736842106e-06,
+      "loss": 1.6956,
+      "step": 15696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.516130208969116,
+      "learning_rate": 5.656980994152047e-06,
+      "loss": 1.3762,
+      "step": 15697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.657759666442871,
+      "learning_rate": 5.6560672514619885e-06,
+      "loss": 1.5534,
+      "step": 15698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.510507345199585,
+      "learning_rate": 5.65515350877193e-06,
+      "loss": 1.4147,
+      "step": 15699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.469437599182129,
+      "learning_rate": 5.654239766081871e-06,
+      "loss": 1.3004,
+      "step": 15700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6779773235321045,
+      "learning_rate": 5.653326023391813e-06,
+      "loss": 1.3699,
+      "step": 15701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.49525785446167,
+      "learning_rate": 5.652412280701755e-06,
+      "loss": 1.406,
+      "step": 15702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.148442268371582,
+      "learning_rate": 5.651498538011696e-06,
+      "loss": 1.2896,
+      "step": 15703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.104008913040161,
+      "learning_rate": 5.650584795321638e-06,
+      "loss": 1.5103,
+      "step": 15704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.014718770980835,
+      "learning_rate": 5.649671052631579e-06,
+      "loss": 1.3183,
+      "step": 15705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.807687282562256,
+      "learning_rate": 5.64875730994152e-06,
+      "loss": 1.3895,
+      "step": 15706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.563201427459717,
+      "learning_rate": 5.647843567251462e-06,
+      "loss": 1.407,
+      "step": 15707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5854501724243164,
+      "learning_rate": 5.646929824561403e-06,
+      "loss": 1.6938,
+      "step": 15708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.311446189880371,
+      "learning_rate": 5.646016081871345e-06,
+      "loss": 1.3357,
+      "step": 15709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2575643062591553,
+      "learning_rate": 5.645102339181287e-06,
+      "loss": 1.8504,
+      "step": 15710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.170297622680664,
+      "learning_rate": 5.644188596491229e-06,
+      "loss": 1.3409,
+      "step": 15711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1359689235687256,
+      "learning_rate": 5.64327485380117e-06,
+      "loss": 1.1288,
+      "step": 15712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.325863838195801,
+      "learning_rate": 5.642361111111112e-06,
+      "loss": 1.3844,
+      "step": 15713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.388248920440674,
+      "learning_rate": 5.641447368421054e-06,
+      "loss": 1.2753,
+      "step": 15714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2834315299987793,
+      "learning_rate": 5.640533625730995e-06,
+      "loss": 1.2503,
+      "step": 15715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0326457023620605,
+      "learning_rate": 5.639619883040937e-06,
+      "loss": 1.0723,
+      "step": 15716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3640084266662598,
+      "learning_rate": 5.638706140350878e-06,
+      "loss": 1.4438,
+      "step": 15717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3867251873016357,
+      "learning_rate": 5.637792397660819e-06,
+      "loss": 1.3018,
+      "step": 15718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.681205987930298,
+      "learning_rate": 5.636878654970761e-06,
+      "loss": 1.6137,
+      "step": 15719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2119686603546143,
+      "learning_rate": 5.635964912280702e-06,
+      "loss": 1.1426,
+      "step": 15720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.791877031326294,
+      "learning_rate": 5.635051169590644e-06,
+      "loss": 1.5955,
+      "step": 15721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.282787799835205,
+      "learning_rate": 5.634137426900586e-06,
+      "loss": 1.8304,
+      "step": 15722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2246551513671875,
+      "learning_rate": 5.633223684210527e-06,
+      "loss": 1.2071,
+      "step": 15723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6983163356781006,
+      "learning_rate": 5.6323099415204685e-06,
+      "loss": 1.2087,
+      "step": 15724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5908164978027344,
+      "learning_rate": 5.63139619883041e-06,
+      "loss": 1.5947,
+      "step": 15725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.434450149536133,
+      "learning_rate": 5.630482456140351e-06,
+      "loss": 1.7021,
+      "step": 15726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0282390117645264,
+      "learning_rate": 5.629568713450293e-06,
+      "loss": 0.9828,
+      "step": 15727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.044590711593628,
+      "learning_rate": 5.628654970760235e-06,
+      "loss": 1.1266,
+      "step": 15728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5435898303985596,
+      "learning_rate": 5.627741228070176e-06,
+      "loss": 1.5963,
+      "step": 15729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4807686805725098,
+      "learning_rate": 5.626827485380118e-06,
+      "loss": 1.4867,
+      "step": 15730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1703548431396484,
+      "learning_rate": 5.625913742690059e-06,
+      "loss": 1.0573,
+      "step": 15731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3337388038635254,
+      "learning_rate": 5.625e-06,
+      "loss": 1.2837,
+      "step": 15732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.144988775253296,
+      "learning_rate": 5.624086257309942e-06,
+      "loss": 1.2825,
+      "step": 15733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3106331825256348,
+      "learning_rate": 5.623172514619883e-06,
+      "loss": 1.2074,
+      "step": 15734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.198870897293091,
+      "learning_rate": 5.622258771929825e-06,
+      "loss": 1.4384,
+      "step": 15735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.067490577697754,
+      "learning_rate": 5.621345029239767e-06,
+      "loss": 1.1316,
+      "step": 15736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2777085304260254,
+      "learning_rate": 5.620431286549708e-06,
+      "loss": 1.2084,
+      "step": 15737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4877023696899414,
+      "learning_rate": 5.6195175438596494e-06,
+      "loss": 1.3174,
+      "step": 15738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8028178215026855,
+      "learning_rate": 5.618603801169591e-06,
+      "loss": 1.8896,
+      "step": 15739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4735569953918457,
+      "learning_rate": 5.617690058479532e-06,
+      "loss": 1.2331,
+      "step": 15740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6210875511169434,
+      "learning_rate": 5.6167763157894736e-06,
+      "loss": 1.5143,
+      "step": 15741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3051464557647705,
+      "learning_rate": 5.615862573099416e-06,
+      "loss": 1.6496,
+      "step": 15742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9251201152801514,
+      "learning_rate": 5.614948830409357e-06,
+      "loss": 1.529,
+      "step": 15743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.717642307281494,
+      "learning_rate": 5.6140350877192985e-06,
+      "loss": 1.7613,
+      "step": 15744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5643060207366943,
+      "learning_rate": 5.61312134502924e-06,
+      "loss": 1.8189,
+      "step": 15745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.50728178024292,
+      "learning_rate": 5.612207602339181e-06,
+      "loss": 1.4847,
+      "step": 15746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.168752670288086,
+      "learning_rate": 5.611293859649123e-06,
+      "loss": 1.3368,
+      "step": 15747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2579429149627686,
+      "learning_rate": 5.610380116959064e-06,
+      "loss": 1.4148,
+      "step": 15748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0524933338165283,
+      "learning_rate": 5.609466374269006e-06,
+      "loss": 1.4115,
+      "step": 15749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.149397611618042,
+      "learning_rate": 5.608552631578948e-06,
+      "loss": 1.6088,
+      "step": 15750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.425198793411255,
+      "learning_rate": 5.607638888888889e-06,
+      "loss": 1.753,
+      "step": 15751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.13157057762146,
+      "learning_rate": 5.60672514619883e-06,
+      "loss": 1.1261,
+      "step": 15752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8905811309814453,
+      "learning_rate": 5.605811403508772e-06,
+      "loss": 1.7055,
+      "step": 15753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.088787794113159,
+      "learning_rate": 5.604897660818715e-06,
+      "loss": 1.2627,
+      "step": 15754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8666446208953857,
+      "learning_rate": 5.603983918128656e-06,
+      "loss": 1.5276,
+      "step": 15755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.311631679534912,
+      "learning_rate": 5.6030701754385976e-06,
+      "loss": 1.1174,
+      "step": 15756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.873893976211548,
+      "learning_rate": 5.602156432748539e-06,
+      "loss": 1.6127,
+      "step": 15757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0260138511657715,
+      "learning_rate": 5.60124269005848e-06,
+      "loss": 1.5039,
+      "step": 15758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7927048206329346,
+      "learning_rate": 5.600328947368422e-06,
+      "loss": 1.2991,
+      "step": 15759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.624868631362915,
+      "learning_rate": 5.599415204678363e-06,
+      "loss": 1.2014,
+      "step": 15760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.752645969390869,
+      "learning_rate": 5.598501461988305e-06,
+      "loss": 1.8719,
+      "step": 15761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7748820781707764,
+      "learning_rate": 5.597587719298247e-06,
+      "loss": 1.6881,
+      "step": 15762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.975433111190796,
+      "learning_rate": 5.596673976608188e-06,
+      "loss": 1.6035,
+      "step": 15763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.367724657058716,
+      "learning_rate": 5.595760233918129e-06,
+      "loss": 1.2063,
+      "step": 15764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7979564666748047,
+      "learning_rate": 5.594846491228071e-06,
+      "loss": 1.6171,
+      "step": 15765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.54819917678833,
+      "learning_rate": 5.593932748538012e-06,
+      "loss": 1.4918,
+      "step": 15766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4184720516204834,
+      "learning_rate": 5.5930190058479535e-06,
+      "loss": 1.4495,
+      "step": 15767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4523098468780518,
+      "learning_rate": 5.592105263157896e-06,
+      "loss": 1.3759,
+      "step": 15768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6678693294525146,
+      "learning_rate": 5.591191520467837e-06,
+      "loss": 1.6766,
+      "step": 15769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6961333751678467,
+      "learning_rate": 5.5902777777777785e-06,
+      "loss": 1.7451,
+      "step": 15770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0553324222564697,
+      "learning_rate": 5.58936403508772e-06,
+      "loss": 1.4073,
+      "step": 15771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.666496515274048,
+      "learning_rate": 5.588450292397661e-06,
+      "loss": 1.6531,
+      "step": 15772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5875465869903564,
+      "learning_rate": 5.587536549707603e-06,
+      "loss": 1.442,
+      "step": 15773
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1707465648651123,
+      "learning_rate": 5.586622807017544e-06,
+      "loss": 1.5746,
+      "step": 15774
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.225372076034546,
+      "learning_rate": 5.585709064327486e-06,
+      "loss": 1.7218,
+      "step": 15775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.42172908782959,
+      "learning_rate": 5.584795321637428e-06,
+      "loss": 1.2893,
+      "step": 15776
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4155473709106445,
+      "learning_rate": 5.583881578947369e-06,
+      "loss": 2.131,
+      "step": 15777
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5366604328155518,
+      "learning_rate": 5.58296783625731e-06,
+      "loss": 1.5419,
+      "step": 15778
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.265807867050171,
+      "learning_rate": 5.582054093567252e-06,
+      "loss": 1.3554,
+      "step": 15779
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8254860639572144,
+      "learning_rate": 5.581140350877193e-06,
+      "loss": 1.0126,
+      "step": 15780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9835495948791504,
+      "learning_rate": 5.5802266081871345e-06,
+      "loss": 1.055,
+      "step": 15781
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.034327268600464,
+      "learning_rate": 5.579312865497077e-06,
+      "loss": 1.2627,
+      "step": 15782
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3204782009124756,
+      "learning_rate": 5.578399122807018e-06,
+      "loss": 1.3796,
+      "step": 15783
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3788340091705322,
+      "learning_rate": 5.5774853801169594e-06,
+      "loss": 1.3748,
+      "step": 15784
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5155985355377197,
+      "learning_rate": 5.576571637426901e-06,
+      "loss": 1.5622,
+      "step": 15785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.743155002593994,
+      "learning_rate": 5.575657894736842e-06,
+      "loss": 1.5896,
+      "step": 15786
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.67034649848938,
+      "learning_rate": 5.5747441520467836e-06,
+      "loss": 1.335,
+      "step": 15787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.284674882888794,
+      "learning_rate": 5.573830409356725e-06,
+      "loss": 1.436,
+      "step": 15788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5856854915618896,
+      "learning_rate": 5.572916666666667e-06,
+      "loss": 1.5644,
+      "step": 15789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.524212598800659,
+      "learning_rate": 5.5720029239766085e-06,
+      "loss": 1.3539,
+      "step": 15790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8330633640289307,
+      "learning_rate": 5.57108918128655e-06,
+      "loss": 1.5687,
+      "step": 15791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.808713436126709,
+      "learning_rate": 5.570175438596491e-06,
+      "loss": 1.4849,
+      "step": 15792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.134613513946533,
+      "learning_rate": 5.569261695906433e-06,
+      "loss": 1.7345,
+      "step": 15793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4275832176208496,
+      "learning_rate": 5.568347953216374e-06,
+      "loss": 1.33,
+      "step": 15794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2269139289855957,
+      "learning_rate": 5.567434210526315e-06,
+      "loss": 1.1395,
+      "step": 15795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5989415645599365,
+      "learning_rate": 5.566520467836258e-06,
+      "loss": 1.8374,
+      "step": 15796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.377406358718872,
+      "learning_rate": 5.5656067251462e-06,
+      "loss": 1.4474,
+      "step": 15797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.487194538116455,
+      "learning_rate": 5.564692982456141e-06,
+      "loss": 1.2205,
+      "step": 15798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.301098108291626,
+      "learning_rate": 5.563779239766083e-06,
+      "loss": 2.2637,
+      "step": 15799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2818028926849365,
+      "learning_rate": 5.562865497076024e-06,
+      "loss": 1.415,
+      "step": 15800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7137694358825684,
+      "learning_rate": 5.561951754385966e-06,
+      "loss": 1.714,
+      "step": 15801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.859135866165161,
+      "learning_rate": 5.5610380116959076e-06,
+      "loss": 1.7293,
+      "step": 15802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.910004138946533,
+      "learning_rate": 5.560124269005849e-06,
+      "loss": 1.3608,
+      "step": 15803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2328901290893555,
+      "learning_rate": 5.55921052631579e-06,
+      "loss": 1.2868,
+      "step": 15804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7353227138519287,
+      "learning_rate": 5.558296783625732e-06,
+      "loss": 1.5696,
+      "step": 15805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3390228748321533,
+      "learning_rate": 5.557383040935673e-06,
+      "loss": 1.2536,
+      "step": 15806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.396815061569214,
+      "learning_rate": 5.5564692982456144e-06,
+      "loss": 1.5569,
+      "step": 15807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.741669178009033,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.373,
+      "step": 15808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9166507720947266,
+      "learning_rate": 5.554641812865498e-06,
+      "loss": 1.6317,
+      "step": 15809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.204146385192871,
+      "learning_rate": 5.553728070175439e-06,
+      "loss": 1.1286,
+      "step": 15810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.351773738861084,
+      "learning_rate": 5.552814327485381e-06,
+      "loss": 1.2914,
+      "step": 15811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9325437545776367,
+      "learning_rate": 5.551900584795322e-06,
+      "loss": 1.365,
+      "step": 15812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3209307193756104,
+      "learning_rate": 5.5509868421052635e-06,
+      "loss": 1.2123,
+      "step": 15813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6649727821350098,
+      "learning_rate": 5.550073099415205e-06,
+      "loss": 1.6378,
+      "step": 15814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.499685764312744,
+      "learning_rate": 5.549159356725147e-06,
+      "loss": 1.5786,
+      "step": 15815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.234938383102417,
+      "learning_rate": 5.5482456140350885e-06,
+      "loss": 1.2493,
+      "step": 15816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.69272780418396,
+      "learning_rate": 5.54733187134503e-06,
+      "loss": 1.4894,
+      "step": 15817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3955156803131104,
+      "learning_rate": 5.546418128654971e-06,
+      "loss": 1.3776,
+      "step": 15818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.727705240249634,
+      "learning_rate": 5.545504385964913e-06,
+      "loss": 1.4166,
+      "step": 15819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7205684185028076,
+      "learning_rate": 5.544590643274854e-06,
+      "loss": 1.4319,
+      "step": 15820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3708043098449707,
+      "learning_rate": 5.543676900584795e-06,
+      "loss": 1.3433,
+      "step": 15821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5483486652374268,
+      "learning_rate": 5.542763157894737e-06,
+      "loss": 1.3095,
+      "step": 15822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9987982511520386,
+      "learning_rate": 5.541849415204679e-06,
+      "loss": 1.195,
+      "step": 15823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8746235370635986,
+      "learning_rate": 5.54093567251462e-06,
+      "loss": 1.4801,
+      "step": 15824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7107419967651367,
+      "learning_rate": 5.540021929824562e-06,
+      "loss": 1.5572,
+      "step": 15825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.266092538833618,
+      "learning_rate": 5.539108187134503e-06,
+      "loss": 1.3235,
+      "step": 15826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.875434398651123,
+      "learning_rate": 5.5381944444444445e-06,
+      "loss": 1.4981,
+      "step": 15827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4216437339782715,
+      "learning_rate": 5.537280701754386e-06,
+      "loss": 1.4354,
+      "step": 15828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1720943450927734,
+      "learning_rate": 5.536366959064327e-06,
+      "loss": 1.2431,
+      "step": 15829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4289450645446777,
+      "learning_rate": 5.5354532163742694e-06,
+      "loss": 1.3906,
+      "step": 15830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9396084547042847,
+      "learning_rate": 5.534539473684211e-06,
+      "loss": 1.0338,
+      "step": 15831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.086235523223877,
+      "learning_rate": 5.533625730994152e-06,
+      "loss": 1.369,
+      "step": 15832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.620995283126831,
+      "learning_rate": 5.5327119883040936e-06,
+      "loss": 1.3459,
+      "step": 15833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6816227436065674,
+      "learning_rate": 5.531798245614035e-06,
+      "loss": 1.8608,
+      "step": 15834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0440027713775635,
+      "learning_rate": 5.530884502923976e-06,
+      "loss": 1.648,
+      "step": 15835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6084887981414795,
+      "learning_rate": 5.529970760233918e-06,
+      "loss": 1.5642,
+      "step": 15836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.358266830444336,
+      "learning_rate": 5.52905701754386e-06,
+      "loss": 1.2388,
+      "step": 15837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.730473756790161,
+      "learning_rate": 5.528143274853801e-06,
+      "loss": 1.1663,
+      "step": 15838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5339152812957764,
+      "learning_rate": 5.5272295321637435e-06,
+      "loss": 1.6584,
+      "step": 15839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.606196641921997,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 1.6114,
+      "step": 15840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0809032917022705,
+      "learning_rate": 5.525402046783626e-06,
+      "loss": 1.111,
+      "step": 15841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2256271839141846,
+      "learning_rate": 5.5244883040935685e-06,
+      "loss": 1.3485,
+      "step": 15842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2103917598724365,
+      "learning_rate": 5.52357456140351e-06,
+      "loss": 1.8081,
+      "step": 15843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.99497652053833,
+      "learning_rate": 5.522660818713451e-06,
+      "loss": 2.1764,
+      "step": 15844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1164114475250244,
+      "learning_rate": 5.521747076023393e-06,
+      "loss": 1.2368,
+      "step": 15845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4031825065612793,
+      "learning_rate": 5.520833333333334e-06,
+      "loss": 1.1376,
+      "step": 15846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.501279592514038,
+      "learning_rate": 5.519919590643275e-06,
+      "loss": 1.3301,
+      "step": 15847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3417327404022217,
+      "learning_rate": 5.519005847953217e-06,
+      "loss": 1.3041,
+      "step": 15848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5142056941986084,
+      "learning_rate": 5.518092105263159e-06,
+      "loss": 1.4385,
+      "step": 15849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.562168836593628,
+      "learning_rate": 5.5171783625731e-06,
+      "loss": 1.2056,
+      "step": 15850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4149978160858154,
+      "learning_rate": 5.516264619883042e-06,
+      "loss": 1.2216,
+      "step": 15851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.557819366455078,
+      "learning_rate": 5.515350877192983e-06,
+      "loss": 1.3445,
+      "step": 15852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.629887104034424,
+      "learning_rate": 5.5144371345029244e-06,
+      "loss": 1.5313,
+      "step": 15853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.288205623626709,
+      "learning_rate": 5.513523391812866e-06,
+      "loss": 1.0744,
+      "step": 15854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9921817779541016,
+      "learning_rate": 5.512609649122807e-06,
+      "loss": 1.5783,
+      "step": 15855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.325725793838501,
+      "learning_rate": 5.511695906432749e-06,
+      "loss": 1.3714,
+      "step": 15856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7995288372039795,
+      "learning_rate": 5.510782163742691e-06,
+      "loss": 1.6679,
+      "step": 15857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.581820487976074,
+      "learning_rate": 5.509868421052632e-06,
+      "loss": 1.6438,
+      "step": 15858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.616891860961914,
+      "learning_rate": 5.5089546783625735e-06,
+      "loss": 1.501,
+      "step": 15859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7203872203826904,
+      "learning_rate": 5.508040935672515e-06,
+      "loss": 1.7074,
+      "step": 15860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.557817220687866,
+      "learning_rate": 5.507127192982456e-06,
+      "loss": 1.5721,
+      "step": 15861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.784411072731018,
+      "learning_rate": 5.506213450292398e-06,
+      "loss": 1.0091,
+      "step": 15862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.602539539337158,
+      "learning_rate": 5.50529970760234e-06,
+      "loss": 1.6255,
+      "step": 15863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.86333429813385,
+      "learning_rate": 5.504385964912281e-06,
+      "loss": 1.0892,
+      "step": 15864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.568052053451538,
+      "learning_rate": 5.503472222222223e-06,
+      "loss": 1.6794,
+      "step": 15865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2686374187469482,
+      "learning_rate": 5.502558479532164e-06,
+      "loss": 1.4237,
+      "step": 15866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.409193754196167,
+      "learning_rate": 5.501644736842105e-06,
+      "loss": 1.4192,
+      "step": 15867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.377207040786743,
+      "learning_rate": 5.500730994152047e-06,
+      "loss": 1.4463,
+      "step": 15868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4076120853424072,
+      "learning_rate": 5.499817251461988e-06,
+      "loss": 1.7293,
+      "step": 15869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6552271842956543,
+      "learning_rate": 5.49890350877193e-06,
+      "loss": 1.624,
+      "step": 15870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.408661127090454,
+      "learning_rate": 5.497989766081872e-06,
+      "loss": 1.4781,
+      "step": 15871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2259788513183594,
+      "learning_rate": 5.497076023391813e-06,
+      "loss": 1.1646,
+      "step": 15872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.341714382171631,
+      "learning_rate": 5.4961622807017545e-06,
+      "loss": 1.2525,
+      "step": 15873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.41542387008667,
+      "learning_rate": 5.495248538011696e-06,
+      "loss": 1.1608,
+      "step": 15874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.952476978302002,
+      "learning_rate": 5.494334795321637e-06,
+      "loss": 1.4811,
+      "step": 15875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.339024543762207,
+      "learning_rate": 5.493421052631579e-06,
+      "loss": 1.44,
+      "step": 15876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.467351198196411,
+      "learning_rate": 5.492507309941521e-06,
+      "loss": 1.4542,
+      "step": 15877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.977696657180786,
+      "learning_rate": 5.491593567251462e-06,
+      "loss": 1.4943,
+      "step": 15878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5549533367156982,
+      "learning_rate": 5.4906798245614036e-06,
+      "loss": 1.3983,
+      "step": 15879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.724719762802124,
+      "learning_rate": 5.489766081871345e-06,
+      "loss": 1.5469,
+      "step": 15880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7574572563171387,
+      "learning_rate": 5.488852339181286e-06,
+      "loss": 1.677,
+      "step": 15881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.472299575805664,
+      "learning_rate": 5.487938596491229e-06,
+      "loss": 1.544,
+      "step": 15882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0938196182250977,
+      "learning_rate": 5.487024853801171e-06,
+      "loss": 1.1907,
+      "step": 15883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6478991508483887,
+      "learning_rate": 5.486111111111112e-06,
+      "loss": 1.1534,
+      "step": 15884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.294461965560913,
+      "learning_rate": 5.4851973684210535e-06,
+      "loss": 1.4112,
+      "step": 15885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.607226610183716,
+      "learning_rate": 5.484283625730995e-06,
+      "loss": 1.588,
+      "step": 15886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.69089674949646,
+      "learning_rate": 5.483369883040936e-06,
+      "loss": 1.2183,
+      "step": 15887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7145678997039795,
+      "learning_rate": 5.482456140350878e-06,
+      "loss": 1.557,
+      "step": 15888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.512263059616089,
+      "learning_rate": 5.48154239766082e-06,
+      "loss": 1.6299,
+      "step": 15889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2118496894836426,
+      "learning_rate": 5.480628654970761e-06,
+      "loss": 1.1387,
+      "step": 15890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.579106330871582,
+      "learning_rate": 5.479714912280703e-06,
+      "loss": 1.7485,
+      "step": 15891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8882992267608643,
+      "learning_rate": 5.478801169590644e-06,
+      "loss": 1.1689,
+      "step": 15892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.431527614593506,
+      "learning_rate": 5.477887426900585e-06,
+      "loss": 1.3747,
+      "step": 15893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7715442180633545,
+      "learning_rate": 5.476973684210527e-06,
+      "loss": 1.5183,
+      "step": 15894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.45804500579834,
+      "learning_rate": 5.476059941520468e-06,
+      "loss": 1.4821,
+      "step": 15895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9622914791107178,
+      "learning_rate": 5.47514619883041e-06,
+      "loss": 1.3081,
+      "step": 15896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7754266262054443,
+      "learning_rate": 5.474232456140352e-06,
+      "loss": 1.4952,
+      "step": 15897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5794947147369385,
+      "learning_rate": 5.473318713450293e-06,
+      "loss": 1.6236,
+      "step": 15898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.382936477661133,
+      "learning_rate": 5.4724049707602344e-06,
+      "loss": 1.2679,
+      "step": 15899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.689863443374634,
+      "learning_rate": 5.471491228070176e-06,
+      "loss": 1.7414,
+      "step": 15900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.674297571182251,
+      "learning_rate": 5.470577485380117e-06,
+      "loss": 1.5426,
+      "step": 15901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.632033109664917,
+      "learning_rate": 5.4696637426900586e-06,
+      "loss": 1.6509,
+      "step": 15902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8819847106933594,
+      "learning_rate": 5.468750000000001e-06,
+      "loss": 1.5494,
+      "step": 15903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5917751789093018,
+      "learning_rate": 5.467836257309942e-06,
+      "loss": 1.5317,
+      "step": 15904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7337512969970703,
+      "learning_rate": 5.4669225146198835e-06,
+      "loss": 1.4243,
+      "step": 15905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.272099733352661,
+      "learning_rate": 5.466008771929825e-06,
+      "loss": 1.9901,
+      "step": 15906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.177665710449219,
+      "learning_rate": 5.465095029239766e-06,
+      "loss": 1.7796,
+      "step": 15907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.658811330795288,
+      "learning_rate": 5.464181286549708e-06,
+      "loss": 1.5488,
+      "step": 15908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.528897762298584,
+      "learning_rate": 5.463267543859649e-06,
+      "loss": 1.3819,
+      "step": 15909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3110525608062744,
+      "learning_rate": 5.462353801169591e-06,
+      "loss": 1.2326,
+      "step": 15910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.259394407272339,
+      "learning_rate": 5.461440058479533e-06,
+      "loss": 1.2531,
+      "step": 15911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.328064441680908,
+      "learning_rate": 5.460526315789474e-06,
+      "loss": 1.7829,
+      "step": 15912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5795834064483643,
+      "learning_rate": 5.459612573099415e-06,
+      "loss": 1.1809,
+      "step": 15913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.830775499343872,
+      "learning_rate": 5.458698830409357e-06,
+      "loss": 1.7349,
+      "step": 15914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6446921825408936,
+      "learning_rate": 5.457785087719298e-06,
+      "loss": 1.7833,
+      "step": 15915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.333148241043091,
+      "learning_rate": 5.4568713450292395e-06,
+      "loss": 1.2937,
+      "step": 15916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.342402219772339,
+      "learning_rate": 5.455957602339182e-06,
+      "loss": 1.2787,
+      "step": 15917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1643474102020264,
+      "learning_rate": 5.455043859649123e-06,
+      "loss": 1.3397,
+      "step": 15918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.47499418258667,
+      "learning_rate": 5.4541301169590645e-06,
+      "loss": 1.4138,
+      "step": 15919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.281620502471924,
+      "learning_rate": 5.453216374269006e-06,
+      "loss": 1.5229,
+      "step": 15920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7407114505767822,
+      "learning_rate": 5.452302631578947e-06,
+      "loss": 1.4824,
+      "step": 15921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7970845699310303,
+      "learning_rate": 5.451388888888889e-06,
+      "loss": 1.4877,
+      "step": 15922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.725712537765503,
+      "learning_rate": 5.45047514619883e-06,
+      "loss": 1.6438,
+      "step": 15923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.342972993850708,
+      "learning_rate": 5.449561403508772e-06,
+      "loss": 1.4471,
+      "step": 15924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.703096866607666,
+      "learning_rate": 5.448647660818714e-06,
+      "loss": 1.4748,
+      "step": 15925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.416659355163574,
+      "learning_rate": 5.447733918128656e-06,
+      "loss": 1.3696,
+      "step": 15926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9797396659851074,
+      "learning_rate": 5.446820175438597e-06,
+      "loss": 1.5791,
+      "step": 15927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5303432941436768,
+      "learning_rate": 5.4459064327485385e-06,
+      "loss": 1.6299,
+      "step": 15928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0328681468963623,
+      "learning_rate": 5.444992690058481e-06,
+      "loss": 1.9958,
+      "step": 15929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3274223804473877,
+      "learning_rate": 5.444078947368422e-06,
+      "loss": 1.5322,
+      "step": 15930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.555950164794922,
+      "learning_rate": 5.4431652046783635e-06,
+      "loss": 1.5274,
+      "step": 15931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.542229652404785,
+      "learning_rate": 5.442251461988305e-06,
+      "loss": 1.379,
+      "step": 15932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.872328519821167,
+      "learning_rate": 5.441337719298246e-06,
+      "loss": 1.5449,
+      "step": 15933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.264110565185547,
+      "learning_rate": 5.440423976608188e-06,
+      "loss": 1.2674,
+      "step": 15934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.55350399017334,
+      "learning_rate": 5.439510233918129e-06,
+      "loss": 1.5282,
+      "step": 15935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.984492301940918,
+      "learning_rate": 5.438596491228071e-06,
+      "loss": 1.5689,
+      "step": 15936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2424697875976562,
+      "learning_rate": 5.437682748538013e-06,
+      "loss": 1.1277,
+      "step": 15937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.480011224746704,
+      "learning_rate": 5.436769005847954e-06,
+      "loss": 1.6455,
+      "step": 15938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6281540393829346,
+      "learning_rate": 5.435855263157895e-06,
+      "loss": 1.4912,
+      "step": 15939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2311177253723145,
+      "learning_rate": 5.434941520467837e-06,
+      "loss": 1.9243,
+      "step": 15940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7537436485290527,
+      "learning_rate": 5.434027777777778e-06,
+      "loss": 2.0738,
+      "step": 15941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.65694260597229,
+      "learning_rate": 5.4331140350877195e-06,
+      "loss": 1.4796,
+      "step": 15942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.837914228439331,
+      "learning_rate": 5.432200292397662e-06,
+      "loss": 1.7401,
+      "step": 15943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.487805128097534,
+      "learning_rate": 5.431286549707603e-06,
+      "loss": 1.403,
+      "step": 15944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.535301923751831,
+      "learning_rate": 5.4303728070175444e-06,
+      "loss": 1.4631,
+      "step": 15945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6798436641693115,
+      "learning_rate": 5.429459064327486e-06,
+      "loss": 1.5435,
+      "step": 15946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.995457649230957,
+      "learning_rate": 5.428545321637427e-06,
+      "loss": 1.5937,
+      "step": 15947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.306786298751831,
+      "learning_rate": 5.4276315789473686e-06,
+      "loss": 1.322,
+      "step": 15948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.656512975692749,
+      "learning_rate": 5.42671783625731e-06,
+      "loss": 1.6339,
+      "step": 15949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.533719778060913,
+      "learning_rate": 5.425804093567252e-06,
+      "loss": 1.5884,
+      "step": 15950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0218071937561035,
+      "learning_rate": 5.4248903508771935e-06,
+      "loss": 1.6071,
+      "step": 15951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.133986234664917,
+      "learning_rate": 5.423976608187135e-06,
+      "loss": 1.1723,
+      "step": 15952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8653206825256348,
+      "learning_rate": 5.423062865497076e-06,
+      "loss": 1.5237,
+      "step": 15953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.662518262863159,
+      "learning_rate": 5.422149122807018e-06,
+      "loss": 1.5855,
+      "step": 15954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.184340000152588,
+      "learning_rate": 5.421235380116959e-06,
+      "loss": 1.4267,
+      "step": 15955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5018012523651123,
+      "learning_rate": 5.4203216374269e-06,
+      "loss": 1.3542,
+      "step": 15956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.306126356124878,
+      "learning_rate": 5.419407894736843e-06,
+      "loss": 1.3149,
+      "step": 15957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.371913433074951,
+      "learning_rate": 5.418494152046784e-06,
+      "loss": 1.4445,
+      "step": 15958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.504743814468384,
+      "learning_rate": 5.417580409356725e-06,
+      "loss": 1.5001,
+      "step": 15959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.897582530975342,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 1.9244,
+      "step": 15960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1326310634613037,
+      "learning_rate": 5.415752923976608e-06,
+      "loss": 1.411,
+      "step": 15961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.589900493621826,
+      "learning_rate": 5.4148391812865495e-06,
+      "loss": 1.1226,
+      "step": 15962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.394958257675171,
+      "learning_rate": 5.413925438596491e-06,
+      "loss": 1.4609,
+      "step": 15963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.612534999847412,
+      "learning_rate": 5.413011695906433e-06,
+      "loss": 1.5458,
+      "step": 15964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.419588804244995,
+      "learning_rate": 5.4120979532163745e-06,
+      "loss": 1.4563,
+      "step": 15965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8356285095214844,
+      "learning_rate": 5.411184210526316e-06,
+      "loss": 1.5896,
+      "step": 15966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9094855785369873,
+      "learning_rate": 5.410270467836257e-06,
+      "loss": 1.3097,
+      "step": 15967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.657872438430786,
+      "learning_rate": 5.4093567251461994e-06,
+      "loss": 1.6526,
+      "step": 15968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5271334648132324,
+      "learning_rate": 5.408442982456142e-06,
+      "loss": 1.5261,
+      "step": 15969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.996504545211792,
+      "learning_rate": 5.407529239766083e-06,
+      "loss": 1.1604,
+      "step": 15970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7840263843536377,
+      "learning_rate": 5.406615497076024e-06,
+      "loss": 1.3716,
+      "step": 15971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1767780780792236,
+      "learning_rate": 5.405701754385966e-06,
+      "loss": 1.1642,
+      "step": 15972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6576685905456543,
+      "learning_rate": 5.404788011695907e-06,
+      "loss": 1.6365,
+      "step": 15973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2618870735168457,
+      "learning_rate": 5.4038742690058485e-06,
+      "loss": 1.4614,
+      "step": 15974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.169900417327881,
+      "learning_rate": 5.40296052631579e-06,
+      "loss": 1.6004,
+      "step": 15975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4913713932037354,
+      "learning_rate": 5.402046783625732e-06,
+      "loss": 1.5995,
+      "step": 15976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.128549575805664,
+      "learning_rate": 5.4011330409356735e-06,
+      "loss": 1.7983,
+      "step": 15977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.700284957885742,
+      "learning_rate": 5.400219298245615e-06,
+      "loss": 1.5944,
+      "step": 15978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3983869552612305,
+      "learning_rate": 5.399305555555556e-06,
+      "loss": 1.3665,
+      "step": 15979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0802557468414307,
+      "learning_rate": 5.398391812865498e-06,
+      "loss": 1.1161,
+      "step": 15980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.534891366958618,
+      "learning_rate": 5.397478070175439e-06,
+      "loss": 1.7164,
+      "step": 15981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4176392555236816,
+      "learning_rate": 5.39656432748538e-06,
+      "loss": 1.3555,
+      "step": 15982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2917778491973877,
+      "learning_rate": 5.395650584795323e-06,
+      "loss": 1.583,
+      "step": 15983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2224879264831543,
+      "learning_rate": 5.394736842105264e-06,
+      "loss": 1.445,
+      "step": 15984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.458019256591797,
+      "learning_rate": 5.393823099415205e-06,
+      "loss": 1.2849,
+      "step": 15985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0220301151275635,
+      "learning_rate": 5.392909356725147e-06,
+      "loss": 1.7424,
+      "step": 15986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.680692672729492,
+      "learning_rate": 5.391995614035088e-06,
+      "loss": 1.4286,
+      "step": 15987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.762026786804199,
+      "learning_rate": 5.3910818713450295e-06,
+      "loss": 1.5196,
+      "step": 15988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9315550327301025,
+      "learning_rate": 5.390168128654971e-06,
+      "loss": 1.8904,
+      "step": 15989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.92832088470459,
+      "learning_rate": 5.389254385964913e-06,
+      "loss": 1.5641,
+      "step": 15990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3072147369384766,
+      "learning_rate": 5.3883406432748544e-06,
+      "loss": 1.2641,
+      "step": 15991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8375730514526367,
+      "learning_rate": 5.387426900584796e-06,
+      "loss": 1.5762,
+      "step": 15992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.057147741317749,
+      "learning_rate": 5.386513157894737e-06,
+      "loss": 1.2128,
+      "step": 15993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.267321825027466,
+      "learning_rate": 5.3855994152046786e-06,
+      "loss": 1.4192,
+      "step": 15994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9203438758850098,
+      "learning_rate": 5.38468567251462e-06,
+      "loss": 1.4632,
+      "step": 15995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.885506510734558,
+      "learning_rate": 5.383771929824561e-06,
+      "loss": 1.2716,
+      "step": 15996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.494882583618164,
+      "learning_rate": 5.3828581871345035e-06,
+      "loss": 1.4756,
+      "step": 15997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.837319016456604,
+      "learning_rate": 5.381944444444445e-06,
+      "loss": 1.1485,
+      "step": 15998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7552683353424072,
+      "learning_rate": 5.381030701754386e-06,
+      "loss": 1.3107,
+      "step": 15999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.412094831466675,
+      "learning_rate": 5.380116959064328e-06,
+      "loss": 1.2641,
+      "step": 16000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4018666744232178,
+      "learning_rate": 5.379203216374269e-06,
+      "loss": 1.5069,
+      "step": 16001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9261929988861084,
+      "learning_rate": 5.37828947368421e-06,
+      "loss": 1.584,
+      "step": 16002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.701504945755005,
+      "learning_rate": 5.377375730994152e-06,
+      "loss": 1.1458,
+      "step": 16003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.741910219192505,
+      "learning_rate": 5.376461988304094e-06,
+      "loss": 1.5476,
+      "step": 16004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.309699296951294,
+      "learning_rate": 5.375548245614035e-06,
+      "loss": 1.3013,
+      "step": 16005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.741580009460449,
+      "learning_rate": 5.374634502923977e-06,
+      "loss": 1.4039,
+      "step": 16006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4366447925567627,
+      "learning_rate": 5.373720760233918e-06,
+      "loss": 1.5466,
+      "step": 16007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2765283584594727,
+      "learning_rate": 5.3728070175438595e-06,
+      "loss": 1.2875,
+      "step": 16008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.444098949432373,
+      "learning_rate": 5.371893274853801e-06,
+      "loss": 1.3191,
+      "step": 16009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.199779987335205,
+      "learning_rate": 5.370979532163744e-06,
+      "loss": 1.4147,
+      "step": 16010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4849047660827637,
+      "learning_rate": 5.370065789473685e-06,
+      "loss": 1.5531,
+      "step": 16011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1979126930236816,
+      "learning_rate": 5.369152046783627e-06,
+      "loss": 1.3411,
+      "step": 16012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2669789791107178,
+      "learning_rate": 5.368238304093568e-06,
+      "loss": 1.0558,
+      "step": 16013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1131911277770996,
+      "learning_rate": 5.3673245614035094e-06,
+      "loss": 1.0186,
+      "step": 16014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7521486282348633,
+      "learning_rate": 5.366410818713451e-06,
+      "loss": 1.5358,
+      "step": 16015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9743566513061523,
+      "learning_rate": 5.365497076023393e-06,
+      "loss": 1.6012,
+      "step": 16016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.778269052505493,
+      "learning_rate": 5.364583333333334e-06,
+      "loss": 1.5791,
+      "step": 16017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4173741340637207,
+      "learning_rate": 5.363669590643276e-06,
+      "loss": 1.5343,
+      "step": 16018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.539121150970459,
+      "learning_rate": 5.362755847953217e-06,
+      "loss": 1.477,
+      "step": 16019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4083831310272217,
+      "learning_rate": 5.3618421052631585e-06,
+      "loss": 1.3907,
+      "step": 16020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.299023389816284,
+      "learning_rate": 5.3609283625731e-06,
+      "loss": 1.2569,
+      "step": 16021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.630147933959961,
+      "learning_rate": 5.360014619883041e-06,
+      "loss": 1.5477,
+      "step": 16022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4499704837799072,
+      "learning_rate": 5.3591008771929835e-06,
+      "loss": 1.5615,
+      "step": 16023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.630798578262329,
+      "learning_rate": 5.358187134502925e-06,
+      "loss": 1.405,
+      "step": 16024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8604681491851807,
+      "learning_rate": 5.357273391812866e-06,
+      "loss": 1.7335,
+      "step": 16025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6330630779266357,
+      "learning_rate": 5.356359649122808e-06,
+      "loss": 1.4502,
+      "step": 16026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6993470191955566,
+      "learning_rate": 5.355445906432749e-06,
+      "loss": 1.4312,
+      "step": 16027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5160765647888184,
+      "learning_rate": 5.35453216374269e-06,
+      "loss": 1.3223,
+      "step": 16028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5492944717407227,
+      "learning_rate": 5.353618421052632e-06,
+      "loss": 1.3629,
+      "step": 16029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9640963077545166,
+      "learning_rate": 5.352704678362574e-06,
+      "loss": 1.4586,
+      "step": 16030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.397465229034424,
+      "learning_rate": 5.351790935672515e-06,
+      "loss": 1.4708,
+      "step": 16031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7351176738739014,
+      "learning_rate": 5.350877192982457e-06,
+      "loss": 1.6135,
+      "step": 16032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4046854972839355,
+      "learning_rate": 5.349963450292398e-06,
+      "loss": 1.4532,
+      "step": 16033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.491877317428589,
+      "learning_rate": 5.3490497076023395e-06,
+      "loss": 1.5292,
+      "step": 16034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2552826404571533,
+      "learning_rate": 5.348135964912281e-06,
+      "loss": 1.4142,
+      "step": 16035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.562457799911499,
+      "learning_rate": 5.347222222222222e-06,
+      "loss": 1.7649,
+      "step": 16036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20877742767334,
+      "learning_rate": 5.3463084795321644e-06,
+      "loss": 1.312,
+      "step": 16037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0561647415161133,
+      "learning_rate": 5.345394736842106e-06,
+      "loss": 1.1452,
+      "step": 16038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7801473140716553,
+      "learning_rate": 5.344480994152047e-06,
+      "loss": 1.7196,
+      "step": 16039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.130007028579712,
+      "learning_rate": 5.3435672514619886e-06,
+      "loss": 2.0395,
+      "step": 16040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8562484979629517,
+      "learning_rate": 5.34265350877193e-06,
+      "loss": 1.2701,
+      "step": 16041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.169569492340088,
+      "learning_rate": 5.341739766081871e-06,
+      "loss": 1.1955,
+      "step": 16042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7648138999938965,
+      "learning_rate": 5.340826023391813e-06,
+      "loss": 1.3933,
+      "step": 16043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3267765045166016,
+      "learning_rate": 5.339912280701755e-06,
+      "loss": 1.2611,
+      "step": 16044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.447901487350464,
+      "learning_rate": 5.338998538011696e-06,
+      "loss": 1.4406,
+      "step": 16045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7377400398254395,
+      "learning_rate": 5.338084795321638e-06,
+      "loss": 1.4011,
+      "step": 16046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5164270401000977,
+      "learning_rate": 5.337171052631579e-06,
+      "loss": 1.4496,
+      "step": 16047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.428051710128784,
+      "learning_rate": 5.33625730994152e-06,
+      "loss": 1.1817,
+      "step": 16048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3536906242370605,
+      "learning_rate": 5.335343567251462e-06,
+      "loss": 1.4599,
+      "step": 16049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.319549798965454,
+      "learning_rate": 5.334429824561403e-06,
+      "loss": 1.4995,
+      "step": 16050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.219261646270752,
+      "learning_rate": 5.3335160818713445e-06,
+      "loss": 1.1896,
+      "step": 16051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.130364179611206,
+      "learning_rate": 5.332602339181287e-06,
+      "loss": 1.0755,
+      "step": 16052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6736366748809814,
+      "learning_rate": 5.331688596491229e-06,
+      "loss": 1.4768,
+      "step": 16053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.029771566390991,
+      "learning_rate": 5.33077485380117e-06,
+      "loss": 1.2648,
+      "step": 16054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3098809719085693,
+      "learning_rate": 5.329861111111112e-06,
+      "loss": 1.3125,
+      "step": 16055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4850122928619385,
+      "learning_rate": 5.328947368421054e-06,
+      "loss": 1.499,
+      "step": 16056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1058645248413086,
+      "learning_rate": 5.328033625730995e-06,
+      "loss": 1.4589,
+      "step": 16057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.454929828643799,
+      "learning_rate": 5.327119883040937e-06,
+      "loss": 1.4337,
+      "step": 16058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4698376655578613,
+      "learning_rate": 5.326206140350878e-06,
+      "loss": 1.475,
+      "step": 16059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.728057622909546,
+      "learning_rate": 5.3252923976608194e-06,
+      "loss": 1.4841,
+      "step": 16060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6853692531585693,
+      "learning_rate": 5.324378654970761e-06,
+      "loss": 1.2857,
+      "step": 16061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7297627925872803,
+      "learning_rate": 5.323464912280702e-06,
+      "loss": 1.5433,
+      "step": 16062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6293022632598877,
+      "learning_rate": 5.322551169590644e-06,
+      "loss": 1.506,
+      "step": 16063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.543456554412842,
+      "learning_rate": 5.321637426900586e-06,
+      "loss": 1.4671,
+      "step": 16064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5477685928344727,
+      "learning_rate": 5.320723684210527e-06,
+      "loss": 1.6555,
+      "step": 16065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5697572231292725,
+      "learning_rate": 5.3198099415204685e-06,
+      "loss": 1.4613,
+      "step": 16066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.426861047744751,
+      "learning_rate": 5.31889619883041e-06,
+      "loss": 1.4947,
+      "step": 16067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3404855728149414,
+      "learning_rate": 5.317982456140351e-06,
+      "loss": 1.3318,
+      "step": 16068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.005168914794922,
+      "learning_rate": 5.317068713450293e-06,
+      "loss": 1.6425,
+      "step": 16069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6233482360839844,
+      "learning_rate": 5.316154970760234e-06,
+      "loss": 1.6127,
+      "step": 16070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0731048583984375,
+      "learning_rate": 5.315241228070176e-06,
+      "loss": 1.8515,
+      "step": 16071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0481905937194824,
+      "learning_rate": 5.314327485380118e-06,
+      "loss": 1.0837,
+      "step": 16072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5400776863098145,
+      "learning_rate": 5.313413742690059e-06,
+      "loss": 1.4602,
+      "step": 16073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3124241828918457,
+      "learning_rate": 5.3125e-06,
+      "loss": 1.4486,
+      "step": 16074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1619675159454346,
+      "learning_rate": 5.311586257309942e-06,
+      "loss": 1.391,
+      "step": 16075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7863314151763916,
+      "learning_rate": 5.310672514619883e-06,
+      "loss": 1.432,
+      "step": 16076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.506922960281372,
+      "learning_rate": 5.3097587719298245e-06,
+      "loss": 1.3157,
+      "step": 16077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.378575086593628,
+      "learning_rate": 5.308845029239767e-06,
+      "loss": 1.2684,
+      "step": 16078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.400219202041626,
+      "learning_rate": 5.307931286549708e-06,
+      "loss": 1.9026,
+      "step": 16079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.794416666030884,
+      "learning_rate": 5.3070175438596495e-06,
+      "loss": 1.5631,
+      "step": 16080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9247305393218994,
+      "learning_rate": 5.306103801169591e-06,
+      "loss": 1.5354,
+      "step": 16081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.388465166091919,
+      "learning_rate": 5.305190058479532e-06,
+      "loss": 1.1534,
+      "step": 16082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.315610408782959,
+      "learning_rate": 5.304276315789474e-06,
+      "loss": 1.0455,
+      "step": 16083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4958183765411377,
+      "learning_rate": 5.303362573099415e-06,
+      "loss": 1.0835,
+      "step": 16084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9028866291046143,
+      "learning_rate": 5.302448830409357e-06,
+      "loss": 1.6894,
+      "step": 16085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5842509269714355,
+      "learning_rate": 5.3015350877192986e-06,
+      "loss": 1.4967,
+      "step": 16086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.464951515197754,
+      "learning_rate": 5.30062134502924e-06,
+      "loss": 1.1993,
+      "step": 16087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9495925903320312,
+      "learning_rate": 5.299707602339181e-06,
+      "loss": 1.4921,
+      "step": 16088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.059636354446411,
+      "learning_rate": 5.298793859649123e-06,
+      "loss": 1.6001,
+      "step": 16089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0721514225006104,
+      "learning_rate": 5.297880116959064e-06,
+      "loss": 1.4292,
+      "step": 16090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8224642276763916,
+      "learning_rate": 5.2969663742690054e-06,
+      "loss": 1.7864,
+      "step": 16091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.517521381378174,
+      "learning_rate": 5.296052631578948e-06,
+      "loss": 1.3024,
+      "step": 16092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5857605934143066,
+      "learning_rate": 5.295138888888889e-06,
+      "loss": 1.583,
+      "step": 16093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2419803142547607,
+      "learning_rate": 5.29422514619883e-06,
+      "loss": 1.3067,
+      "step": 16094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.555152177810669,
+      "learning_rate": 5.293311403508772e-06,
+      "loss": 1.3804,
+      "step": 16095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.791365385055542,
+      "learning_rate": 5.292397660818714e-06,
+      "loss": 1.5907,
+      "step": 16096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5030102729797363,
+      "learning_rate": 5.291483918128656e-06,
+      "loss": 1.3346,
+      "step": 16097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.307561159133911,
+      "learning_rate": 5.290570175438598e-06,
+      "loss": 1.1676,
+      "step": 16098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2667713165283203,
+      "learning_rate": 5.289656432748539e-06,
+      "loss": 1.6391,
+      "step": 16099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.503589153289795,
+      "learning_rate": 5.28874269005848e-06,
+      "loss": 1.3936,
+      "step": 16100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0352962017059326,
+      "learning_rate": 5.287828947368422e-06,
+      "loss": 1.1554,
+      "step": 16101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8916736841201782,
+      "learning_rate": 5.286915204678363e-06,
+      "loss": 1.1222,
+      "step": 16102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.16628098487854,
+      "learning_rate": 5.2860014619883045e-06,
+      "loss": 1.4664,
+      "step": 16103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.185819149017334,
+      "learning_rate": 5.285087719298247e-06,
+      "loss": 1.2707,
+      "step": 16104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5740861892700195,
+      "learning_rate": 5.284173976608188e-06,
+      "loss": 1.6491,
+      "step": 16105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6595287322998047,
+      "learning_rate": 5.2832602339181294e-06,
+      "loss": 1.5058,
+      "step": 16106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6235814094543457,
+      "learning_rate": 5.282346491228071e-06,
+      "loss": 1.0416,
+      "step": 16107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6454200744628906,
+      "learning_rate": 5.281432748538012e-06,
+      "loss": 1.4446,
+      "step": 16108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.630995035171509,
+      "learning_rate": 5.2805190058479536e-06,
+      "loss": 1.5894,
+      "step": 16109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.742403507232666,
+      "learning_rate": 5.279605263157895e-06,
+      "loss": 1.8085,
+      "step": 16110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.571657419204712,
+      "learning_rate": 5.278691520467837e-06,
+      "loss": 1.7115,
+      "step": 16111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9290426969528198,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 1.0603,
+      "step": 16112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.529726982116699,
+      "learning_rate": 5.27686403508772e-06,
+      "loss": 1.5066,
+      "step": 16113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4392168521881104,
+      "learning_rate": 5.275950292397661e-06,
+      "loss": 1.4262,
+      "step": 16114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6488444805145264,
+      "learning_rate": 5.275036549707603e-06,
+      "loss": 1.5123,
+      "step": 16115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6110968589782715,
+      "learning_rate": 5.274122807017544e-06,
+      "loss": 1.602,
+      "step": 16116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.798835039138794,
+      "learning_rate": 5.273209064327485e-06,
+      "loss": 1.7326,
+      "step": 16117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2113327980041504,
+      "learning_rate": 5.272295321637428e-06,
+      "loss": 1.2348,
+      "step": 16118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3769469261169434,
+      "learning_rate": 5.271381578947369e-06,
+      "loss": 1.5211,
+      "step": 16119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2868540287017822,
+      "learning_rate": 5.27046783625731e-06,
+      "loss": 1.2861,
+      "step": 16120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.690528631210327,
+      "learning_rate": 5.269554093567252e-06,
+      "loss": 1.4242,
+      "step": 16121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.722726583480835,
+      "learning_rate": 5.268640350877193e-06,
+      "loss": 1.9543,
+      "step": 16122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.408142328262329,
+      "learning_rate": 5.2677266081871345e-06,
+      "loss": 1.234,
+      "step": 16123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8084094524383545,
+      "learning_rate": 5.266812865497076e-06,
+      "loss": 1.375,
+      "step": 16124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.181046962738037,
+      "learning_rate": 5.265899122807018e-06,
+      "loss": 1.3623,
+      "step": 16125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4085988998413086,
+      "learning_rate": 5.2649853801169595e-06,
+      "loss": 1.6131,
+      "step": 16126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9233739376068115,
+      "learning_rate": 5.264071637426901e-06,
+      "loss": 1.0791,
+      "step": 16127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3730099201202393,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 1.3726,
+      "step": 16128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1151254177093506,
+      "learning_rate": 5.262244152046784e-06,
+      "loss": 1.3511,
+      "step": 16129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.15775990486145,
+      "learning_rate": 5.261330409356725e-06,
+      "loss": 1.578,
+      "step": 16130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.568882703781128,
+      "learning_rate": 5.260416666666666e-06,
+      "loss": 1.1685,
+      "step": 16131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.649386405944824,
+      "learning_rate": 5.2595029239766086e-06,
+      "loss": 1.4857,
+      "step": 16132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4176878929138184,
+      "learning_rate": 5.25858918128655e-06,
+      "loss": 1.534,
+      "step": 16133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9934556484222412,
+      "learning_rate": 5.257675438596491e-06,
+      "loss": 1.002,
+      "step": 16134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4383151531219482,
+      "learning_rate": 5.256761695906433e-06,
+      "loss": 1.2575,
+      "step": 16135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.026698589324951,
+      "learning_rate": 5.255847953216374e-06,
+      "loss": 1.1453,
+      "step": 16136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3398303985595703,
+      "learning_rate": 5.2549342105263154e-06,
+      "loss": 1.4094,
+      "step": 16137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.121767997741699,
+      "learning_rate": 5.254020467836257e-06,
+      "loss": 1.1677,
+      "step": 16138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6043195724487305,
+      "learning_rate": 5.2531067251462e-06,
+      "loss": 1.3459,
+      "step": 16139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.606773614883423,
+      "learning_rate": 5.252192982456141e-06,
+      "loss": 1.2231,
+      "step": 16140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.875680923461914,
+      "learning_rate": 5.251279239766083e-06,
+      "loss": 1.1657,
+      "step": 16141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7463502883911133,
+      "learning_rate": 5.250365497076024e-06,
+      "loss": 1.3663,
+      "step": 16142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2529261112213135,
+      "learning_rate": 5.249451754385965e-06,
+      "loss": 1.063,
+      "step": 16143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4110312461853027,
+      "learning_rate": 5.248538011695908e-06,
+      "loss": 1.5649,
+      "step": 16144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3568594455718994,
+      "learning_rate": 5.247624269005849e-06,
+      "loss": 1.2979,
+      "step": 16145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.504958152770996,
+      "learning_rate": 5.24671052631579e-06,
+      "loss": 1.5448,
+      "step": 16146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3765709400177,
+      "learning_rate": 5.245796783625732e-06,
+      "loss": 1.4161,
+      "step": 16147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4641306400299072,
+      "learning_rate": 5.244883040935673e-06,
+      "loss": 1.5013,
+      "step": 16148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.350111722946167,
+      "learning_rate": 5.2439692982456145e-06,
+      "loss": 1.4389,
+      "step": 16149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.939088821411133,
+      "learning_rate": 5.243055555555556e-06,
+      "loss": 1.7163,
+      "step": 16150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.344275712966919,
+      "learning_rate": 5.242141812865498e-06,
+      "loss": 1.5501,
+      "step": 16151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2766330242156982,
+      "learning_rate": 5.2412280701754394e-06,
+      "loss": 2.1107,
+      "step": 16152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4257211685180664,
+      "learning_rate": 5.240314327485381e-06,
+      "loss": 1.9788,
+      "step": 16153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.556574583053589,
+      "learning_rate": 5.239400584795322e-06,
+      "loss": 1.6036,
+      "step": 16154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7731332778930664,
+      "learning_rate": 5.2384868421052636e-06,
+      "loss": 1.4983,
+      "step": 16155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.27228045463562,
+      "learning_rate": 5.237573099415205e-06,
+      "loss": 1.3088,
+      "step": 16156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9222488403320312,
+      "learning_rate": 5.236659356725146e-06,
+      "loss": 1.3823,
+      "step": 16157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3188445568084717,
+      "learning_rate": 5.2357456140350885e-06,
+      "loss": 1.4287,
+      "step": 16158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.43243145942688,
+      "learning_rate": 5.23483187134503e-06,
+      "loss": 1.2882,
+      "step": 16159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0213050842285156,
+      "learning_rate": 5.233918128654971e-06,
+      "loss": 1.1216,
+      "step": 16160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.776158571243286,
+      "learning_rate": 5.233004385964913e-06,
+      "loss": 1.5623,
+      "step": 16161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5478999614715576,
+      "learning_rate": 5.232090643274854e-06,
+      "loss": 1.2582,
+      "step": 16162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.848426103591919,
+      "learning_rate": 5.231176900584795e-06,
+      "loss": 1.7047,
+      "step": 16163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.81776762008667,
+      "learning_rate": 5.230263157894737e-06,
+      "loss": 1.3436,
+      "step": 16164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5439040660858154,
+      "learning_rate": 5.229349415204679e-06,
+      "loss": 1.3059,
+      "step": 16165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1277546882629395,
+      "learning_rate": 5.22843567251462e-06,
+      "loss": 1.5887,
+      "step": 16166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6492178440093994,
+      "learning_rate": 5.227521929824562e-06,
+      "loss": 1.422,
+      "step": 16167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.828064203262329,
+      "learning_rate": 5.226608187134503e-06,
+      "loss": 1.7262,
+      "step": 16168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1867825984954834,
+      "learning_rate": 5.2256944444444445e-06,
+      "loss": 1.343,
+      "step": 16169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.506399631500244,
+      "learning_rate": 5.224780701754386e-06,
+      "loss": 1.5552,
+      "step": 16170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.285504102706909,
+      "learning_rate": 5.223866959064327e-06,
+      "loss": 1.3613,
+      "step": 16171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1503427028656006,
+      "learning_rate": 5.2229532163742695e-06,
+      "loss": 1.1744,
+      "step": 16172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.573789596557617,
+      "learning_rate": 5.222039473684211e-06,
+      "loss": 1.5518,
+      "step": 16173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.904866933822632,
+      "learning_rate": 5.221125730994152e-06,
+      "loss": 1.3855,
+      "step": 16174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3675873279571533,
+      "learning_rate": 5.220211988304094e-06,
+      "loss": 1.2585,
+      "step": 16175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.667105197906494,
+      "learning_rate": 5.219298245614035e-06,
+      "loss": 1.3511,
+      "step": 16176
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.162579298019409,
+      "learning_rate": 5.218384502923976e-06,
+      "loss": 1.2405,
+      "step": 16177
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.447150707244873,
+      "learning_rate": 5.217470760233918e-06,
+      "loss": 1.423,
+      "step": 16178
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9413734674453735,
+      "learning_rate": 5.21655701754386e-06,
+      "loss": 1.1864,
+      "step": 16179
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.276740074157715,
+      "learning_rate": 5.215643274853801e-06,
+      "loss": 1.361,
+      "step": 16180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6252048015594482,
+      "learning_rate": 5.2147295321637435e-06,
+      "loss": 1.5154,
+      "step": 16181
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.727689504623413,
+      "learning_rate": 5.213815789473685e-06,
+      "loss": 1.7091,
+      "step": 16182
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.483057975769043,
+      "learning_rate": 5.212902046783626e-06,
+      "loss": 1.6881,
+      "step": 16183
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3906686305999756,
+      "learning_rate": 5.2119883040935685e-06,
+      "loss": 1.2593,
+      "step": 16184
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.286736488342285,
+      "learning_rate": 5.21107456140351e-06,
+      "loss": 1.4682,
+      "step": 16185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.299689292907715,
+      "learning_rate": 5.210160818713451e-06,
+      "loss": 1.5028,
+      "step": 16186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2461256980895996,
+      "learning_rate": 5.209247076023393e-06,
+      "loss": 1.7026,
+      "step": 16187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.486036777496338,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 1.3999,
+      "step": 16188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.169844627380371,
+      "learning_rate": 5.207419590643275e-06,
+      "loss": 1.462,
+      "step": 16189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.444485902786255,
+      "learning_rate": 5.206505847953217e-06,
+      "loss": 1.4903,
+      "step": 16190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.108565092086792,
+      "learning_rate": 5.205592105263159e-06,
+      "loss": 1.7133,
+      "step": 16191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.516580581665039,
+      "learning_rate": 5.2046783625731e-06,
+      "loss": 1.3339,
+      "step": 16192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.133589267730713,
+      "learning_rate": 5.203764619883042e-06,
+      "loss": 1.343,
+      "step": 16193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.160067558288574,
+      "learning_rate": 5.202850877192983e-06,
+      "loss": 1.1118,
+      "step": 16194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7769627571105957,
+      "learning_rate": 5.2019371345029245e-06,
+      "loss": 1.4543,
+      "step": 16195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1633856296539307,
+      "learning_rate": 5.201023391812866e-06,
+      "loss": 1.2118,
+      "step": 16196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5618362426757812,
+      "learning_rate": 5.200109649122807e-06,
+      "loss": 1.4159,
+      "step": 16197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3040385246276855,
+      "learning_rate": 5.1991959064327494e-06,
+      "loss": 1.2035,
+      "step": 16198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1922550201416016,
+      "learning_rate": 5.198282163742691e-06,
+      "loss": 1.215,
+      "step": 16199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.85163950920105,
+      "learning_rate": 5.197368421052632e-06,
+      "loss": 1.5421,
+      "step": 16200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.684840440750122,
+      "learning_rate": 5.1964546783625736e-06,
+      "loss": 1.5089,
+      "step": 16201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3637993335723877,
+      "learning_rate": 5.195540935672515e-06,
+      "loss": 1.1983,
+      "step": 16202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.600173234939575,
+      "learning_rate": 5.194627192982456e-06,
+      "loss": 1.3994,
+      "step": 16203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4008185863494873,
+      "learning_rate": 5.193713450292398e-06,
+      "loss": 1.7126,
+      "step": 16204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6123220920562744,
+      "learning_rate": 5.19279970760234e-06,
+      "loss": 1.6619,
+      "step": 16205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6698696613311768,
+      "learning_rate": 5.191885964912281e-06,
+      "loss": 1.5812,
+      "step": 16206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3344852924346924,
+      "learning_rate": 5.190972222222223e-06,
+      "loss": 1.1951,
+      "step": 16207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.396608829498291,
+      "learning_rate": 5.190058479532164e-06,
+      "loss": 1.3791,
+      "step": 16208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4071969985961914,
+      "learning_rate": 5.189144736842105e-06,
+      "loss": 1.3235,
+      "step": 16209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8457987308502197,
+      "learning_rate": 5.188230994152047e-06,
+      "loss": 1.7939,
+      "step": 16210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.844026803970337,
+      "learning_rate": 5.187317251461988e-06,
+      "loss": 2.045,
+      "step": 16211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0163052082061768,
+      "learning_rate": 5.18640350877193e-06,
+      "loss": 2.0022,
+      "step": 16212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.235689163208008,
+      "learning_rate": 5.185489766081872e-06,
+      "loss": 1.3608,
+      "step": 16213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.241421937942505,
+      "learning_rate": 5.184576023391813e-06,
+      "loss": 1.3927,
+      "step": 16214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.63794207572937,
+      "learning_rate": 5.1836622807017545e-06,
+      "loss": 1.7517,
+      "step": 16215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0972957611083984,
+      "learning_rate": 5.182748538011696e-06,
+      "loss": 1.6716,
+      "step": 16216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1969261169433594,
+      "learning_rate": 5.181834795321637e-06,
+      "loss": 1.2308,
+      "step": 16217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4561595916748047,
+      "learning_rate": 5.180921052631579e-06,
+      "loss": 1.4839,
+      "step": 16218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7125799655914307,
+      "learning_rate": 5.180007309941521e-06,
+      "loss": 1.8852,
+      "step": 16219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7612786293029785,
+      "learning_rate": 5.179093567251462e-06,
+      "loss": 1.7101,
+      "step": 16220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4830095767974854,
+      "learning_rate": 5.178179824561404e-06,
+      "loss": 1.4285,
+      "step": 16221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5001795291900635,
+      "learning_rate": 5.177266081871345e-06,
+      "loss": 1.333,
+      "step": 16222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.496953010559082,
+      "learning_rate": 5.176352339181286e-06,
+      "loss": 1.6824,
+      "step": 16223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1177573204040527,
+      "learning_rate": 5.175438596491229e-06,
+      "loss": 1.7347,
+      "step": 16224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1334309577941895,
+      "learning_rate": 5.174524853801171e-06,
+      "loss": 1.3893,
+      "step": 16225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0599377155303955,
+      "learning_rate": 5.173611111111112e-06,
+      "loss": 1.5556,
+      "step": 16226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3962972164154053,
+      "learning_rate": 5.1726973684210535e-06,
+      "loss": 1.4909,
+      "step": 16227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2320868968963623,
+      "learning_rate": 5.171783625730995e-06,
+      "loss": 1.1759,
+      "step": 16228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.618708848953247,
+      "learning_rate": 5.170869883040936e-06,
+      "loss": 1.5842,
+      "step": 16229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2255172729492188,
+      "learning_rate": 5.169956140350878e-06,
+      "loss": 1.2584,
+      "step": 16230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.307053327560425,
+      "learning_rate": 5.16904239766082e-06,
+      "loss": 1.4944,
+      "step": 16231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9732611179351807,
+      "learning_rate": 5.168128654970761e-06,
+      "loss": 1.4899,
+      "step": 16232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.410398006439209,
+      "learning_rate": 5.167214912280703e-06,
+      "loss": 1.4952,
+      "step": 16233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.148843288421631,
+      "learning_rate": 5.166301169590644e-06,
+      "loss": 1.2058,
+      "step": 16234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.647249221801758,
+      "learning_rate": 5.165387426900585e-06,
+      "loss": 1.2111,
+      "step": 16235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5132455825805664,
+      "learning_rate": 5.164473684210527e-06,
+      "loss": 1.572,
+      "step": 16236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3812265396118164,
+      "learning_rate": 5.163559941520468e-06,
+      "loss": 1.3249,
+      "step": 16237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1754627227783203,
+      "learning_rate": 5.16264619883041e-06,
+      "loss": 1.3788,
+      "step": 16238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.00494122505188,
+      "learning_rate": 5.161732456140352e-06,
+      "loss": 1.2199,
+      "step": 16239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.350706100463867,
+      "learning_rate": 5.160818713450293e-06,
+      "loss": 1.2582,
+      "step": 16240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.83766770362854,
+      "learning_rate": 5.1599049707602345e-06,
+      "loss": 1.831,
+      "step": 16241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.857480525970459,
+      "learning_rate": 5.158991228070176e-06,
+      "loss": 1.6918,
+      "step": 16242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.634910821914673,
+      "learning_rate": 5.158077485380117e-06,
+      "loss": 1.745,
+      "step": 16243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7324442863464355,
+      "learning_rate": 5.157163742690059e-06,
+      "loss": 1.4532,
+      "step": 16244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5481209754943848,
+      "learning_rate": 5.156250000000001e-06,
+      "loss": 1.7188,
+      "step": 16245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.761428117752075,
+      "learning_rate": 5.155336257309942e-06,
+      "loss": 1.7189,
+      "step": 16246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6693878173828125,
+      "learning_rate": 5.1544225146198836e-06,
+      "loss": 1.2106,
+      "step": 16247
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8512675762176514,
+      "learning_rate": 5.153508771929825e-06,
+      "loss": 1.7627,
+      "step": 16248
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5434842109680176,
+      "learning_rate": 5.152595029239766e-06,
+      "loss": 1.8738,
+      "step": 16249
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0937750339508057,
+      "learning_rate": 5.151681286549708e-06,
+      "loss": 1.3662,
+      "step": 16250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.176913261413574,
+      "learning_rate": 5.150767543859649e-06,
+      "loss": 1.2496,
+      "step": 16251
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2647545337677,
+      "learning_rate": 5.149853801169591e-06,
+      "loss": 1.8068,
+      "step": 16252
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6081647872924805,
+      "learning_rate": 5.148940058479533e-06,
+      "loss": 1.4727,
+      "step": 16253
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4030070304870605,
+      "learning_rate": 5.148026315789474e-06,
+      "loss": 1.5876,
+      "step": 16254
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.184211730957031,
+      "learning_rate": 5.147112573099415e-06,
+      "loss": 1.7828,
+      "step": 16255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1506948471069336,
+      "learning_rate": 5.146198830409357e-06,
+      "loss": 1.2202,
+      "step": 16256
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8918380737304688,
+      "learning_rate": 5.145285087719298e-06,
+      "loss": 1.7042,
+      "step": 16257
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4674015045166016,
+      "learning_rate": 5.1443713450292395e-06,
+      "loss": 1.4057,
+      "step": 16258
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5253822803497314,
+      "learning_rate": 5.143457602339182e-06,
+      "loss": 1.5635,
+      "step": 16259
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.68947696685791,
+      "learning_rate": 5.142543859649123e-06,
+      "loss": 1.6212,
+      "step": 16260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.905034065246582,
+      "learning_rate": 5.1416301169590645e-06,
+      "loss": 1.4273,
+      "step": 16261
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.692070960998535,
+      "learning_rate": 5.140716374269006e-06,
+      "loss": 1.3306,
+      "step": 16262
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.765566825866699,
+      "learning_rate": 5.139802631578947e-06,
+      "loss": 1.0682,
+      "step": 16263
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.578594446182251,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 1.4589,
+      "step": 16264
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5209481716156006,
+      "learning_rate": 5.13797514619883e-06,
+      "loss": 1.4987,
+      "step": 16265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.165969133377075,
+      "learning_rate": 5.137061403508772e-06,
+      "loss": 1.4374,
+      "step": 16266
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.919834613800049,
+      "learning_rate": 5.1361476608187144e-06,
+      "loss": 1.4795,
+      "step": 16267
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.530989646911621,
+      "learning_rate": 5.135233918128656e-06,
+      "loss": 1.3467,
+      "step": 16268
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2542665004730225,
+      "learning_rate": 5.134320175438597e-06,
+      "loss": 1.2407,
+      "step": 16269
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5132293701171875,
+      "learning_rate": 5.1334064327485386e-06,
+      "loss": 1.5621,
+      "step": 16270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.531886339187622,
+      "learning_rate": 5.132492690058481e-06,
+      "loss": 1.2636,
+      "step": 16271
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0134377479553223,
+      "learning_rate": 5.131578947368422e-06,
+      "loss": 1.1833,
+      "step": 16272
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1445565223693848,
+      "learning_rate": 5.1306652046783635e-06,
+      "loss": 1.2635,
+      "step": 16273
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.460538148880005,
+      "learning_rate": 5.129751461988305e-06,
+      "loss": 1.4353,
+      "step": 16274
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5397427082061768,
+      "learning_rate": 5.128837719298246e-06,
+      "loss": 1.4608,
+      "step": 16275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.176992177963257,
+      "learning_rate": 5.127923976608188e-06,
+      "loss": 1.1771,
+      "step": 16276
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5984511375427246,
+      "learning_rate": 5.127010233918129e-06,
+      "loss": 1.3984,
+      "step": 16277
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.484670400619507,
+      "learning_rate": 5.126096491228071e-06,
+      "loss": 1.5373,
+      "step": 16278
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9980077743530273,
+      "learning_rate": 5.125182748538013e-06,
+      "loss": 1.424,
+      "step": 16279
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.463715076446533,
+      "learning_rate": 5.124269005847954e-06,
+      "loss": 1.4145,
+      "step": 16280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.068300247192383,
+      "learning_rate": 5.123355263157895e-06,
+      "loss": 1.7075,
+      "step": 16281
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2257790565490723,
+      "learning_rate": 5.122441520467837e-06,
+      "loss": 1.3585,
+      "step": 16282
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5248093605041504,
+      "learning_rate": 5.121527777777778e-06,
+      "loss": 1.4672,
+      "step": 16283
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.384986162185669,
+      "learning_rate": 5.1206140350877195e-06,
+      "loss": 1.1348,
+      "step": 16284
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3113315105438232,
+      "learning_rate": 5.119700292397662e-06,
+      "loss": 1.5747,
+      "step": 16285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6285948753356934,
+      "learning_rate": 5.118786549707603e-06,
+      "loss": 1.4268,
+      "step": 16286
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.812767505645752,
+      "learning_rate": 5.1178728070175445e-06,
+      "loss": 1.5081,
+      "step": 16287
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.91723370552063,
+      "learning_rate": 5.116959064327486e-06,
+      "loss": 1.7292,
+      "step": 16288
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8733415603637695,
+      "learning_rate": 5.116045321637427e-06,
+      "loss": 1.448,
+      "step": 16289
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.167863130569458,
+      "learning_rate": 5.115131578947369e-06,
+      "loss": 1.9776,
+      "step": 16290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.226228952407837,
+      "learning_rate": 5.11421783625731e-06,
+      "loss": 1.3622,
+      "step": 16291
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3674705028533936,
+      "learning_rate": 5.113304093567252e-06,
+      "loss": 1.418,
+      "step": 16292
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.84967303276062,
+      "learning_rate": 5.1123903508771936e-06,
+      "loss": 1.4808,
+      "step": 16293
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9170777797698975,
+      "learning_rate": 5.111476608187135e-06,
+      "loss": 1.4465,
+      "step": 16294
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.186072587966919,
+      "learning_rate": 5.110562865497076e-06,
+      "loss": 1.2912,
+      "step": 16295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.555589437484741,
+      "learning_rate": 5.109649122807018e-06,
+      "loss": 1.5783,
+      "step": 16296
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3162906169891357,
+      "learning_rate": 5.108735380116959e-06,
+      "loss": 1.2718,
+      "step": 16297
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0232434272766113,
+      "learning_rate": 5.1078216374269004e-06,
+      "loss": 1.1839,
+      "step": 16298
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.166062593460083,
+      "learning_rate": 5.106907894736842e-06,
+      "loss": 1.0528,
+      "step": 16299
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5440595149993896,
+      "learning_rate": 5.105994152046784e-06,
+      "loss": 1.5036,
+      "step": 16300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3541224002838135,
+      "learning_rate": 5.105080409356725e-06,
+      "loss": 1.1849,
+      "step": 16301
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.08250093460083,
+      "learning_rate": 5.104166666666667e-06,
+      "loss": 2.1388,
+      "step": 16302
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1204118728637695,
+      "learning_rate": 5.103252923976608e-06,
+      "loss": 1.3136,
+      "step": 16303
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8075144290924072,
+      "learning_rate": 5.1023391812865495e-06,
+      "loss": 1.7879,
+      "step": 16304
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.768716335296631,
+      "learning_rate": 5.101425438596491e-06,
+      "loss": 1.6643,
+      "step": 16305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.006281614303589,
+      "learning_rate": 5.100511695906432e-06,
+      "loss": 1.2251,
+      "step": 16306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6826934814453125,
+      "learning_rate": 5.0995979532163745e-06,
+      "loss": 1.6479,
+      "step": 16307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0342648029327393,
+      "learning_rate": 5.098684210526316e-06,
+      "loss": 1.7678,
+      "step": 16308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7277615070343018,
+      "learning_rate": 5.097770467836257e-06,
+      "loss": 1.4311,
+      "step": 16309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7341763973236084,
+      "learning_rate": 5.0968567251461995e-06,
+      "loss": 1.4608,
+      "step": 16310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.549326181411743,
+      "learning_rate": 5.095942982456142e-06,
+      "loss": 1.2384,
+      "step": 16311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.464938163757324,
+      "learning_rate": 5.095029239766083e-06,
+      "loss": 1.3303,
+      "step": 16312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2836639881134033,
+      "learning_rate": 5.0941154970760244e-06,
+      "loss": 1.9883,
+      "step": 16313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0236148834228516,
+      "learning_rate": 5.093201754385966e-06,
+      "loss": 1.2971,
+      "step": 16314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3693411350250244,
+      "learning_rate": 5.092288011695907e-06,
+      "loss": 1.5961,
+      "step": 16315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3286733627319336,
+      "learning_rate": 5.0913742690058486e-06,
+      "loss": 1.4133,
+      "step": 16316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.763399362564087,
+      "learning_rate": 5.09046052631579e-06,
+      "loss": 1.7619,
+      "step": 16317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.154634952545166,
+      "learning_rate": 5.089546783625731e-06,
+      "loss": 1.4005,
+      "step": 16318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5686259269714355,
+      "learning_rate": 5.0886330409356735e-06,
+      "loss": 1.3703,
+      "step": 16319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.403776168823242,
+      "learning_rate": 5.087719298245615e-06,
+      "loss": 1.1236,
+      "step": 16320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.469848394393921,
+      "learning_rate": 5.086805555555556e-06,
+      "loss": 1.3686,
+      "step": 16321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1283819675445557,
+      "learning_rate": 5.085891812865498e-06,
+      "loss": 1.3534,
+      "step": 16322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.75264835357666,
+      "learning_rate": 5.084978070175439e-06,
+      "loss": 1.655,
+      "step": 16323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2922463417053223,
+      "learning_rate": 5.08406432748538e-06,
+      "loss": 1.1921,
+      "step": 16324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.695621967315674,
+      "learning_rate": 5.083150584795322e-06,
+      "loss": 1.5119,
+      "step": 16325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7678117752075195,
+      "learning_rate": 5.082236842105264e-06,
+      "loss": 1.3395,
+      "step": 16326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7450900077819824,
+      "learning_rate": 5.081323099415205e-06,
+      "loss": 1.6397,
+      "step": 16327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6075801849365234,
+      "learning_rate": 5.080409356725147e-06,
+      "loss": 1.4365,
+      "step": 16328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5418684482574463,
+      "learning_rate": 5.079495614035088e-06,
+      "loss": 1.6031,
+      "step": 16329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1879169940948486,
+      "learning_rate": 5.0785818713450295e-06,
+      "loss": 1.4037,
+      "step": 16330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.677751302719116,
+      "learning_rate": 5.077668128654971e-06,
+      "loss": 1.3803,
+      "step": 16331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3099825382232666,
+      "learning_rate": 5.076754385964912e-06,
+      "loss": 1.5749,
+      "step": 16332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.749037504196167,
+      "learning_rate": 5.0758406432748545e-06,
+      "loss": 1.692,
+      "step": 16333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.542051076889038,
+      "learning_rate": 5.074926900584796e-06,
+      "loss": 1.4236,
+      "step": 16334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8051278591156006,
+      "learning_rate": 5.074013157894737e-06,
+      "loss": 1.471,
+      "step": 16335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4783148765563965,
+      "learning_rate": 5.073099415204679e-06,
+      "loss": 1.5915,
+      "step": 16336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.299548864364624,
+      "learning_rate": 5.07218567251462e-06,
+      "loss": 1.5457,
+      "step": 16337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.709054946899414,
+      "learning_rate": 5.071271929824561e-06,
+      "loss": 1.5815,
+      "step": 16338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.739893913269043,
+      "learning_rate": 5.070358187134503e-06,
+      "loss": 1.7,
+      "step": 16339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5729763507843018,
+      "learning_rate": 5.069444444444445e-06,
+      "loss": 1.417,
+      "step": 16340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.159759998321533,
+      "learning_rate": 5.068530701754386e-06,
+      "loss": 1.4384,
+      "step": 16341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.662353754043579,
+      "learning_rate": 5.067616959064328e-06,
+      "loss": 1.5313,
+      "step": 16342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6211187839508057,
+      "learning_rate": 5.066703216374269e-06,
+      "loss": 1.5265,
+      "step": 16343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2875421047210693,
+      "learning_rate": 5.0657894736842104e-06,
+      "loss": 1.3937,
+      "step": 16344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.099187135696411,
+      "learning_rate": 5.064875730994152e-06,
+      "loss": 1.6103,
+      "step": 16345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6531190872192383,
+      "learning_rate": 5.063961988304093e-06,
+      "loss": 1.7426,
+      "step": 16346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5120158195495605,
+      "learning_rate": 5.063048245614035e-06,
+      "loss": 1.3487,
+      "step": 16347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3709936141967773,
+      "learning_rate": 5.062134502923977e-06,
+      "loss": 1.2849,
+      "step": 16348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.674197196960449,
+      "learning_rate": 5.061220760233918e-06,
+      "loss": 1.5341,
+      "step": 16349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.312530517578125,
+      "learning_rate": 5.0603070175438595e-06,
+      "loss": 1.1175,
+      "step": 16350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.102747917175293,
+      "learning_rate": 5.059393274853801e-06,
+      "loss": 1.1244,
+      "step": 16351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5180420875549316,
+      "learning_rate": 5.058479532163744e-06,
+      "loss": 1.6368,
+      "step": 16352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4360952377319336,
+      "learning_rate": 5.057565789473685e-06,
+      "loss": 1.7011,
+      "step": 16353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8003761768341064,
+      "learning_rate": 5.056652046783627e-06,
+      "loss": 1.6486,
+      "step": 16354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9142985343933105,
+      "learning_rate": 5.055738304093568e-06,
+      "loss": 1.6186,
+      "step": 16355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9165854454040527,
+      "learning_rate": 5.0548245614035095e-06,
+      "loss": 1.5044,
+      "step": 16356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6313469409942627,
+      "learning_rate": 5.053910818713451e-06,
+      "loss": 1.5264,
+      "step": 16357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8362622261047363,
+      "learning_rate": 5.052997076023392e-06,
+      "loss": 1.573,
+      "step": 16358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.65348219871521,
+      "learning_rate": 5.0520833333333344e-06,
+      "loss": 1.5496,
+      "step": 16359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4859893321990967,
+      "learning_rate": 5.051169590643276e-06,
+      "loss": 1.2114,
+      "step": 16360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.937577486038208,
+      "learning_rate": 5.050255847953217e-06,
+      "loss": 1.2814,
+      "step": 16361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.527081251144409,
+      "learning_rate": 5.0493421052631586e-06,
+      "loss": 1.203,
+      "step": 16362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6831352710723877,
+      "learning_rate": 5.0484283625731e-06,
+      "loss": 1.382,
+      "step": 16363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1647191047668457,
+      "learning_rate": 5.047514619883041e-06,
+      "loss": 1.2537,
+      "step": 16364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.44767689704895,
+      "learning_rate": 5.046600877192983e-06,
+      "loss": 1.4956,
+      "step": 16365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3922290802001953,
+      "learning_rate": 5.045687134502925e-06,
+      "loss": 1.3302,
+      "step": 16366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0738654136657715,
+      "learning_rate": 5.044773391812866e-06,
+      "loss": 1.3425,
+      "step": 16367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4967052936553955,
+      "learning_rate": 5.043859649122808e-06,
+      "loss": 1.4243,
+      "step": 16368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5310544967651367,
+      "learning_rate": 5.042945906432749e-06,
+      "loss": 1.3698,
+      "step": 16369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0023629665374756,
+      "learning_rate": 5.04203216374269e-06,
+      "loss": 1.266,
+      "step": 16370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.836756467819214,
+      "learning_rate": 5.041118421052632e-06,
+      "loss": 1.1388,
+      "step": 16371
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7296950817108154,
+      "learning_rate": 5.040204678362573e-06,
+      "loss": 1.525,
+      "step": 16372
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.388972282409668,
+      "learning_rate": 5.039290935672515e-06,
+      "loss": 1.5376,
+      "step": 16373
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.58558988571167,
+      "learning_rate": 5.038377192982457e-06,
+      "loss": 1.4689,
+      "step": 16374
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.664793014526367,
+      "learning_rate": 5.037463450292398e-06,
+      "loss": 1.5028,
+      "step": 16375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9901769161224365,
+      "learning_rate": 5.0365497076023395e-06,
+      "loss": 1.5331,
+      "step": 16376
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.666149139404297,
+      "learning_rate": 5.035635964912281e-06,
+      "loss": 1.2688,
+      "step": 16377
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.49364972114563,
+      "learning_rate": 5.034722222222222e-06,
+      "loss": 1.6567,
+      "step": 16378
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2112388610839844,
+      "learning_rate": 5.033808479532164e-06,
+      "loss": 1.1674,
+      "step": 16379
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9908828735351562,
+      "learning_rate": 5.032894736842106e-06,
+      "loss": 1.4206,
+      "step": 16380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.402144432067871,
+      "learning_rate": 5.031980994152047e-06,
+      "loss": 1.3769,
+      "step": 16381
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2672464847564697,
+      "learning_rate": 5.031067251461989e-06,
+      "loss": 1.404,
+      "step": 16382
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5602734088897705,
+      "learning_rate": 5.03015350877193e-06,
+      "loss": 1.1669,
+      "step": 16383
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4502816200256348,
+      "learning_rate": 5.029239766081871e-06,
+      "loss": 1.5472,
+      "step": 16384
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.151925563812256,
+      "learning_rate": 5.028326023391813e-06,
+      "loss": 1.0187,
+      "step": 16385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1739394664764404,
+      "learning_rate": 5.027412280701754e-06,
+      "loss": 1.2218,
+      "step": 16386
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4431660175323486,
+      "learning_rate": 5.026498538011696e-06,
+      "loss": 1.3004,
+      "step": 16387
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.456583023071289,
+      "learning_rate": 5.025584795321638e-06,
+      "loss": 1.3382,
+      "step": 16388
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1499526500701904,
+      "learning_rate": 5.024671052631579e-06,
+      "loss": 1.2662,
+      "step": 16389
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.984107494354248,
+      "learning_rate": 5.0237573099415204e-06,
+      "loss": 1.0152,
+      "step": 16390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.448003053665161,
+      "learning_rate": 5.022843567251462e-06,
+      "loss": 1.4732,
+      "step": 16391
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5114693641662598,
+      "learning_rate": 5.021929824561403e-06,
+      "loss": 1.5158,
+      "step": 16392
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6329762935638428,
+      "learning_rate": 5.0210160818713446e-06,
+      "loss": 1.832,
+      "step": 16393
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.990990400314331,
+      "learning_rate": 5.020102339181287e-06,
+      "loss": 1.9423,
+      "step": 16394
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.315230131149292,
+      "learning_rate": 5.019188596491229e-06,
+      "loss": 1.7922,
+      "step": 16395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1912195682525635,
+      "learning_rate": 5.01827485380117e-06,
+      "loss": 1.6804,
+      "step": 16396
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0420751571655273,
+      "learning_rate": 5.017361111111112e-06,
+      "loss": 1.7777,
+      "step": 16397
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4219720363616943,
+      "learning_rate": 5.016447368421053e-06,
+      "loss": 1.7363,
+      "step": 16398
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.470599889755249,
+      "learning_rate": 5.015533625730995e-06,
+      "loss": 1.3849,
+      "step": 16399
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1229655742645264,
+      "learning_rate": 5.014619883040937e-06,
+      "loss": 1.7542,
+      "step": 16400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.050670623779297,
+      "learning_rate": 5.013706140350878e-06,
+      "loss": 1.2264,
+      "step": 16401
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.558129072189331,
+      "learning_rate": 5.0127923976608195e-06,
+      "loss": 1.7197,
+      "step": 16402
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5973172187805176,
+      "learning_rate": 5.011878654970761e-06,
+      "loss": 1.6179,
+      "step": 16403
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1389830112457275,
+      "learning_rate": 5.010964912280702e-06,
+      "loss": 1.2947,
+      "step": 16404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6205074787139893,
+      "learning_rate": 5.010051169590644e-06,
+      "loss": 1.5392,
+      "step": 16405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6763687133789062,
+      "learning_rate": 5.009137426900586e-06,
+      "loss": 1.6961,
+      "step": 16406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3494579792022705,
+      "learning_rate": 5.008223684210527e-06,
+      "loss": 1.4463,
+      "step": 16407
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7698884010314941,
+      "learning_rate": 5.0073099415204686e-06,
+      "loss": 1.1257,
+      "step": 16408
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5855910778045654,
+      "learning_rate": 5.00639619883041e-06,
+      "loss": 1.3861,
+      "step": 16409
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9099104404449463,
+      "learning_rate": 5.005482456140351e-06,
+      "loss": 1.1748,
+      "step": 16410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.451909065246582,
+      "learning_rate": 5.004568713450293e-06,
+      "loss": 1.4059,
+      "step": 16411
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.634427785873413,
+      "learning_rate": 5.003654970760234e-06,
+      "loss": 1.8905,
+      "step": 16412
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4107532501220703,
+      "learning_rate": 5.002741228070176e-06,
+      "loss": 1.701,
+      "step": 16413
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.595092296600342,
+      "learning_rate": 5.001827485380118e-06,
+      "loss": 1.5426,
+      "step": 16414
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.267085313796997,
+      "learning_rate": 5.000913742690059e-06,
+      "loss": 1.3784,
+      "step": 16415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.91243577003479,
+      "learning_rate": 5e-06,
+      "loss": 1.4159,
+      "step": 16416
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9095797538757324,
+      "learning_rate": 4.999086257309942e-06,
+      "loss": 1.2011,
+      "step": 16417
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.707321882247925,
+      "learning_rate": 4.998172514619883e-06,
+      "loss": 1.2968,
+      "step": 16418
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7538363933563232,
+      "learning_rate": 4.9972587719298245e-06,
+      "loss": 1.5294,
+      "step": 16419
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4010305404663086,
+      "learning_rate": 4.996345029239767e-06,
+      "loss": 1.3887,
+      "step": 16420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.591179847717285,
+      "learning_rate": 4.995431286549708e-06,
+      "loss": 1.5625,
+      "step": 16421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.136510133743286,
+      "learning_rate": 4.9945175438596495e-06,
+      "loss": 1.3593,
+      "step": 16422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8967653512954712,
+      "learning_rate": 4.993603801169591e-06,
+      "loss": 1.0145,
+      "step": 16423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.349400043487549,
+      "learning_rate": 4.992690058479532e-06,
+      "loss": 1.47,
+      "step": 16424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.741769790649414,
+      "learning_rate": 4.991776315789474e-06,
+      "loss": 1.9265,
+      "step": 16425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.882404327392578,
+      "learning_rate": 4.990862573099415e-06,
+      "loss": 1.5447,
+      "step": 16426
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.981339693069458,
+      "learning_rate": 4.989948830409357e-06,
+      "loss": 1.8977,
+      "step": 16427
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6963319778442383,
+      "learning_rate": 4.989035087719299e-06,
+      "loss": 1.5668,
+      "step": 16428
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.28661847114563,
+      "learning_rate": 4.98812134502924e-06,
+      "loss": 1.2046,
+      "step": 16429
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.355724334716797,
+      "learning_rate": 4.987207602339182e-06,
+      "loss": 1.1562,
+      "step": 16430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.122068405151367,
+      "learning_rate": 4.9862938596491236e-06,
+      "loss": 1.3708,
+      "step": 16431
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.53888201713562,
+      "learning_rate": 4.985380116959065e-06,
+      "loss": 1.3984,
+      "step": 16432
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8658995628356934,
+      "learning_rate": 4.984466374269006e-06,
+      "loss": 1.5472,
+      "step": 16433
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.545524835586548,
+      "learning_rate": 4.983552631578948e-06,
+      "loss": 1.5307,
+      "step": 16434
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9875550270080566,
+      "learning_rate": 4.982638888888889e-06,
+      "loss": 1.8499,
+      "step": 16435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.558694839477539,
+      "learning_rate": 4.9817251461988304e-06,
+      "loss": 1.6393,
+      "step": 16436
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8037614822387695,
+      "learning_rate": 4.980811403508773e-06,
+      "loss": 1.8013,
+      "step": 16437
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.540189743041992,
+      "learning_rate": 4.979897660818714e-06,
+      "loss": 1.6555,
+      "step": 16438
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.308112382888794,
+      "learning_rate": 4.978983918128655e-06,
+      "loss": 1.311,
+      "step": 16439
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8945648670196533,
+      "learning_rate": 4.978070175438597e-06,
+      "loss": 1.6951,
+      "step": 16440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.188504219055176,
+      "learning_rate": 4.977156432748538e-06,
+      "loss": 1.3105,
+      "step": 16441
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6385574340820312,
+      "learning_rate": 4.9762426900584795e-06,
+      "loss": 1.4065,
+      "step": 16442
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2563388347625732,
+      "learning_rate": 4.975328947368421e-06,
+      "loss": 1.4069,
+      "step": 16443
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6945858001708984,
+      "learning_rate": 4.974415204678363e-06,
+      "loss": 1.4563,
+      "step": 16444
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6228973865509033,
+      "learning_rate": 4.9735014619883045e-06,
+      "loss": 1.4817,
+      "step": 16445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4107720851898193,
+      "learning_rate": 4.972587719298246e-06,
+      "loss": 1.3816,
+      "step": 16446
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.241279125213623,
+      "learning_rate": 4.971673976608187e-06,
+      "loss": 1.2075,
+      "step": 16447
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0840718746185303,
+      "learning_rate": 4.970760233918129e-06,
+      "loss": 1.1362,
+      "step": 16448
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6825053691864014,
+      "learning_rate": 4.969846491228071e-06,
+      "loss": 1.6685,
+      "step": 16449
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7242348194122314,
+      "learning_rate": 4.968932748538012e-06,
+      "loss": 1.4768,
+      "step": 16450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.343893051147461,
+      "learning_rate": 4.968019005847954e-06,
+      "loss": 1.4098,
+      "step": 16451
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1330106258392334,
+      "learning_rate": 4.967105263157895e-06,
+      "loss": 1.9331,
+      "step": 16452
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6517326831817627,
+      "learning_rate": 4.966191520467837e-06,
+      "loss": 1.3226,
+      "step": 16453
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7991943359375,
+      "learning_rate": 4.9652777777777786e-06,
+      "loss": 1.4685,
+      "step": 16454
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.214250087738037,
+      "learning_rate": 4.96436403508772e-06,
+      "loss": 1.4735,
+      "step": 16455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.035413980484009,
+      "learning_rate": 4.963450292397661e-06,
+      "loss": 2.0548,
+      "step": 16456
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0177910327911377,
+      "learning_rate": 4.962536549707603e-06,
+      "loss": 0.9825,
+      "step": 16457
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.536172389984131,
+      "learning_rate": 4.961622807017544e-06,
+      "loss": 1.0997,
+      "step": 16458
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.215268611907959,
+      "learning_rate": 4.9607090643274854e-06,
+      "loss": 1.7778,
+      "step": 16459
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7635629177093506,
+      "learning_rate": 4.959795321637428e-06,
+      "loss": 1.4906,
+      "step": 16460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.525869369506836,
+      "learning_rate": 4.958881578947369e-06,
+      "loss": 1.6163,
+      "step": 16461
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6199355125427246,
+      "learning_rate": 4.95796783625731e-06,
+      "loss": 1.4252,
+      "step": 16462
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7053966522216797,
+      "learning_rate": 4.957054093567252e-06,
+      "loss": 1.5342,
+      "step": 16463
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4555206298828125,
+      "learning_rate": 4.956140350877193e-06,
+      "loss": 1.3397,
+      "step": 16464
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.431711196899414,
+      "learning_rate": 4.9552266081871345e-06,
+      "loss": 1.4166,
+      "step": 16465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2096006870269775,
+      "learning_rate": 4.954312865497076e-06,
+      "loss": 1.2822,
+      "step": 16466
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9100751876831055,
+      "learning_rate": 4.953399122807018e-06,
+      "loss": 1.605,
+      "step": 16467
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.00697660446167,
+      "learning_rate": 4.9524853801169595e-06,
+      "loss": 1.2159,
+      "step": 16468
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1664140224456787,
+      "learning_rate": 4.951571637426901e-06,
+      "loss": 1.2984,
+      "step": 16469
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2547314167022705,
+      "learning_rate": 4.950657894736843e-06,
+      "loss": 1.3321,
+      "step": 16470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1406314373016357,
+      "learning_rate": 4.9497441520467845e-06,
+      "loss": 1.144,
+      "step": 16471
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5062999725341797,
+      "learning_rate": 4.948830409356726e-06,
+      "loss": 1.5127,
+      "step": 16472
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6269185543060303,
+      "learning_rate": 4.947916666666667e-06,
+      "loss": 1.3578,
+      "step": 16473
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6069247722625732,
+      "learning_rate": 4.947002923976609e-06,
+      "loss": 1.5092,
+      "step": 16474
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3591554164886475,
+      "learning_rate": 4.94608918128655e-06,
+      "loss": 1.2881,
+      "step": 16475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1333231925964355,
+      "learning_rate": 4.945175438596491e-06,
+      "loss": 1.2253,
+      "step": 16476
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2767512798309326,
+      "learning_rate": 4.9442616959064336e-06,
+      "loss": 1.6571,
+      "step": 16477
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4268500804901123,
+      "learning_rate": 4.943347953216375e-06,
+      "loss": 1.272,
+      "step": 16478
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0266523361206055,
+      "learning_rate": 4.942434210526316e-06,
+      "loss": 1.6114,
+      "step": 16479
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8975088596343994,
+      "learning_rate": 4.941520467836258e-06,
+      "loss": 1.5404,
+      "step": 16480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.521899700164795,
+      "learning_rate": 4.940606725146199e-06,
+      "loss": 1.5679,
+      "step": 16481
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5009617805480957,
+      "learning_rate": 4.9396929824561404e-06,
+      "loss": 1.5397,
+      "step": 16482
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4141788482666016,
+      "learning_rate": 4.938779239766082e-06,
+      "loss": 1.4843,
+      "step": 16483
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.281102418899536,
+      "learning_rate": 4.937865497076024e-06,
+      "loss": 1.4904,
+      "step": 16484
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.506891965866089,
+      "learning_rate": 4.936951754385965e-06,
+      "loss": 1.9051,
+      "step": 16485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8992117643356323,
+      "learning_rate": 4.936038011695907e-06,
+      "loss": 1.1373,
+      "step": 16486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0786147117614746,
+      "learning_rate": 4.935124269005848e-06,
+      "loss": 1.127,
+      "step": 16487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0083467960357666,
+      "learning_rate": 4.9342105263157895e-06,
+      "loss": 1.4382,
+      "step": 16488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5105082988739014,
+      "learning_rate": 4.933296783625731e-06,
+      "loss": 1.347,
+      "step": 16489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7377209663391113,
+      "learning_rate": 4.932383040935672e-06,
+      "loss": 1.7199,
+      "step": 16490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1444897651672363,
+      "learning_rate": 4.9314692982456145e-06,
+      "loss": 1.5444,
+      "step": 16491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2018184661865234,
+      "learning_rate": 4.930555555555556e-06,
+      "loss": 1.4706,
+      "step": 16492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9568963050842285,
+      "learning_rate": 4.929641812865498e-06,
+      "loss": 1.5212,
+      "step": 16493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6634697914123535,
+      "learning_rate": 4.9287280701754395e-06,
+      "loss": 1.6442,
+      "step": 16494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.959125280380249,
+      "learning_rate": 4.927814327485381e-06,
+      "loss": 1.5838,
+      "step": 16495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2951693534851074,
+      "learning_rate": 4.926900584795322e-06,
+      "loss": 1.3265,
+      "step": 16496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2855172157287598,
+      "learning_rate": 4.925986842105264e-06,
+      "loss": 1.3109,
+      "step": 16497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.590510129928589,
+      "learning_rate": 4.925073099415205e-06,
+      "loss": 1.553,
+      "step": 16498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.360837697982788,
+      "learning_rate": 4.924159356725146e-06,
+      "loss": 1.318,
+      "step": 16499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.768406629562378,
+      "learning_rate": 4.9232456140350886e-06,
+      "loss": 1.6883,
+      "step": 16500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1245861053466797,
+      "learning_rate": 4.92233187134503e-06,
+      "loss": 1.2152,
+      "step": 16501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.63112211227417,
+      "learning_rate": 4.921418128654971e-06,
+      "loss": 1.2671,
+      "step": 16502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0647387504577637,
+      "learning_rate": 4.920504385964913e-06,
+      "loss": 1.2774,
+      "step": 16503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2674713134765625,
+      "learning_rate": 4.919590643274854e-06,
+      "loss": 1.1939,
+      "step": 16504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.487097978591919,
+      "learning_rate": 4.9186769005847954e-06,
+      "loss": 1.4109,
+      "step": 16505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.580817699432373,
+      "learning_rate": 4.917763157894737e-06,
+      "loss": 1.6643,
+      "step": 16506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7690787315368652,
+      "learning_rate": 4.916849415204679e-06,
+      "loss": 1.7971,
+      "step": 16507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6604461669921875,
+      "learning_rate": 4.91593567251462e-06,
+      "loss": 1.6621,
+      "step": 16508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.238450288772583,
+      "learning_rate": 4.915021929824562e-06,
+      "loss": 1.45,
+      "step": 16509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.027837038040161,
+      "learning_rate": 4.914108187134503e-06,
+      "loss": 0.9994,
+      "step": 16510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.213167428970337,
+      "learning_rate": 4.9131944444444445e-06,
+      "loss": 1.4617,
+      "step": 16511
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.779116630554199,
+      "learning_rate": 4.912280701754386e-06,
+      "loss": 1.4507,
+      "step": 16512
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.972759962081909,
+      "learning_rate": 4.911366959064328e-06,
+      "loss": 1.6766,
+      "step": 16513
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.385577917098999,
+      "learning_rate": 4.9104532163742695e-06,
+      "loss": 1.3697,
+      "step": 16514
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2349913120269775,
+      "learning_rate": 4.909539473684211e-06,
+      "loss": 1.5197,
+      "step": 16515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9680161476135254,
+      "learning_rate": 4.908625730994152e-06,
+      "loss": 1.3471,
+      "step": 16516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.008702278137207,
+      "learning_rate": 4.9077119883040945e-06,
+      "loss": 1.7167,
+      "step": 16517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3984100818634033,
+      "learning_rate": 4.906798245614036e-06,
+      "loss": 1.7115,
+      "step": 16518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.786599636077881,
+      "learning_rate": 4.905884502923977e-06,
+      "loss": 1.5497,
+      "step": 16519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.428447723388672,
+      "learning_rate": 4.904970760233919e-06,
+      "loss": 1.3817,
+      "step": 16520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7019214630126953,
+      "learning_rate": 4.90405701754386e-06,
+      "loss": 1.627,
+      "step": 16521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.722804069519043,
+      "learning_rate": 4.903143274853801e-06,
+      "loss": 1.4302,
+      "step": 16522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.734999418258667,
+      "learning_rate": 4.902229532163743e-06,
+      "loss": 1.5272,
+      "step": 16523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3408243656158447,
+      "learning_rate": 4.901315789473685e-06,
+      "loss": 1.2473,
+      "step": 16524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.609069585800171,
+      "learning_rate": 4.900402046783626e-06,
+      "loss": 1.5101,
+      "step": 16525
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5515215396881104,
+      "learning_rate": 4.899488304093568e-06,
+      "loss": 1.2743,
+      "step": 16526
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4200518131256104,
+      "learning_rate": 4.898574561403509e-06,
+      "loss": 1.5917,
+      "step": 16527
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9112732410430908,
+      "learning_rate": 4.8976608187134504e-06,
+      "loss": 1.1373,
+      "step": 16528
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3202385902404785,
+      "learning_rate": 4.896747076023392e-06,
+      "loss": 1.339,
+      "step": 16529
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.721000909805298,
+      "learning_rate": 4.895833333333333e-06,
+      "loss": 1.4192,
+      "step": 16530
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.696458101272583,
+      "learning_rate": 4.894919590643275e-06,
+      "loss": 1.3028,
+      "step": 16531
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1401612758636475,
+      "learning_rate": 4.894005847953217e-06,
+      "loss": 1.3826,
+      "step": 16532
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.329054117202759,
+      "learning_rate": 4.893092105263158e-06,
+      "loss": 1.2588,
+      "step": 16533
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.542163848876953,
+      "learning_rate": 4.8921783625731e-06,
+      "loss": 1.2023,
+      "step": 16534
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7338321208953857,
+      "learning_rate": 4.891264619883042e-06,
+      "loss": 1.4417,
+      "step": 16535
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3029229640960693,
+      "learning_rate": 4.890350877192983e-06,
+      "loss": 1.2165,
+      "step": 16536
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.223483085632324,
+      "learning_rate": 4.8894371345029245e-06,
+      "loss": 1.9214,
+      "step": 16537
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.541002035140991,
+      "learning_rate": 4.888523391812866e-06,
+      "loss": 1.6168,
+      "step": 16538
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3884525299072266,
+      "learning_rate": 4.887609649122807e-06,
+      "loss": 1.433,
+      "step": 16539
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.691206455230713,
+      "learning_rate": 4.886695906432749e-06,
+      "loss": 1.4163,
+      "step": 16540
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.279686689376831,
+      "learning_rate": 4.885782163742691e-06,
+      "loss": 1.1217,
+      "step": 16541
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.565980911254883,
+      "learning_rate": 4.884868421052632e-06,
+      "loss": 1.5305,
+      "step": 16542
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.666654586791992,
+      "learning_rate": 4.883954678362574e-06,
+      "loss": 2.0023,
+      "step": 16543
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3603456020355225,
+      "learning_rate": 4.883040935672515e-06,
+      "loss": 1.4038,
+      "step": 16544
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1910037994384766,
+      "learning_rate": 4.882127192982456e-06,
+      "loss": 1.6515,
+      "step": 16545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5513644218444824,
+      "learning_rate": 4.881213450292398e-06,
+      "loss": 1.3514,
+      "step": 16546
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3288586139678955,
+      "learning_rate": 4.880299707602339e-06,
+      "loss": 1.2256,
+      "step": 16547
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.202045202255249,
+      "learning_rate": 4.879385964912281e-06,
+      "loss": 1.0997,
+      "step": 16548
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5585334300994873,
+      "learning_rate": 4.878472222222223e-06,
+      "loss": 1.4321,
+      "step": 16549
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7412221431732178,
+      "learning_rate": 4.877558479532164e-06,
+      "loss": 1.3966,
+      "step": 16550
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.332780122756958,
+      "learning_rate": 4.8766447368421054e-06,
+      "loss": 1.2346,
+      "step": 16551
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8950436115264893,
+      "learning_rate": 4.875730994152047e-06,
+      "loss": 1.7846,
+      "step": 16552
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9401330947875977,
+      "learning_rate": 4.874817251461988e-06,
+      "loss": 1.7306,
+      "step": 16553
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5871429443359375,
+      "learning_rate": 4.8739035087719296e-06,
+      "loss": 1.3425,
+      "step": 16554
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8120460510253906,
+      "learning_rate": 4.872989766081872e-06,
+      "loss": 1.6472,
+      "step": 16555
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.484528064727783,
+      "learning_rate": 4.872076023391813e-06,
+      "loss": 1.37,
+      "step": 16556
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6268186569213867,
+      "learning_rate": 4.871162280701755e-06,
+      "loss": 1.2649,
+      "step": 16557
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8476650714874268,
+      "learning_rate": 4.870248538011697e-06,
+      "loss": 1.6379,
+      "step": 16558
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7733817100524902,
+      "learning_rate": 4.869334795321638e-06,
+      "loss": 1.4405,
+      "step": 16559
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2949650287628174,
+      "learning_rate": 4.8684210526315795e-06,
+      "loss": 1.0708,
+      "step": 16560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3533029556274414,
+      "learning_rate": 4.867507309941521e-06,
+      "loss": 1.5549,
+      "step": 16561
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.327012777328491,
+      "learning_rate": 4.866593567251462e-06,
+      "loss": 1.4307,
+      "step": 16562
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4978885650634766,
+      "learning_rate": 4.865679824561404e-06,
+      "loss": 1.4175,
+      "step": 16563
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4828193187713623,
+      "learning_rate": 4.864766081871346e-06,
+      "loss": 1.6384,
+      "step": 16564
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6688289642333984,
+      "learning_rate": 4.863852339181287e-06,
+      "loss": 1.4813,
+      "step": 16565
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.431929111480713,
+      "learning_rate": 4.862938596491229e-06,
+      "loss": 1.3104,
+      "step": 16566
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.445200204849243,
+      "learning_rate": 4.86202485380117e-06,
+      "loss": 1.5448,
+      "step": 16567
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2865519523620605,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 1.3981,
+      "step": 16568
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0812020301818848,
+      "learning_rate": 4.860197368421053e-06,
+      "loss": 1.3846,
+      "step": 16569
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3823485374450684,
+      "learning_rate": 4.859283625730994e-06,
+      "loss": 1.1375,
+      "step": 16570
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3375296592712402,
+      "learning_rate": 4.858369883040936e-06,
+      "loss": 1.3638,
+      "step": 16571
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.723729133605957,
+      "learning_rate": 4.857456140350878e-06,
+      "loss": 1.4695,
+      "step": 16572
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6447298526763916,
+      "learning_rate": 4.856542397660819e-06,
+      "loss": 1.6134,
+      "step": 16573
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.331552267074585,
+      "learning_rate": 4.8556286549707604e-06,
+      "loss": 1.3354,
+      "step": 16574
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3836841583251953,
+      "learning_rate": 4.854714912280702e-06,
+      "loss": 1.3324,
+      "step": 16575
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0103790760040283,
+      "learning_rate": 4.853801169590643e-06,
+      "loss": 1.1277,
+      "step": 16576
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6110475063323975,
+      "learning_rate": 4.852887426900585e-06,
+      "loss": 1.7179,
+      "step": 16577
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.252138614654541,
+      "learning_rate": 4.851973684210527e-06,
+      "loss": 1.2572,
+      "step": 16578
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.269512414932251,
+      "learning_rate": 4.851059941520468e-06,
+      "loss": 1.2481,
+      "step": 16579
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9805006980895996,
+      "learning_rate": 4.8501461988304095e-06,
+      "loss": 1.5992,
+      "step": 16580
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8201379776000977,
+      "learning_rate": 4.849232456140352e-06,
+      "loss": 1.7761,
+      "step": 16581
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5403800010681152,
+      "learning_rate": 4.848318713450293e-06,
+      "loss": 1.4827,
+      "step": 16582
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.166085958480835,
+      "learning_rate": 4.8474049707602345e-06,
+      "loss": 1.3022,
+      "step": 16583
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.228299856185913,
+      "learning_rate": 4.846491228070176e-06,
+      "loss": 1.1416,
+      "step": 16584
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4330058097839355,
+      "learning_rate": 4.845577485380117e-06,
+      "loss": 1.1997,
+      "step": 16585
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.918221950531006,
+      "learning_rate": 4.844663742690059e-06,
+      "loss": 1.7263,
+      "step": 16586
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2981925010681152,
+      "learning_rate": 4.84375e-06,
+      "loss": 1.146,
+      "step": 16587
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1696205139160156,
+      "learning_rate": 4.842836257309942e-06,
+      "loss": 1.1192,
+      "step": 16588
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7188873291015625,
+      "learning_rate": 4.841922514619884e-06,
+      "loss": 1.6085,
+      "step": 16589
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.157329797744751,
+      "learning_rate": 4.841008771929825e-06,
+      "loss": 1.4098,
+      "step": 16590
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.725314140319824,
+      "learning_rate": 4.840095029239766e-06,
+      "loss": 1.694,
+      "step": 16591
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.465007781982422,
+      "learning_rate": 4.839181286549708e-06,
+      "loss": 1.5024,
+      "step": 16592
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.75931453704834,
+      "learning_rate": 4.838267543859649e-06,
+      "loss": 1.5205,
+      "step": 16593
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.790814161300659,
+      "learning_rate": 4.8373538011695905e-06,
+      "loss": 1.5225,
+      "step": 16594
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9829336404800415,
+      "learning_rate": 4.836440058479533e-06,
+      "loss": 1.2286,
+      "step": 16595
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.992356777191162,
+      "learning_rate": 4.835526315789474e-06,
+      "loss": 1.6389,
+      "step": 16596
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.578993558883667,
+      "learning_rate": 4.8346125730994154e-06,
+      "loss": 1.2599,
+      "step": 16597
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1304919719696045,
+      "learning_rate": 4.833698830409358e-06,
+      "loss": 1.623,
+      "step": 16598
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7424845695495605,
+      "learning_rate": 4.832785087719299e-06,
+      "loss": 1.2439,
+      "step": 16599
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.215413808822632,
+      "learning_rate": 4.83187134502924e-06,
+      "loss": 1.3258,
+      "step": 16600
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7711360454559326,
+      "learning_rate": 4.830957602339182e-06,
+      "loss": 1.6268,
+      "step": 16601
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.91098690032959,
+      "learning_rate": 4.830043859649123e-06,
+      "loss": 1.6133,
+      "step": 16602
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.159575939178467,
+      "learning_rate": 4.8291301169590645e-06,
+      "loss": 1.1588,
+      "step": 16603
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3577282428741455,
+      "learning_rate": 4.828216374269007e-06,
+      "loss": 1.3388,
+      "step": 16604
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.796337127685547,
+      "learning_rate": 4.827302631578948e-06,
+      "loss": 2.0611,
+      "step": 16605
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.878922462463379,
+      "learning_rate": 4.8263888888888895e-06,
+      "loss": 1.7253,
+      "step": 16606
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.452101945877075,
+      "learning_rate": 4.825475146198831e-06,
+      "loss": 1.4996,
+      "step": 16607
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.723355531692505,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 1.6289,
+      "step": 16608
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9135926961898804,
+      "learning_rate": 4.823647660818714e-06,
+      "loss": 1.2115,
+      "step": 16609
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.66726016998291,
+      "learning_rate": 4.822733918128655e-06,
+      "loss": 1.4047,
+      "step": 16610
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.336440324783325,
+      "learning_rate": 4.821820175438597e-06,
+      "loss": 1.6074,
+      "step": 16611
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.813880681991577,
+      "learning_rate": 4.820906432748539e-06,
+      "loss": 1.9133,
+      "step": 16612
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.664598226547241,
+      "learning_rate": 4.81999269005848e-06,
+      "loss": 1.3785,
+      "step": 16613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3319714069366455,
+      "learning_rate": 4.819078947368421e-06,
+      "loss": 1.3643,
+      "step": 16614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.206454277038574,
+      "learning_rate": 4.818165204678363e-06,
+      "loss": 1.4943,
+      "step": 16615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.383925676345825,
+      "learning_rate": 4.817251461988304e-06,
+      "loss": 1.2304,
+      "step": 16616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2385334968566895,
+      "learning_rate": 4.8163377192982455e-06,
+      "loss": 1.4478,
+      "step": 16617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.069248676300049,
+      "learning_rate": 4.815423976608188e-06,
+      "loss": 1.3023,
+      "step": 16618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.61867094039917,
+      "learning_rate": 4.814510233918129e-06,
+      "loss": 1.3089,
+      "step": 16619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.153362274169922,
+      "learning_rate": 4.8135964912280704e-06,
+      "loss": 1.0992,
+      "step": 16620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.561183214187622,
+      "learning_rate": 4.812682748538013e-06,
+      "loss": 1.4087,
+      "step": 16621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.103907823562622,
+      "learning_rate": 4.811769005847954e-06,
+      "loss": 1.11,
+      "step": 16622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5288808345794678,
+      "learning_rate": 4.810855263157895e-06,
+      "loss": 1.252,
+      "step": 16623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.370063543319702,
+      "learning_rate": 4.809941520467837e-06,
+      "loss": 1.5613,
+      "step": 16624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6789169311523438,
+      "learning_rate": 4.809027777777778e-06,
+      "loss": 1.49,
+      "step": 16625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4524717330932617,
+      "learning_rate": 4.8081140350877195e-06,
+      "loss": 1.4445,
+      "step": 16626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.373350143432617,
+      "learning_rate": 4.807200292397661e-06,
+      "loss": 1.6134,
+      "step": 16627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0704925060272217,
+      "learning_rate": 4.806286549707603e-06,
+      "loss": 1.6723,
+      "step": 16628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.240107297897339,
+      "learning_rate": 4.8053728070175445e-06,
+      "loss": 1.2327,
+      "step": 16629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4031620025634766,
+      "learning_rate": 4.804459064327486e-06,
+      "loss": 1.506,
+      "step": 16630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.324594259262085,
+      "learning_rate": 4.803545321637427e-06,
+      "loss": 1.7288,
+      "step": 16631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1089608669281006,
+      "learning_rate": 4.802631578947369e-06,
+      "loss": 1.2121,
+      "step": 16632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.42618465423584,
+      "learning_rate": 4.80171783625731e-06,
+      "loss": 1.3013,
+      "step": 16633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.571873664855957,
+      "learning_rate": 4.800804093567251e-06,
+      "loss": 1.4703,
+      "step": 16634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3855879306793213,
+      "learning_rate": 4.799890350877194e-06,
+      "loss": 1.3354,
+      "step": 16635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3405981063842773,
+      "learning_rate": 4.798976608187135e-06,
+      "loss": 1.9222,
+      "step": 16636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3483283519744873,
+      "learning_rate": 4.798062865497076e-06,
+      "loss": 1.3721,
+      "step": 16637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7023444175720215,
+      "learning_rate": 4.797149122807018e-06,
+      "loss": 1.7544,
+      "step": 16638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7797181606292725,
+      "learning_rate": 4.796235380116959e-06,
+      "loss": 1.4117,
+      "step": 16639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9509990215301514,
+      "learning_rate": 4.7953216374269005e-06,
+      "loss": 1.4881,
+      "step": 16640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.532071590423584,
+      "learning_rate": 4.794407894736843e-06,
+      "loss": 1.826,
+      "step": 16641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3901119232177734,
+      "learning_rate": 4.793494152046784e-06,
+      "loss": 1.3983,
+      "step": 16642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1794564723968506,
+      "learning_rate": 4.7925804093567254e-06,
+      "loss": 1.3489,
+      "step": 16643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4832046031951904,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 1.5343,
+      "step": 16644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.259288787841797,
+      "learning_rate": 4.790752923976609e-06,
+      "loss": 1.4398,
+      "step": 16645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.314101457595825,
+      "learning_rate": 4.78983918128655e-06,
+      "loss": 1.2591,
+      "step": 16646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.554124116897583,
+      "learning_rate": 4.788925438596492e-06,
+      "loss": 1.2478,
+      "step": 16647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2972562313079834,
+      "learning_rate": 4.788011695906433e-06,
+      "loss": 1.3829,
+      "step": 16648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4774510860443115,
+      "learning_rate": 4.7870979532163745e-06,
+      "loss": 1.6026,
+      "step": 16649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2249763011932373,
+      "learning_rate": 4.786184210526316e-06,
+      "loss": 1.4031,
+      "step": 16650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3211982250213623,
+      "learning_rate": 4.785270467836258e-06,
+      "loss": 1.5305,
+      "step": 16651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0821306705474854,
+      "learning_rate": 4.7843567251461995e-06,
+      "loss": 1.5569,
+      "step": 16652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6396288871765137,
+      "learning_rate": 4.783442982456141e-06,
+      "loss": 1.7729,
+      "step": 16653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4047436714172363,
+      "learning_rate": 4.782529239766082e-06,
+      "loss": 1.25,
+      "step": 16654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.047569751739502,
+      "learning_rate": 4.781615497076024e-06,
+      "loss": 1.8739,
+      "step": 16655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1239428520202637,
+      "learning_rate": 4.780701754385965e-06,
+      "loss": 1.1762,
+      "step": 16656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.356166124343872,
+      "learning_rate": 4.779788011695906e-06,
+      "loss": 1.4698,
+      "step": 16657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3744077682495117,
+      "learning_rate": 4.778874269005848e-06,
+      "loss": 1.5637,
+      "step": 16658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5547034740448,
+      "learning_rate": 4.77796052631579e-06,
+      "loss": 1.3853,
+      "step": 16659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2085213661193848,
+      "learning_rate": 4.777046783625731e-06,
+      "loss": 1.4841,
+      "step": 16660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3777143955230713,
+      "learning_rate": 4.776133040935673e-06,
+      "loss": 1.4432,
+      "step": 16661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3324711322784424,
+      "learning_rate": 4.775219298245615e-06,
+      "loss": 1.226,
+      "step": 16662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5639426708221436,
+      "learning_rate": 4.774305555555556e-06,
+      "loss": 1.4045,
+      "step": 16663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6265740394592285,
+      "learning_rate": 4.773391812865498e-06,
+      "loss": 1.6517,
+      "step": 16664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2668399810791016,
+      "learning_rate": 4.772478070175439e-06,
+      "loss": 1.4698,
+      "step": 16665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.014207363128662,
+      "learning_rate": 4.7715643274853804e-06,
+      "loss": 1.7824,
+      "step": 16666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5753772258758545,
+      "learning_rate": 4.770650584795322e-06,
+      "loss": 1.434,
+      "step": 16667
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.686840534210205,
+      "learning_rate": 4.769736842105264e-06,
+      "loss": 1.5883,
+      "step": 16668
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4715628623962402,
+      "learning_rate": 4.768823099415205e-06,
+      "loss": 1.3744,
+      "step": 16669
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2141175270080566,
+      "learning_rate": 4.767909356725147e-06,
+      "loss": 1.2209,
+      "step": 16670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8439457416534424,
+      "learning_rate": 4.766995614035088e-06,
+      "loss": 1.4204,
+      "step": 16671
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.165968418121338,
+      "learning_rate": 4.7660818713450295e-06,
+      "loss": 1.2816,
+      "step": 16672
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.514481544494629,
+      "learning_rate": 4.765168128654971e-06,
+      "loss": 1.5608,
+      "step": 16673
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.30839204788208,
+      "learning_rate": 4.764254385964912e-06,
+      "loss": 1.6213,
+      "step": 16674
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5683350563049316,
+      "learning_rate": 4.7633406432748545e-06,
+      "loss": 1.3923,
+      "step": 16675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.148228645324707,
+      "learning_rate": 4.762426900584796e-06,
+      "loss": 1.2665,
+      "step": 16676
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0681679248809814,
+      "learning_rate": 4.761513157894737e-06,
+      "loss": 1.1281,
+      "step": 16677
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.613631010055542,
+      "learning_rate": 4.760599415204679e-06,
+      "loss": 1.5847,
+      "step": 16678
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.894727110862732,
+      "learning_rate": 4.75968567251462e-06,
+      "loss": 0.9911,
+      "step": 16679
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3904716968536377,
+      "learning_rate": 4.758771929824561e-06,
+      "loss": 1.3158,
+      "step": 16680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6023685932159424,
+      "learning_rate": 4.757858187134503e-06,
+      "loss": 1.4822,
+      "step": 16681
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.102579116821289,
+      "learning_rate": 4.756944444444445e-06,
+      "loss": 1.3282,
+      "step": 16682
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.630309820175171,
+      "learning_rate": 4.756030701754386e-06,
+      "loss": 1.5788,
+      "step": 16683
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.502737283706665,
+      "learning_rate": 4.755116959064328e-06,
+      "loss": 1.311,
+      "step": 16684
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.671769857406616,
+      "learning_rate": 4.75420321637427e-06,
+      "loss": 1.496,
+      "step": 16685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.36044979095459,
+      "learning_rate": 4.753289473684211e-06,
+      "loss": 1.4453,
+      "step": 16686
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7952847480773926,
+      "learning_rate": 4.752375730994153e-06,
+      "loss": 1.5085,
+      "step": 16687
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8839242458343506,
+      "learning_rate": 4.751461988304094e-06,
+      "loss": 1.5703,
+      "step": 16688
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.701338291168213,
+      "learning_rate": 4.7505482456140354e-06,
+      "loss": 1.7172,
+      "step": 16689
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5350961685180664,
+      "learning_rate": 4.749634502923977e-06,
+      "loss": 1.6004,
+      "step": 16690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3695924282073975,
+      "learning_rate": 4.748720760233918e-06,
+      "loss": 1.6307,
+      "step": 16691
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.467907428741455,
+      "learning_rate": 4.74780701754386e-06,
+      "loss": 1.66,
+      "step": 16692
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5425984859466553,
+      "learning_rate": 4.746893274853802e-06,
+      "loss": 1.6555,
+      "step": 16693
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0567712783813477,
+      "learning_rate": 4.745979532163743e-06,
+      "loss": 1.4173,
+      "step": 16694
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4498515129089355,
+      "learning_rate": 4.7450657894736845e-06,
+      "loss": 1.5922,
+      "step": 16695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4558820724487305,
+      "learning_rate": 4.744152046783626e-06,
+      "loss": 1.367,
+      "step": 16696
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.020081043243408,
+      "learning_rate": 4.743238304093567e-06,
+      "loss": 1.8606,
+      "step": 16697
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.657026767730713,
+      "learning_rate": 4.742324561403509e-06,
+      "loss": 1.5133,
+      "step": 16698
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2097671031951904,
+      "learning_rate": 4.741410818713451e-06,
+      "loss": 1.2582,
+      "step": 16699
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7582061290740967,
+      "learning_rate": 4.740497076023392e-06,
+      "loss": 1.5609,
+      "step": 16700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.022845983505249,
+      "learning_rate": 4.739583333333334e-06,
+      "loss": 1.0299,
+      "step": 16701
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.899249315261841,
+      "learning_rate": 4.738669590643275e-06,
+      "loss": 1.758,
+      "step": 16702
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.594310998916626,
+      "learning_rate": 4.737755847953216e-06,
+      "loss": 1.2963,
+      "step": 16703
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9511137008666992,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 1.2322,
+      "step": 16704
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.198064088821411,
+      "learning_rate": 4.7359283625731e-06,
+      "loss": 1.2823,
+      "step": 16705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.65254807472229,
+      "learning_rate": 4.735014619883041e-06,
+      "loss": 1.5318,
+      "step": 16706
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8258140087127686,
+      "learning_rate": 4.734100877192983e-06,
+      "loss": 1.8744,
+      "step": 16707
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2065744400024414,
+      "learning_rate": 4.733187134502925e-06,
+      "loss": 1.8984,
+      "step": 16708
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6823017597198486,
+      "learning_rate": 4.732273391812866e-06,
+      "loss": 1.3533,
+      "step": 16709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.894838333129883,
+      "learning_rate": 4.731359649122808e-06,
+      "loss": 1.4956,
+      "step": 16710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.506938934326172,
+      "learning_rate": 4.730445906432749e-06,
+      "loss": 1.6161,
+      "step": 16711
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.842928647994995,
+      "learning_rate": 4.7295321637426904e-06,
+      "loss": 1.7025,
+      "step": 16712
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.20298433303833,
+      "learning_rate": 4.728618421052632e-06,
+      "loss": 1.4219,
+      "step": 16713
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0287508964538574,
+      "learning_rate": 4.727704678362573e-06,
+      "loss": 1.1986,
+      "step": 16714
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2552237510681152,
+      "learning_rate": 4.726790935672515e-06,
+      "loss": 1.299,
+      "step": 16715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.591907024383545,
+      "learning_rate": 4.725877192982457e-06,
+      "loss": 1.3837,
+      "step": 16716
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1734843254089355,
+      "learning_rate": 4.724963450292398e-06,
+      "loss": 1.3564,
+      "step": 16717
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3045129776000977,
+      "learning_rate": 4.7240497076023395e-06,
+      "loss": 1.3038,
+      "step": 16718
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6671876907348633,
+      "learning_rate": 4.723135964912281e-06,
+      "loss": 1.4819,
+      "step": 16719
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.733259439468384,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 1.5169,
+      "step": 16720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1087470054626465,
+      "learning_rate": 4.721308479532164e-06,
+      "loss": 1.1307,
+      "step": 16721
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4039151668548584,
+      "learning_rate": 4.720394736842106e-06,
+      "loss": 1.6279,
+      "step": 16722
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7128374576568604,
+      "learning_rate": 4.719480994152047e-06,
+      "loss": 1.3644,
+      "step": 16723
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.865355968475342,
+      "learning_rate": 4.718567251461989e-06,
+      "loss": 1.4052,
+      "step": 16724
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5354745388031006,
+      "learning_rate": 4.71765350877193e-06,
+      "loss": 1.5077,
+      "step": 16725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6481618881225586,
+      "learning_rate": 4.716739766081872e-06,
+      "loss": 1.4218,
+      "step": 16726
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5636932849884033,
+      "learning_rate": 4.715826023391814e-06,
+      "loss": 1.6394,
+      "step": 16727
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3744049072265625,
+      "learning_rate": 4.714912280701755e-06,
+      "loss": 1.2289,
+      "step": 16728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7569074630737305,
+      "learning_rate": 4.713998538011696e-06,
+      "loss": 1.6426,
+      "step": 16729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4793412685394287,
+      "learning_rate": 4.713084795321638e-06,
+      "loss": 1.8528,
+      "step": 16730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4149515628814697,
+      "learning_rate": 4.712171052631579e-06,
+      "loss": 1.5329,
+      "step": 16731
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6285006999969482,
+      "learning_rate": 4.711257309941521e-06,
+      "loss": 2.1015,
+      "step": 16732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.804551839828491,
+      "learning_rate": 4.710343567251463e-06,
+      "loss": 1.8144,
+      "step": 16733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.744145393371582,
+      "learning_rate": 4.709429824561404e-06,
+      "loss": 1.5725,
+      "step": 16734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7991113662719727,
+      "learning_rate": 4.7085160818713454e-06,
+      "loss": 1.476,
+      "step": 16735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7755126953125,
+      "learning_rate": 4.707602339181287e-06,
+      "loss": 1.5959,
+      "step": 16736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.274691104888916,
+      "learning_rate": 4.706688596491228e-06,
+      "loss": 1.3281,
+      "step": 16737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.320025682449341,
+      "learning_rate": 4.7057748538011696e-06,
+      "loss": 1.252,
+      "step": 16738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.387068271636963,
+      "learning_rate": 4.704861111111112e-06,
+      "loss": 1.388,
+      "step": 16739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.217802047729492,
+      "learning_rate": 4.703947368421053e-06,
+      "loss": 1.3724,
+      "step": 16740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8743185997009277,
+      "learning_rate": 4.7030336257309945e-06,
+      "loss": 2.1019,
+      "step": 16741
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0871617794036865,
+      "learning_rate": 4.702119883040936e-06,
+      "loss": 1.8117,
+      "step": 16742
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3511857986450195,
+      "learning_rate": 4.701206140350877e-06,
+      "loss": 1.3348,
+      "step": 16743
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.133485794067383,
+      "learning_rate": 4.700292397660819e-06,
+      "loss": 1.1512,
+      "step": 16744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3663387298583984,
+      "learning_rate": 4.69937865497076e-06,
+      "loss": 1.3307,
+      "step": 16745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3773767948150635,
+      "learning_rate": 4.698464912280702e-06,
+      "loss": 1.1591,
+      "step": 16746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0143818855285645,
+      "learning_rate": 4.697551169590644e-06,
+      "loss": 1.6524,
+      "step": 16747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.560183048248291,
+      "learning_rate": 4.696637426900586e-06,
+      "loss": 1.5732,
+      "step": 16748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7783970832824707,
+      "learning_rate": 4.695723684210527e-06,
+      "loss": 1.3706,
+      "step": 16749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4151804447174072,
+      "learning_rate": 4.694809941520469e-06,
+      "loss": 1.2157,
+      "step": 16750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7650766372680664,
+      "learning_rate": 4.69389619883041e-06,
+      "loss": 1.4705,
+      "step": 16751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.927549362182617,
+      "learning_rate": 4.692982456140351e-06,
+      "loss": 1.4908,
+      "step": 16752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9711551666259766,
+      "learning_rate": 4.692068713450293e-06,
+      "loss": 1.5299,
+      "step": 16753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6431851387023926,
+      "learning_rate": 4.691154970760234e-06,
+      "loss": 1.5301,
+      "step": 16754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0631227493286133,
+      "learning_rate": 4.690241228070176e-06,
+      "loss": 1.8369,
+      "step": 16755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.625000476837158,
+      "learning_rate": 4.689327485380118e-06,
+      "loss": 1.734,
+      "step": 16756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.631627321243286,
+      "learning_rate": 4.688413742690059e-06,
+      "loss": 1.4264,
+      "step": 16757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5632431507110596,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 1.5324,
+      "step": 16758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.85829758644104,
+      "learning_rate": 4.686586257309942e-06,
+      "loss": 1.7851,
+      "step": 16759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.146799087524414,
+      "learning_rate": 4.685672514619883e-06,
+      "loss": 1.3934,
+      "step": 16760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.271162509918213,
+      "learning_rate": 4.6847587719298246e-06,
+      "loss": 1.4195,
+      "step": 16761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.706102132797241,
+      "learning_rate": 4.683845029239767e-06,
+      "loss": 1.6028,
+      "step": 16762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.533269166946411,
+      "learning_rate": 4.682931286549708e-06,
+      "loss": 1.4989,
+      "step": 16763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.615614175796509,
+      "learning_rate": 4.6820175438596495e-06,
+      "loss": 1.5648,
+      "step": 16764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.34490966796875,
+      "learning_rate": 4.681103801169591e-06,
+      "loss": 1.6582,
+      "step": 16765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.402688980102539,
+      "learning_rate": 4.680190058479532e-06,
+      "loss": 1.3024,
+      "step": 16766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.673025131225586,
+      "learning_rate": 4.679276315789474e-06,
+      "loss": 1.7321,
+      "step": 16767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.762723207473755,
+      "learning_rate": 4.678362573099415e-06,
+      "loss": 1.6857,
+      "step": 16768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8868203163146973,
+      "learning_rate": 4.677448830409357e-06,
+      "loss": 1.666,
+      "step": 16769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4699878692626953,
+      "learning_rate": 4.676535087719299e-06,
+      "loss": 1.4926,
+      "step": 16770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.867480516433716,
+      "learning_rate": 4.67562134502924e-06,
+      "loss": 1.617,
+      "step": 16771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.402191162109375,
+      "learning_rate": 4.674707602339182e-06,
+      "loss": 1.3784,
+      "step": 16772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.571722984313965,
+      "learning_rate": 4.673793859649124e-06,
+      "loss": 1.4672,
+      "step": 16773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.667327404022217,
+      "learning_rate": 4.672880116959065e-06,
+      "loss": 1.5735,
+      "step": 16774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.634242296218872,
+      "learning_rate": 4.671966374269006e-06,
+      "loss": 1.3676,
+      "step": 16775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6817259788513184,
+      "learning_rate": 4.671052631578948e-06,
+      "loss": 1.3816,
+      "step": 16776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7446446418762207,
+      "learning_rate": 4.670138888888889e-06,
+      "loss": 1.419,
+      "step": 16777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.93896746635437,
+      "learning_rate": 4.6692251461988305e-06,
+      "loss": 1.5418,
+      "step": 16778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2847864627838135,
+      "learning_rate": 4.668311403508773e-06,
+      "loss": 1.3637,
+      "step": 16779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.735891342163086,
+      "learning_rate": 4.667397660818714e-06,
+      "loss": 1.4785,
+      "step": 16780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5135254859924316,
+      "learning_rate": 4.6664839181286554e-06,
+      "loss": 1.338,
+      "step": 16781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7693936824798584,
+      "learning_rate": 4.665570175438597e-06,
+      "loss": 1.7213,
+      "step": 16782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0521137714385986,
+      "learning_rate": 4.664656432748538e-06,
+      "loss": 1.6565,
+      "step": 16783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1383306980133057,
+      "learning_rate": 4.6637426900584796e-06,
+      "loss": 1.4812,
+      "step": 16784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5772526264190674,
+      "learning_rate": 4.662828947368421e-06,
+      "loss": 1.4014,
+      "step": 16785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.838977098464966,
+      "learning_rate": 4.661915204678363e-06,
+      "loss": 1.3582,
+      "step": 16786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7476470470428467,
+      "learning_rate": 4.6610014619883045e-06,
+      "loss": 1.6329,
+      "step": 16787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8733279705047607,
+      "learning_rate": 4.660087719298246e-06,
+      "loss": 2.1405,
+      "step": 16788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8284876346588135,
+      "learning_rate": 4.659173976608187e-06,
+      "loss": 1.8638,
+      "step": 16789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5297396183013916,
+      "learning_rate": 4.658260233918129e-06,
+      "loss": 1.3103,
+      "step": 16790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.678334951400757,
+      "learning_rate": 4.657346491228071e-06,
+      "loss": 1.5285,
+      "step": 16791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.823085308074951,
+      "learning_rate": 4.656432748538012e-06,
+      "loss": 1.7218,
+      "step": 16792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2735671997070312,
+      "learning_rate": 4.655519005847954e-06,
+      "loss": 1.3603,
+      "step": 16793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2480247020721436,
+      "learning_rate": 4.654605263157895e-06,
+      "loss": 1.1411,
+      "step": 16794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3207626342773438,
+      "learning_rate": 4.653691520467836e-06,
+      "loss": 1.4704,
+      "step": 16795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.832019805908203,
+      "learning_rate": 4.652777777777779e-06,
+      "loss": 1.5838,
+      "step": 16796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.457794189453125,
+      "learning_rate": 4.65186403508772e-06,
+      "loss": 1.3224,
+      "step": 16797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.31542706489563,
+      "learning_rate": 4.650950292397661e-06,
+      "loss": 1.28,
+      "step": 16798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5137085914611816,
+      "learning_rate": 4.650036549707603e-06,
+      "loss": 1.4703,
+      "step": 16799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5394442081451416,
+      "learning_rate": 4.649122807017544e-06,
+      "loss": 1.6529,
+      "step": 16800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7012317180633545,
+      "learning_rate": 4.6482090643274855e-06,
+      "loss": 1.543,
+      "step": 16801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4682490825653076,
+      "learning_rate": 4.647295321637427e-06,
+      "loss": 1.6101,
+      "step": 16802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.577671527862549,
+      "learning_rate": 4.646381578947369e-06,
+      "loss": 1.12,
+      "step": 16803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1720871925354004,
+      "learning_rate": 4.6454678362573104e-06,
+      "loss": 1.3652,
+      "step": 16804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1299920082092285,
+      "learning_rate": 4.644554093567252e-06,
+      "loss": 1.3994,
+      "step": 16805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.89255690574646,
+      "learning_rate": 4.643640350877193e-06,
+      "loss": 1.8292,
+      "step": 16806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.529170274734497,
+      "learning_rate": 4.6427266081871346e-06,
+      "loss": 1.3876,
+      "step": 16807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4877192974090576,
+      "learning_rate": 4.641812865497076e-06,
+      "loss": 1.4546,
+      "step": 16808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7198855876922607,
+      "learning_rate": 4.640899122807017e-06,
+      "loss": 1.6052,
+      "step": 16809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0425541400909424,
+      "learning_rate": 4.6399853801169595e-06,
+      "loss": 1.786,
+      "step": 16810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8814806938171387,
+      "learning_rate": 4.639071637426901e-06,
+      "loss": 0.9274,
+      "step": 16811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.886077404022217,
+      "learning_rate": 4.638157894736843e-06,
+      "loss": 1.6048,
+      "step": 16812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6118218898773193,
+      "learning_rate": 4.6372441520467845e-06,
+      "loss": 1.7708,
+      "step": 16813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9671852588653564,
+      "learning_rate": 4.636330409356726e-06,
+      "loss": 1.3583,
+      "step": 16814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5784263610839844,
+      "learning_rate": 4.635416666666667e-06,
+      "loss": 1.5873,
+      "step": 16815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3134005069732666,
+      "learning_rate": 4.634502923976609e-06,
+      "loss": 1.4055,
+      "step": 16816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6457271575927734,
+      "learning_rate": 4.63358918128655e-06,
+      "loss": 1.5958,
+      "step": 16817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.857553005218506,
+      "learning_rate": 4.632675438596491e-06,
+      "loss": 1.5662,
+      "step": 16818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2411673069000244,
+      "learning_rate": 4.631761695906434e-06,
+      "loss": 1.5387,
+      "step": 16819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.281553268432617,
+      "learning_rate": 4.630847953216375e-06,
+      "loss": 1.2612,
+      "step": 16820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.626627206802368,
+      "learning_rate": 4.629934210526316e-06,
+      "loss": 1.6999,
+      "step": 16821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.76479434967041,
+      "learning_rate": 4.629020467836258e-06,
+      "loss": 1.3015,
+      "step": 16822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.861868381500244,
+      "learning_rate": 4.628106725146199e-06,
+      "loss": 1.6911,
+      "step": 16823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.321268320083618,
+      "learning_rate": 4.6271929824561405e-06,
+      "loss": 1.3063,
+      "step": 16824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4603488445281982,
+      "learning_rate": 4.626279239766082e-06,
+      "loss": 1.5366,
+      "step": 16825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.515285015106201,
+      "learning_rate": 4.625365497076024e-06,
+      "loss": 1.4608,
+      "step": 16826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3070220947265625,
+      "learning_rate": 4.6244517543859654e-06,
+      "loss": 1.165,
+      "step": 16827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3844408988952637,
+      "learning_rate": 4.623538011695907e-06,
+      "loss": 1.3304,
+      "step": 16828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.14581561088562,
+      "learning_rate": 4.622624269005848e-06,
+      "loss": 1.8757,
+      "step": 16829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.271369457244873,
+      "learning_rate": 4.6217105263157896e-06,
+      "loss": 1.322,
+      "step": 16830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2408206462860107,
+      "learning_rate": 4.620796783625731e-06,
+      "loss": 1.3711,
+      "step": 16831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3937084674835205,
+      "learning_rate": 4.619883040935672e-06,
+      "loss": 1.6801,
+      "step": 16832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.133756637573242,
+      "learning_rate": 4.6189692982456145e-06,
+      "loss": 1.3632,
+      "step": 16833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.252448558807373,
+      "learning_rate": 4.618055555555556e-06,
+      "loss": 1.1497,
+      "step": 16834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.510262966156006,
+      "learning_rate": 4.617141812865497e-06,
+      "loss": 1.8692,
+      "step": 16835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.815000534057617,
+      "learning_rate": 4.6162280701754395e-06,
+      "loss": 1.6009,
+      "step": 16836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5379698276519775,
+      "learning_rate": 4.615314327485381e-06,
+      "loss": 1.4227,
+      "step": 16837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7042899131774902,
+      "learning_rate": 4.614400584795322e-06,
+      "loss": 1.5983,
+      "step": 16838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.373756170272827,
+      "learning_rate": 4.613486842105264e-06,
+      "loss": 1.3736,
+      "step": 16839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.736527919769287,
+      "learning_rate": 4.612573099415205e-06,
+      "loss": 1.3079,
+      "step": 16840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9789907932281494,
+      "learning_rate": 4.611659356725146e-06,
+      "loss": 1.0837,
+      "step": 16841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4889261722564697,
+      "learning_rate": 4.610745614035088e-06,
+      "loss": 1.4817,
+      "step": 16842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.024725914001465,
+      "learning_rate": 4.60983187134503e-06,
+      "loss": 1.0064,
+      "step": 16843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0671427249908447,
+      "learning_rate": 4.608918128654971e-06,
+      "loss": 1.2734,
+      "step": 16844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2279293537139893,
+      "learning_rate": 4.608004385964913e-06,
+      "loss": 1.1974,
+      "step": 16845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6011102199554443,
+      "learning_rate": 4.607090643274854e-06,
+      "loss": 1.4774,
+      "step": 16846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.357415199279785,
+      "learning_rate": 4.6061769005847955e-06,
+      "loss": 1.3119,
+      "step": 16847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0971720218658447,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 1.1432,
+      "step": 16848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.004260778427124,
+      "learning_rate": 4.604349415204678e-06,
+      "loss": 2.0007,
+      "step": 16849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4220879077911377,
+      "learning_rate": 4.6034356725146204e-06,
+      "loss": 1.1663,
+      "step": 16850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.162358522415161,
+      "learning_rate": 4.602521929824562e-06,
+      "loss": 1.3144,
+      "step": 16851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1872050762176514,
+      "learning_rate": 4.601608187134503e-06,
+      "loss": 1.1039,
+      "step": 16852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6026668548583984,
+      "learning_rate": 4.6006944444444446e-06,
+      "loss": 1.4614,
+      "step": 16853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7856814861297607,
+      "learning_rate": 4.599780701754386e-06,
+      "loss": 1.9286,
+      "step": 16854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.081970453262329,
+      "learning_rate": 4.598866959064328e-06,
+      "loss": 1.0834,
+      "step": 16855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.311403274536133,
+      "learning_rate": 4.5979532163742695e-06,
+      "loss": 1.1997,
+      "step": 16856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.370748996734619,
+      "learning_rate": 4.597039473684211e-06,
+      "loss": 1.5099,
+      "step": 16857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.035179376602173,
+      "learning_rate": 4.596125730994152e-06,
+      "loss": 1.2797,
+      "step": 16858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0984199047088623,
+      "learning_rate": 4.5952119883040945e-06,
+      "loss": 1.3424,
+      "step": 16859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4799299240112305,
+      "learning_rate": 4.594298245614036e-06,
+      "loss": 1.6117,
+      "step": 16860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5619311332702637,
+      "learning_rate": 4.593384502923977e-06,
+      "loss": 1.5322,
+      "step": 16861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.420614719390869,
+      "learning_rate": 4.592470760233919e-06,
+      "loss": 1.4004,
+      "step": 16862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8164076805114746,
+      "learning_rate": 4.59155701754386e-06,
+      "loss": 1.555,
+      "step": 16863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.573770523071289,
+      "learning_rate": 4.590643274853801e-06,
+      "loss": 1.5174,
+      "step": 16864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.950059652328491,
+      "learning_rate": 4.589729532163743e-06,
+      "loss": 1.5967,
+      "step": 16865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.411722183227539,
+      "learning_rate": 4.588815789473685e-06,
+      "loss": 1.5106,
+      "step": 16866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.671870470046997,
+      "learning_rate": 4.587902046783626e-06,
+      "loss": 1.4706,
+      "step": 16867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4380221366882324,
+      "learning_rate": 4.586988304093568e-06,
+      "loss": 1.5159,
+      "step": 16868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7561111450195312,
+      "learning_rate": 4.586074561403509e-06,
+      "loss": 1.7731,
+      "step": 16869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.219087839126587,
+      "learning_rate": 4.5851608187134505e-06,
+      "loss": 1.3481,
+      "step": 16870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0021233558654785,
+      "learning_rate": 4.584247076023392e-06,
+      "loss": 1.6767,
+      "step": 16871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.27333664894104,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 1.7352,
+      "step": 16872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.204369068145752,
+      "learning_rate": 4.5824195906432754e-06,
+      "loss": 1.301,
+      "step": 16873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.638817548751831,
+      "learning_rate": 4.581505847953217e-06,
+      "loss": 1.394,
+      "step": 16874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.149085283279419,
+      "learning_rate": 4.580592105263158e-06,
+      "loss": 1.6935,
+      "step": 16875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9389593601226807,
+      "learning_rate": 4.5796783625731e-06,
+      "loss": 0.9743,
+      "step": 16876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.632601737976074,
+      "learning_rate": 4.578764619883042e-06,
+      "loss": 1.6355,
+      "step": 16877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.647906541824341,
+      "learning_rate": 4.577850877192983e-06,
+      "loss": 1.4906,
+      "step": 16878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2281155586242676,
+      "learning_rate": 4.5769371345029245e-06,
+      "loss": 1.4488,
+      "step": 16879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.238247871398926,
+      "learning_rate": 4.576023391812866e-06,
+      "loss": 1.4719,
+      "step": 16880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.339409589767456,
+      "learning_rate": 4.575109649122807e-06,
+      "loss": 1.653,
+      "step": 16881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4694507122039795,
+      "learning_rate": 4.574195906432749e-06,
+      "loss": 1.5907,
+      "step": 16882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.788522720336914,
+      "learning_rate": 4.573282163742691e-06,
+      "loss": 1.5082,
+      "step": 16883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2521626949310303,
+      "learning_rate": 4.572368421052632e-06,
+      "loss": 1.5056,
+      "step": 16884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5599617958068848,
+      "learning_rate": 4.571454678362574e-06,
+      "loss": 1.4062,
+      "step": 16885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8852033615112305,
+      "learning_rate": 4.570540935672515e-06,
+      "loss": 1.8591,
+      "step": 16886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9279093742370605,
+      "learning_rate": 4.569627192982456e-06,
+      "loss": 1.6265,
+      "step": 16887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6296939849853516,
+      "learning_rate": 4.568713450292398e-06,
+      "loss": 1.6306,
+      "step": 16888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1215627193450928,
+      "learning_rate": 4.567799707602339e-06,
+      "loss": 1.2892,
+      "step": 16889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7865920066833496,
+      "learning_rate": 4.566885964912281e-06,
+      "loss": 1.4455,
+      "step": 16890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.059269428253174,
+      "learning_rate": 4.565972222222223e-06,
+      "loss": 1.291,
+      "step": 16891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.577993392944336,
+      "learning_rate": 4.565058479532164e-06,
+      "loss": 1.6221,
+      "step": 16892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.435340404510498,
+      "learning_rate": 4.5641447368421055e-06,
+      "loss": 1.3368,
+      "step": 16893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4917750358581543,
+      "learning_rate": 4.563230994152047e-06,
+      "loss": 1.327,
+      "step": 16894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2345330715179443,
+      "learning_rate": 4.562317251461988e-06,
+      "loss": 1.3286,
+      "step": 16895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.561511516571045,
+      "learning_rate": 4.56140350877193e-06,
+      "loss": 1.5985,
+      "step": 16896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.456627368927002,
+      "learning_rate": 4.560489766081872e-06,
+      "loss": 1.4004,
+      "step": 16897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.270507574081421,
+      "learning_rate": 4.559576023391813e-06,
+      "loss": 1.4068,
+      "step": 16898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.314020872116089,
+      "learning_rate": 4.558662280701755e-06,
+      "loss": 1.3176,
+      "step": 16899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9501004219055176,
+      "learning_rate": 4.557748538011697e-06,
+      "loss": 1.5213,
+      "step": 16900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4112040996551514,
+      "learning_rate": 4.556834795321638e-06,
+      "loss": 1.4623,
+      "step": 16901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6153600215911865,
+      "learning_rate": 4.5559210526315795e-06,
+      "loss": 1.6236,
+      "step": 16902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1107239723205566,
+      "learning_rate": 4.555007309941521e-06,
+      "loss": 1.1984,
+      "step": 16903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5233285427093506,
+      "learning_rate": 4.554093567251462e-06,
+      "loss": 1.3958,
+      "step": 16904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4842369556427,
+      "learning_rate": 4.553179824561404e-06,
+      "loss": 1.3492,
+      "step": 16905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.817962646484375,
+      "learning_rate": 4.552266081871345e-06,
+      "loss": 1.7601,
+      "step": 16906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.31459903717041,
+      "learning_rate": 4.551352339181287e-06,
+      "loss": 1.146,
+      "step": 16907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.076996326446533,
+      "learning_rate": 4.550438596491229e-06,
+      "loss": 1.2122,
+      "step": 16908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.304870128631592,
+      "learning_rate": 4.54952485380117e-06,
+      "loss": 1.2817,
+      "step": 16909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.372570037841797,
+      "learning_rate": 4.548611111111111e-06,
+      "loss": 1.2772,
+      "step": 16910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7233846187591553,
+      "learning_rate": 4.547697368421053e-06,
+      "loss": 1.8449,
+      "step": 16911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4577624797821045,
+      "learning_rate": 4.546783625730994e-06,
+      "loss": 1.3985,
+      "step": 16912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.340362310409546,
+      "learning_rate": 4.5458698830409355e-06,
+      "loss": 1.3892,
+      "step": 16913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1735446453094482,
+      "learning_rate": 4.544956140350878e-06,
+      "loss": 1.2915,
+      "step": 16914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8701590299606323,
+      "learning_rate": 4.544042397660819e-06,
+      "loss": 1.1011,
+      "step": 16915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6682825088500977,
+      "learning_rate": 4.5431286549707605e-06,
+      "loss": 1.7438,
+      "step": 16916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.156013250350952,
+      "learning_rate": 4.542214912280702e-06,
+      "loss": 1.125,
+      "step": 16917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.931952476501465,
+      "learning_rate": 4.541301169590643e-06,
+      "loss": 1.1645,
+      "step": 16918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3472604751586914,
+      "learning_rate": 4.5403874269005854e-06,
+      "loss": 1.471,
+      "step": 16919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.494462013244629,
+      "learning_rate": 4.539473684210527e-06,
+      "loss": 1.7145,
+      "step": 16920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4065377712249756,
+      "learning_rate": 4.538559941520468e-06,
+      "loss": 1.4468,
+      "step": 16921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4444072246551514,
+      "learning_rate": 4.5376461988304096e-06,
+      "loss": 1.2236,
+      "step": 16922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.804696798324585,
+      "learning_rate": 4.536732456140352e-06,
+      "loss": 1.5741,
+      "step": 16923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.691572666168213,
+      "learning_rate": 4.535818713450293e-06,
+      "loss": 1.5717,
+      "step": 16924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.563202142715454,
+      "learning_rate": 4.5349049707602345e-06,
+      "loss": 1.6918,
+      "step": 16925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.681349039077759,
+      "learning_rate": 4.533991228070176e-06,
+      "loss": 1.4928,
+      "step": 16926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8583614826202393,
+      "learning_rate": 4.533077485380117e-06,
+      "loss": 1.8536,
+      "step": 16927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8421244621276855,
+      "learning_rate": 4.532163742690059e-06,
+      "loss": 1.6048,
+      "step": 16928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.428175210952759,
+      "learning_rate": 4.53125e-06,
+      "loss": 1.3842,
+      "step": 16929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.63179874420166,
+      "learning_rate": 4.530336257309942e-06,
+      "loss": 1.6136,
+      "step": 16930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5952086448669434,
+      "learning_rate": 4.529422514619884e-06,
+      "loss": 1.6251,
+      "step": 16931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6156938076019287,
+      "learning_rate": 4.528508771929825e-06,
+      "loss": 1.8411,
+      "step": 16932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6721885204315186,
+      "learning_rate": 4.527595029239766e-06,
+      "loss": 1.763,
+      "step": 16933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5279550552368164,
+      "learning_rate": 4.526681286549708e-06,
+      "loss": 1.4591,
+      "step": 16934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2398879528045654,
+      "learning_rate": 4.525767543859649e-06,
+      "loss": 1.5588,
+      "step": 16935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4004878997802734,
+      "learning_rate": 4.5248538011695905e-06,
+      "loss": 1.2367,
+      "step": 16936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8504910469055176,
+      "learning_rate": 4.523940058479533e-06,
+      "loss": 1.4843,
+      "step": 16937
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.55505633354187,
+      "learning_rate": 4.523026315789474e-06,
+      "loss": 1.5236,
+      "step": 16938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9028775691986084,
+      "learning_rate": 4.5221125730994155e-06,
+      "loss": 1.7371,
+      "step": 16939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.347717046737671,
+      "learning_rate": 4.521198830409358e-06,
+      "loss": 1.4816,
+      "step": 16940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3345115184783936,
+      "learning_rate": 4.520285087719299e-06,
+      "loss": 1.6162,
+      "step": 16941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8256428241729736,
+      "learning_rate": 4.5193713450292404e-06,
+      "loss": 1.6775,
+      "step": 16942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.939896821975708,
+      "learning_rate": 4.518457602339182e-06,
+      "loss": 1.4659,
+      "step": 16943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.524129629135132,
+      "learning_rate": 4.517543859649123e-06,
+      "loss": 1.3425,
+      "step": 16944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1960155963897705,
+      "learning_rate": 4.5166301169590646e-06,
+      "loss": 1.2317,
+      "step": 16945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.033428907394409,
+      "learning_rate": 4.515716374269006e-06,
+      "loss": 2.2017,
+      "step": 16946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.091034173965454,
+      "learning_rate": 4.514802631578948e-06,
+      "loss": 1.1875,
+      "step": 16947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3807053565979004,
+      "learning_rate": 4.5138888888888895e-06,
+      "loss": 1.3077,
+      "step": 16948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7164223194122314,
+      "learning_rate": 4.512975146198831e-06,
+      "loss": 1.4792,
+      "step": 16949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2707250118255615,
+      "learning_rate": 4.512061403508772e-06,
+      "loss": 1.5918,
+      "step": 16950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.043670654296875,
+      "learning_rate": 4.511147660818714e-06,
+      "loss": 1.4159,
+      "step": 16951
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6672258377075195,
+      "learning_rate": 4.510233918128655e-06,
+      "loss": 1.4214,
+      "step": 16952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.864474058151245,
+      "learning_rate": 4.509320175438596e-06,
+      "loss": 1.2444,
+      "step": 16953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9397401809692383,
+      "learning_rate": 4.508406432748539e-06,
+      "loss": 1.2779,
+      "step": 16954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2818424701690674,
+      "learning_rate": 4.50749269005848e-06,
+      "loss": 1.6069,
+      "step": 16955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.678161382675171,
+      "learning_rate": 4.506578947368421e-06,
+      "loss": 1.5231,
+      "step": 16956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4886252880096436,
+      "learning_rate": 4.505665204678363e-06,
+      "loss": 1.5244,
+      "step": 16957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.298776865005493,
+      "learning_rate": 4.504751461988304e-06,
+      "loss": 1.2412,
+      "step": 16958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.303602695465088,
+      "learning_rate": 4.5038377192982455e-06,
+      "loss": 1.0339,
+      "step": 16959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9633352756500244,
+      "learning_rate": 4.502923976608187e-06,
+      "loss": 1.7344,
+      "step": 16960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3721814155578613,
+      "learning_rate": 4.502010233918129e-06,
+      "loss": 1.5041,
+      "step": 16961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0104451179504395,
+      "learning_rate": 4.5010964912280705e-06,
+      "loss": 1.1064,
+      "step": 16962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.897953748703003,
+      "learning_rate": 4.500182748538013e-06,
+      "loss": 1.5673,
+      "step": 16963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0903632640838623,
+      "learning_rate": 4.499269005847954e-06,
+      "loss": 1.3221,
+      "step": 16964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6321568489074707,
+      "learning_rate": 4.4983552631578954e-06,
+      "loss": 1.6004,
+      "step": 16965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7130348682403564,
+      "learning_rate": 4.497441520467837e-06,
+      "loss": 1.5488,
+      "step": 16966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5224413871765137,
+      "learning_rate": 4.496527777777778e-06,
+      "loss": 1.4618,
+      "step": 16967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0343406200408936,
+      "learning_rate": 4.4956140350877196e-06,
+      "loss": 0.9564,
+      "step": 16968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.743431329727173,
+      "learning_rate": 4.494700292397661e-06,
+      "loss": 1.8552,
+      "step": 16969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.65155029296875,
+      "learning_rate": 4.493786549707603e-06,
+      "loss": 1.7241,
+      "step": 16970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1333706378936768,
+      "learning_rate": 4.4928728070175445e-06,
+      "loss": 1.3414,
+      "step": 16971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5894415378570557,
+      "learning_rate": 4.491959064327486e-06,
+      "loss": 1.3739,
+      "step": 16972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.065959930419922,
+      "learning_rate": 4.491045321637427e-06,
+      "loss": 1.1607,
+      "step": 16973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0935118198394775,
+      "learning_rate": 4.490131578947369e-06,
+      "loss": 1.2321,
+      "step": 16974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.636542320251465,
+      "learning_rate": 4.48921783625731e-06,
+      "loss": 1.2414,
+      "step": 16975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2383742332458496,
+      "learning_rate": 4.488304093567251e-06,
+      "loss": 1.3149,
+      "step": 16976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.165308952331543,
+      "learning_rate": 4.487390350877194e-06,
+      "loss": 1.3924,
+      "step": 16977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3161356449127197,
+      "learning_rate": 4.486476608187135e-06,
+      "loss": 1.4848,
+      "step": 16978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.883538246154785,
+      "learning_rate": 4.485562865497076e-06,
+      "loss": 1.6413,
+      "step": 16979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.687896490097046,
+      "learning_rate": 4.484649122807018e-06,
+      "loss": 1.3878,
+      "step": 16980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3511743545532227,
+      "learning_rate": 4.483735380116959e-06,
+      "loss": 1.4698,
+      "step": 16981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.074171543121338,
+      "learning_rate": 4.4828216374269005e-06,
+      "loss": 1.1943,
+      "step": 16982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3656506538391113,
+      "learning_rate": 4.481907894736843e-06,
+      "loss": 1.5783,
+      "step": 16983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.318661689758301,
+      "learning_rate": 4.480994152046784e-06,
+      "loss": 1.6618,
+      "step": 16984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2167816162109375,
+      "learning_rate": 4.4800804093567255e-06,
+      "loss": 1.2676,
+      "step": 16985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3771700859069824,
+      "learning_rate": 4.479166666666667e-06,
+      "loss": 1.2932,
+      "step": 16986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8011462688446045,
+      "learning_rate": 4.478252923976609e-06,
+      "loss": 1.7508,
+      "step": 16987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.325927495956421,
+      "learning_rate": 4.4773391812865504e-06,
+      "loss": 1.4416,
+      "step": 16988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.410782814025879,
+      "learning_rate": 4.476425438596492e-06,
+      "loss": 1.5247,
+      "step": 16989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.782700538635254,
+      "learning_rate": 4.475511695906433e-06,
+      "loss": 1.3694,
+      "step": 16990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5434443950653076,
+      "learning_rate": 4.4745979532163746e-06,
+      "loss": 1.5076,
+      "step": 16991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1585090160369873,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 1.7257,
+      "step": 16992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.625901460647583,
+      "learning_rate": 4.472770467836257e-06,
+      "loss": 1.5335,
+      "step": 16993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0883805751800537,
+      "learning_rate": 4.4718567251461995e-06,
+      "loss": 1.0916,
+      "step": 16994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1829004287719727,
+      "learning_rate": 4.470942982456141e-06,
+      "loss": 1.1492,
+      "step": 16995
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.334061622619629,
+      "learning_rate": 4.470029239766082e-06,
+      "loss": 1.5312,
+      "step": 16996
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7130842208862305,
+      "learning_rate": 4.469115497076024e-06,
+      "loss": 1.5274,
+      "step": 16997
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4072611331939697,
+      "learning_rate": 4.468201754385965e-06,
+      "loss": 1.1963,
+      "step": 16998
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6838228702545166,
+      "learning_rate": 4.467288011695906e-06,
+      "loss": 1.6529,
+      "step": 16999
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6417272090911865,
+      "learning_rate": 4.466374269005848e-06,
+      "loss": 1.2752,
+      "step": 17000
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.71897554397583,
+      "learning_rate": 4.46546052631579e-06,
+      "loss": 1.6075,
+      "step": 17001
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8091347217559814,
+      "learning_rate": 4.464546783625731e-06,
+      "loss": 1.6972,
+      "step": 17002
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8464694023132324,
+      "learning_rate": 4.463633040935673e-06,
+      "loss": 1.7433,
+      "step": 17003
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.374305009841919,
+      "learning_rate": 4.462719298245615e-06,
+      "loss": 1.5745,
+      "step": 17004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5800185203552246,
+      "learning_rate": 4.461805555555556e-06,
+      "loss": 1.5159,
+      "step": 17005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4067258834838867,
+      "learning_rate": 4.460891812865498e-06,
+      "loss": 1.3988,
+      "step": 17006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5195038318634033,
+      "learning_rate": 4.459978070175439e-06,
+      "loss": 1.4564,
+      "step": 17007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.745976448059082,
+      "learning_rate": 4.4590643274853805e-06,
+      "loss": 1.6322,
+      "step": 17008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.426671266555786,
+      "learning_rate": 4.458150584795322e-06,
+      "loss": 1.2418,
+      "step": 17009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4456684589385986,
+      "learning_rate": 4.457236842105264e-06,
+      "loss": 1.128,
+      "step": 17010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4386913776397705,
+      "learning_rate": 4.4563230994152054e-06,
+      "loss": 1.3443,
+      "step": 17011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.552682399749756,
+      "learning_rate": 4.455409356725147e-06,
+      "loss": 1.4909,
+      "step": 17012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4013373851776123,
+      "learning_rate": 4.454495614035088e-06,
+      "loss": 1.5387,
+      "step": 17013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4044506549835205,
+      "learning_rate": 4.4535818713450296e-06,
+      "loss": 1.4188,
+      "step": 17014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1460745334625244,
+      "learning_rate": 4.452668128654971e-06,
+      "loss": 1.8507,
+      "step": 17015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.322103977203369,
+      "learning_rate": 4.451754385964912e-06,
+      "loss": 1.6221,
+      "step": 17016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.435274839401245,
+      "learning_rate": 4.450840643274854e-06,
+      "loss": 1.4816,
+      "step": 17017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.156224012374878,
+      "learning_rate": 4.449926900584796e-06,
+      "loss": 1.1534,
+      "step": 17018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1734752655029297,
+      "learning_rate": 4.449013157894737e-06,
+      "loss": 1.1964,
+      "step": 17019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.580327272415161,
+      "learning_rate": 4.448099415204679e-06,
+      "loss": 1.7082,
+      "step": 17020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4354634284973145,
+      "learning_rate": 4.44718567251462e-06,
+      "loss": 1.3059,
+      "step": 17021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2423410415649414,
+      "learning_rate": 4.446271929824561e-06,
+      "loss": 1.5144,
+      "step": 17022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5489580631256104,
+      "learning_rate": 4.445358187134503e-06,
+      "loss": 1.5734,
+      "step": 17023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2016336917877197,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.425,
+      "step": 17024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.114776611328125,
+      "learning_rate": 4.443530701754386e-06,
+      "loss": 1.3728,
+      "step": 17025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.508416175842285,
+      "learning_rate": 4.442616959064328e-06,
+      "loss": 1.3349,
+      "step": 17026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5286357402801514,
+      "learning_rate": 4.44170321637427e-06,
+      "loss": 1.5078,
+      "step": 17027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9457857608795166,
+      "learning_rate": 4.440789473684211e-06,
+      "loss": 1.6163,
+      "step": 17028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2018280029296875,
+      "learning_rate": 4.439875730994153e-06,
+      "loss": 1.2838,
+      "step": 17029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5771257877349854,
+      "learning_rate": 4.438961988304094e-06,
+      "loss": 1.4104,
+      "step": 17030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2017202377319336,
+      "learning_rate": 4.4380482456140355e-06,
+      "loss": 1.7221,
+      "step": 17031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7789196968078613,
+      "learning_rate": 4.437134502923977e-06,
+      "loss": 1.3235,
+      "step": 17032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.892902135848999,
+      "learning_rate": 4.436220760233918e-06,
+      "loss": 1.5353,
+      "step": 17033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2351479530334473,
+      "learning_rate": 4.4353070175438604e-06,
+      "loss": 1.2167,
+      "step": 17034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.576242208480835,
+      "learning_rate": 4.434393274853802e-06,
+      "loss": 1.8155,
+      "step": 17035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.642519474029541,
+      "learning_rate": 4.433479532163743e-06,
+      "loss": 1.6556,
+      "step": 17036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.52176833152771,
+      "learning_rate": 4.4325657894736846e-06,
+      "loss": 1.6701,
+      "step": 17037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4528002738952637,
+      "learning_rate": 4.431652046783626e-06,
+      "loss": 1.4213,
+      "step": 17038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.14512038230896,
+      "learning_rate": 4.430738304093567e-06,
+      "loss": 1.3798,
+      "step": 17039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5672552585601807,
+      "learning_rate": 4.429824561403509e-06,
+      "loss": 1.3599,
+      "step": 17040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3485491275787354,
+      "learning_rate": 4.428910818713451e-06,
+      "loss": 1.2004,
+      "step": 17041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9971649646759033,
+      "learning_rate": 4.427997076023392e-06,
+      "loss": 1.4459,
+      "step": 17042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6947524547576904,
+      "learning_rate": 4.427083333333334e-06,
+      "loss": 1.4836,
+      "step": 17043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8512942790985107,
+      "learning_rate": 4.426169590643275e-06,
+      "loss": 1.7522,
+      "step": 17044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.778367042541504,
+      "learning_rate": 4.425255847953216e-06,
+      "loss": 1.4942,
+      "step": 17045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.64102840423584,
+      "learning_rate": 4.424342105263158e-06,
+      "loss": 1.2814,
+      "step": 17046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2986927032470703,
+      "learning_rate": 4.4234283625731e-06,
+      "loss": 1.3783,
+      "step": 17047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3323006629943848,
+      "learning_rate": 4.422514619883041e-06,
+      "loss": 2.0287,
+      "step": 17048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2030627727508545,
+      "learning_rate": 4.421600877192983e-06,
+      "loss": 1.1357,
+      "step": 17049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8694496154785156,
+      "learning_rate": 4.420687134502924e-06,
+      "loss": 1.7497,
+      "step": 17050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5760812759399414,
+      "learning_rate": 4.419773391812866e-06,
+      "loss": 1.4498,
+      "step": 17051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3780486583709717,
+      "learning_rate": 4.418859649122808e-06,
+      "loss": 1.4396,
+      "step": 17052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1093223094940186,
+      "learning_rate": 4.417945906432749e-06,
+      "loss": 1.3043,
+      "step": 17053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.649181365966797,
+      "learning_rate": 4.4170321637426905e-06,
+      "loss": 1.4467,
+      "step": 17054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2902615070343018,
+      "learning_rate": 4.416118421052632e-06,
+      "loss": 1.3828,
+      "step": 17055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.548341751098633,
+      "learning_rate": 4.415204678362573e-06,
+      "loss": 1.6845,
+      "step": 17056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4287731647491455,
+      "learning_rate": 4.414290935672515e-06,
+      "loss": 1.4094,
+      "step": 17057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9201982021331787,
+      "learning_rate": 4.413377192982457e-06,
+      "loss": 1.5618,
+      "step": 17058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.55301833152771,
+      "learning_rate": 4.412463450292398e-06,
+      "loss": 1.5098,
+      "step": 17059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.231045722961426,
+      "learning_rate": 4.4115497076023396e-06,
+      "loss": 1.1511,
+      "step": 17060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6941261291503906,
+      "learning_rate": 4.410635964912281e-06,
+      "loss": 1.566,
+      "step": 17061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1109511852264404,
+      "learning_rate": 4.409722222222222e-06,
+      "loss": 1.6721,
+      "step": 17062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.230036735534668,
+      "learning_rate": 4.408808479532164e-06,
+      "loss": 1.3414,
+      "step": 17063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5881876945495605,
+      "learning_rate": 4.407894736842105e-06,
+      "loss": 1.3876,
+      "step": 17064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3788046836853027,
+      "learning_rate": 4.406980994152047e-06,
+      "loss": 1.5719,
+      "step": 17065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.302737236022949,
+      "learning_rate": 4.406067251461989e-06,
+      "loss": 1.326,
+      "step": 17066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4886908531188965,
+      "learning_rate": 4.40515350877193e-06,
+      "loss": 1.5372,
+      "step": 17067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.218353748321533,
+      "learning_rate": 4.404239766081872e-06,
+      "loss": 1.1493,
+      "step": 17068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6495296955108643,
+      "learning_rate": 4.403326023391814e-06,
+      "loss": 1.2224,
+      "step": 17069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.548278331756592,
+      "learning_rate": 4.402412280701755e-06,
+      "loss": 1.7208,
+      "step": 17070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.534722328186035,
+      "learning_rate": 4.401498538011696e-06,
+      "loss": 1.345,
+      "step": 17071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1386373043060303,
+      "learning_rate": 4.400584795321638e-06,
+      "loss": 0.8195,
+      "step": 17072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6195380687713623,
+      "learning_rate": 4.399671052631579e-06,
+      "loss": 1.4532,
+      "step": 17073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9006237983703613,
+      "learning_rate": 4.398757309941521e-06,
+      "loss": 1.6071,
+      "step": 17074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.099052906036377,
+      "learning_rate": 4.397843567251463e-06,
+      "loss": 1.7107,
+      "step": 17075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2989699840545654,
+      "learning_rate": 4.396929824561404e-06,
+      "loss": 1.2791,
+      "step": 17076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9661664962768555,
+      "learning_rate": 4.3960160818713455e-06,
+      "loss": 1.5078,
+      "step": 17077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8315608501434326,
+      "learning_rate": 4.395102339181287e-06,
+      "loss": 1.6283,
+      "step": 17078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8825814723968506,
+      "learning_rate": 4.394188596491228e-06,
+      "loss": 1.4479,
+      "step": 17079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.819979667663574,
+      "learning_rate": 4.39327485380117e-06,
+      "loss": 1.8771,
+      "step": 17080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8359148502349854,
+      "learning_rate": 4.392361111111112e-06,
+      "loss": 1.6501,
+      "step": 17081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4656736850738525,
+      "learning_rate": 4.391447368421053e-06,
+      "loss": 1.6708,
+      "step": 17082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3002235889434814,
+      "learning_rate": 4.3905336257309946e-06,
+      "loss": 1.4252,
+      "step": 17083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.215076208114624,
+      "learning_rate": 4.389619883040936e-06,
+      "loss": 1.43,
+      "step": 17084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.549459457397461,
+      "learning_rate": 4.388706140350877e-06,
+      "loss": 1.6465,
+      "step": 17085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.080500364303589,
+      "learning_rate": 4.387792397660819e-06,
+      "loss": 1.1563,
+      "step": 17086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0828263759613037,
+      "learning_rate": 4.38687865497076e-06,
+      "loss": 1.3517,
+      "step": 17087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4777262210845947,
+      "learning_rate": 4.385964912280702e-06,
+      "loss": 1.6745,
+      "step": 17088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2508351802825928,
+      "learning_rate": 4.385051169590644e-06,
+      "loss": 1.3449,
+      "step": 17089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6923091411590576,
+      "learning_rate": 4.384137426900585e-06,
+      "loss": 1.6342,
+      "step": 17090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.290219306945801,
+      "learning_rate": 4.383223684210527e-06,
+      "loss": 1.3538,
+      "step": 17091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.901313543319702,
+      "learning_rate": 4.382309941520469e-06,
+      "loss": 1.2632,
+      "step": 17092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8453760147094727,
+      "learning_rate": 4.38139619883041e-06,
+      "loss": 1.4807,
+      "step": 17093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.526588201522827,
+      "learning_rate": 4.380482456140351e-06,
+      "loss": 1.3004,
+      "step": 17094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.439012050628662,
+      "learning_rate": 4.379568713450293e-06,
+      "loss": 1.5973,
+      "step": 17095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.737159490585327,
+      "learning_rate": 4.378654970760234e-06,
+      "loss": 1.4092,
+      "step": 17096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1787972450256348,
+      "learning_rate": 4.3777412280701755e-06,
+      "loss": 1.2763,
+      "step": 17097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7299113273620605,
+      "learning_rate": 4.376827485380118e-06,
+      "loss": 1.9475,
+      "step": 17098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2956016063690186,
+      "learning_rate": 4.375913742690059e-06,
+      "loss": 1.5747,
+      "step": 17099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4218077659606934,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 1.5391,
+      "step": 17100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4605958461761475,
+      "learning_rate": 4.374086257309942e-06,
+      "loss": 1.5127,
+      "step": 17101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.272313117980957,
+      "learning_rate": 4.373172514619883e-06,
+      "loss": 1.3142,
+      "step": 17102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.692261219024658,
+      "learning_rate": 4.372258771929825e-06,
+      "loss": 2.0939,
+      "step": 17103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5465879440307617,
+      "learning_rate": 4.371345029239766e-06,
+      "loss": 1.4089,
+      "step": 17104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0301730632781982,
+      "learning_rate": 4.370431286549708e-06,
+      "loss": 1.7402,
+      "step": 17105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2188615798950195,
+      "learning_rate": 4.3695175438596496e-06,
+      "loss": 1.0716,
+      "step": 17106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7735469341278076,
+      "learning_rate": 4.368603801169591e-06,
+      "loss": 1.8171,
+      "step": 17107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.486560583114624,
+      "learning_rate": 4.367690058479532e-06,
+      "loss": 1.9272,
+      "step": 17108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4612183570861816,
+      "learning_rate": 4.366776315789474e-06,
+      "loss": 1.261,
+      "step": 17109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1141414642333984,
+      "learning_rate": 4.365862573099415e-06,
+      "loss": 1.0313,
+      "step": 17110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.517728805541992,
+      "learning_rate": 4.364948830409357e-06,
+      "loss": 1.6204,
+      "step": 17111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.176126003265381,
+      "learning_rate": 4.364035087719299e-06,
+      "loss": 1.2894,
+      "step": 17112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.308173179626465,
+      "learning_rate": 4.36312134502924e-06,
+      "loss": 1.3908,
+      "step": 17113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0031585693359375,
+      "learning_rate": 4.362207602339182e-06,
+      "loss": 1.7867,
+      "step": 17114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1956613063812256,
+      "learning_rate": 4.361293859649124e-06,
+      "loss": 1.1036,
+      "step": 17115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.651139497756958,
+      "learning_rate": 4.360380116959065e-06,
+      "loss": 1.6672,
+      "step": 17116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5605502128601074,
+      "learning_rate": 4.359466374269006e-06,
+      "loss": 1.4074,
+      "step": 17117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2733826637268066,
+      "learning_rate": 4.358552631578948e-06,
+      "loss": 1.4722,
+      "step": 17118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0608444213867188,
+      "learning_rate": 4.357638888888889e-06,
+      "loss": 1.5778,
+      "step": 17119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.585566759109497,
+      "learning_rate": 4.3567251461988305e-06,
+      "loss": 1.8279,
+      "step": 17120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.420517921447754,
+      "learning_rate": 4.355811403508773e-06,
+      "loss": 1.5447,
+      "step": 17121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.15580153465271,
+      "learning_rate": 4.354897660818714e-06,
+      "loss": 1.4195,
+      "step": 17122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7505815029144287,
+      "learning_rate": 4.3539839181286555e-06,
+      "loss": 1.8973,
+      "step": 17123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6103410720825195,
+      "learning_rate": 4.353070175438597e-06,
+      "loss": 1.5894,
+      "step": 17124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.894415855407715,
+      "learning_rate": 4.352156432748538e-06,
+      "loss": 1.8514,
+      "step": 17125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4552578926086426,
+      "learning_rate": 4.35124269005848e-06,
+      "loss": 1.7257,
+      "step": 17126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2473955154418945,
+      "learning_rate": 4.350328947368421e-06,
+      "loss": 1.3671,
+      "step": 17127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.29286789894104,
+      "learning_rate": 4.349415204678362e-06,
+      "loss": 1.6222,
+      "step": 17128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5764708518981934,
+      "learning_rate": 4.3485014619883046e-06,
+      "loss": 1.5399,
+      "step": 17129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6151318550109863,
+      "learning_rate": 4.347587719298246e-06,
+      "loss": 1.6565,
+      "step": 17130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.553461790084839,
+      "learning_rate": 4.346673976608187e-06,
+      "loss": 1.4893,
+      "step": 17131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.359700918197632,
+      "learning_rate": 4.345760233918129e-06,
+      "loss": 1.5075,
+      "step": 17132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8267343044281006,
+      "learning_rate": 4.344846491228071e-06,
+      "loss": 1.5533,
+      "step": 17133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.218642234802246,
+      "learning_rate": 4.343932748538012e-06,
+      "loss": 1.6758,
+      "step": 17134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.149740695953369,
+      "learning_rate": 4.343019005847954e-06,
+      "loss": 1.2367,
+      "step": 17135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2045719623565674,
+      "learning_rate": 4.342105263157895e-06,
+      "loss": 1.418,
+      "step": 17136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.481921672821045,
+      "learning_rate": 4.341191520467836e-06,
+      "loss": 1.3857,
+      "step": 17137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4783687591552734,
+      "learning_rate": 4.340277777777779e-06,
+      "loss": 1.5153,
+      "step": 17138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9782638549804688,
+      "learning_rate": 4.33936403508772e-06,
+      "loss": 1.112,
+      "step": 17139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5244760513305664,
+      "learning_rate": 4.338450292397661e-06,
+      "loss": 1.3634,
+      "step": 17140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.456993341445923,
+      "learning_rate": 4.337536549707603e-06,
+      "loss": 1.9429,
+      "step": 17141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7155230045318604,
+      "learning_rate": 4.336622807017544e-06,
+      "loss": 1.533,
+      "step": 17142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.567939519882202,
+      "learning_rate": 4.3357090643274855e-06,
+      "loss": 1.4896,
+      "step": 17143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.745980739593506,
+      "learning_rate": 4.334795321637427e-06,
+      "loss": 1.5858,
+      "step": 17144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8556923866271973,
+      "learning_rate": 4.333881578947369e-06,
+      "loss": 1.1593,
+      "step": 17145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.539853572845459,
+      "learning_rate": 4.3329678362573105e-06,
+      "loss": 1.5518,
+      "step": 17146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4188005924224854,
+      "learning_rate": 4.332054093567252e-06,
+      "loss": 1.2141,
+      "step": 17147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0156426429748535,
+      "learning_rate": 4.331140350877193e-06,
+      "loss": 1.1875,
+      "step": 17148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.166064977645874,
+      "learning_rate": 4.330226608187135e-06,
+      "loss": 1.3643,
+      "step": 17149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1827149391174316,
+      "learning_rate": 4.329312865497076e-06,
+      "loss": 1.2586,
+      "step": 17150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.626431465148926,
+      "learning_rate": 4.328399122807017e-06,
+      "loss": 1.4414,
+      "step": 17151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.254051923751831,
+      "learning_rate": 4.3274853801169596e-06,
+      "loss": 1.7577,
+      "step": 17152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6648001670837402,
+      "learning_rate": 4.326571637426901e-06,
+      "loss": 1.5978,
+      "step": 17153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.426435708999634,
+      "learning_rate": 4.325657894736842e-06,
+      "loss": 1.8222,
+      "step": 17154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5072665214538574,
+      "learning_rate": 4.3247441520467845e-06,
+      "loss": 1.4796,
+      "step": 17155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5834741592407227,
+      "learning_rate": 4.323830409356726e-06,
+      "loss": 1.3423,
+      "step": 17156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1820647716522217,
+      "learning_rate": 4.322916666666667e-06,
+      "loss": 1.1915,
+      "step": 17157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.330953359603882,
+      "learning_rate": 4.322002923976609e-06,
+      "loss": 1.3354,
+      "step": 17158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.678943634033203,
+      "learning_rate": 4.32108918128655e-06,
+      "loss": 1.2341,
+      "step": 17159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8516597747802734,
+      "learning_rate": 4.320175438596491e-06,
+      "loss": 1.3159,
+      "step": 17160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0336387157440186,
+      "learning_rate": 4.319261695906433e-06,
+      "loss": 1.1214,
+      "step": 17161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2582974433898926,
+      "learning_rate": 4.318347953216375e-06,
+      "loss": 1.5045,
+      "step": 17162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.601011037826538,
+      "learning_rate": 4.317434210526316e-06,
+      "loss": 1.4306,
+      "step": 17163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.36849045753479,
+      "learning_rate": 4.316520467836258e-06,
+      "loss": 1.3506,
+      "step": 17164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1911795139312744,
+      "learning_rate": 4.315606725146199e-06,
+      "loss": 1.2296,
+      "step": 17165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.321601152420044,
+      "learning_rate": 4.3146929824561405e-06,
+      "loss": 1.4097,
+      "step": 17166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.600092649459839,
+      "learning_rate": 4.313779239766082e-06,
+      "loss": 1.494,
+      "step": 17167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.376202344894409,
+      "learning_rate": 4.312865497076023e-06,
+      "loss": 1.3468,
+      "step": 17168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.528627872467041,
+      "learning_rate": 4.3119517543859655e-06,
+      "loss": 1.5438,
+      "step": 17169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.422013521194458,
+      "learning_rate": 4.311038011695907e-06,
+      "loss": 2.1946,
+      "step": 17170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2968814373016357,
+      "learning_rate": 4.310124269005848e-06,
+      "loss": 1.4578,
+      "step": 17171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5872130393981934,
+      "learning_rate": 4.30921052631579e-06,
+      "loss": 1.2924,
+      "step": 17172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.830533027648926,
+      "learning_rate": 4.308296783625731e-06,
+      "loss": 1.5563,
+      "step": 17173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.266047239303589,
+      "learning_rate": 4.307383040935672e-06,
+      "loss": 1.5079,
+      "step": 17174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.512399911880493,
+      "learning_rate": 4.3064692982456146e-06,
+      "loss": 1.5554,
+      "step": 17175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9112892150878906,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 1.6979,
+      "step": 17176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.54407000541687,
+      "learning_rate": 4.304641812865497e-06,
+      "loss": 1.7184,
+      "step": 17177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6243703365325928,
+      "learning_rate": 4.3037280701754395e-06,
+      "loss": 1.7264,
+      "step": 17178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.639397621154785,
+      "learning_rate": 4.302814327485381e-06,
+      "loss": 1.2181,
+      "step": 17179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4691338539123535,
+      "learning_rate": 4.301900584795322e-06,
+      "loss": 1.4129,
+      "step": 17180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.097174882888794,
+      "learning_rate": 4.300986842105264e-06,
+      "loss": 1.0622,
+      "step": 17181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0886194705963135,
+      "learning_rate": 4.300073099415205e-06,
+      "loss": 1.3122,
+      "step": 17182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0181021690368652,
+      "learning_rate": 4.299159356725146e-06,
+      "loss": 2.1673,
+      "step": 17183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.213622808456421,
+      "learning_rate": 4.298245614035088e-06,
+      "loss": 1.1296,
+      "step": 17184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6555967330932617,
+      "learning_rate": 4.29733187134503e-06,
+      "loss": 1.6829,
+      "step": 17185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4333438873291016,
+      "learning_rate": 4.296418128654971e-06,
+      "loss": 1.5212,
+      "step": 17186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5715599060058594,
+      "learning_rate": 4.295504385964913e-06,
+      "loss": 1.4288,
+      "step": 17187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7746050357818604,
+      "learning_rate": 4.294590643274854e-06,
+      "loss": 1.5328,
+      "step": 17188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6223843097686768,
+      "learning_rate": 4.2936769005847955e-06,
+      "loss": 1.3817,
+      "step": 17189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9123260974884033,
+      "learning_rate": 4.292763157894737e-06,
+      "loss": 1.8584,
+      "step": 17190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.465578079223633,
+      "learning_rate": 4.291849415204678e-06,
+      "loss": 1.374,
+      "step": 17191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3509764671325684,
+      "learning_rate": 4.2909356725146205e-06,
+      "loss": 1.441,
+      "step": 17192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0575554370880127,
+      "learning_rate": 4.290021929824562e-06,
+      "loss": 1.2793,
+      "step": 17193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.531390905380249,
+      "learning_rate": 4.289108187134503e-06,
+      "loss": 1.4053,
+      "step": 17194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7126214504241943,
+      "learning_rate": 4.288194444444445e-06,
+      "loss": 1.853,
+      "step": 17195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4555580615997314,
+      "learning_rate": 4.287280701754386e-06,
+      "loss": 1.3055,
+      "step": 17196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.590768575668335,
+      "learning_rate": 4.286366959064328e-06,
+      "loss": 1.4201,
+      "step": 17197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.251373291015625,
+      "learning_rate": 4.2854532163742696e-06,
+      "loss": 1.322,
+      "step": 17198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8929288387298584,
+      "learning_rate": 4.284539473684211e-06,
+      "loss": 1.1639,
+      "step": 17199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.983412742614746,
+      "learning_rate": 4.283625730994152e-06,
+      "loss": 1.4716,
+      "step": 17200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6658220291137695,
+      "learning_rate": 4.282711988304094e-06,
+      "loss": 1.2545,
+      "step": 17201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.482783555984497,
+      "learning_rate": 4.281798245614036e-06,
+      "loss": 1.5831,
+      "step": 17202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.68575119972229,
+      "learning_rate": 4.280884502923977e-06,
+      "loss": 1.6481,
+      "step": 17203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3384158611297607,
+      "learning_rate": 4.279970760233919e-06,
+      "loss": 1.2689,
+      "step": 17204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.035583257675171,
+      "learning_rate": 4.27905701754386e-06,
+      "loss": 1.4633,
+      "step": 17205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.49127459526062,
+      "learning_rate": 4.278143274853801e-06,
+      "loss": 1.6069,
+      "step": 17206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.487757921218872,
+      "learning_rate": 4.277229532163743e-06,
+      "loss": 1.3339,
+      "step": 17207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0187418460845947,
+      "learning_rate": 4.276315789473684e-06,
+      "loss": 1.812,
+      "step": 17208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8640732765197754,
+      "learning_rate": 4.275402046783626e-06,
+      "loss": 1.7986,
+      "step": 17209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3027210235595703,
+      "learning_rate": 4.274488304093568e-06,
+      "loss": 1.4023,
+      "step": 17210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3491079807281494,
+      "learning_rate": 4.273574561403509e-06,
+      "loss": 1.2905,
+      "step": 17211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0952494144439697,
+      "learning_rate": 4.2726608187134505e-06,
+      "loss": 1.3108,
+      "step": 17212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.414357900619507,
+      "learning_rate": 4.271747076023392e-06,
+      "loss": 1.2975,
+      "step": 17213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.424557685852051,
+      "learning_rate": 4.270833333333333e-06,
+      "loss": 1.3425,
+      "step": 17214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.381448745727539,
+      "learning_rate": 4.269919590643275e-06,
+      "loss": 1.3382,
+      "step": 17215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5086309909820557,
+      "learning_rate": 4.269005847953217e-06,
+      "loss": 1.3153,
+      "step": 17216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6552984714508057,
+      "learning_rate": 4.268092105263158e-06,
+      "loss": 1.6442,
+      "step": 17217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.248877763748169,
+      "learning_rate": 4.2671783625731004e-06,
+      "loss": 1.5686,
+      "step": 17218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6908740997314453,
+      "learning_rate": 4.266264619883042e-06,
+      "loss": 1.5361,
+      "step": 17219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1839494705200195,
+      "learning_rate": 4.265350877192983e-06,
+      "loss": 1.31,
+      "step": 17220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4197773933410645,
+      "learning_rate": 4.2644371345029246e-06,
+      "loss": 1.521,
+      "step": 17221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8951066732406616,
+      "learning_rate": 4.263523391812866e-06,
+      "loss": 1.0654,
+      "step": 17222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.643383026123047,
+      "learning_rate": 4.262609649122807e-06,
+      "loss": 1.4414,
+      "step": 17223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.768176794052124,
+      "learning_rate": 4.261695906432749e-06,
+      "loss": 1.6705,
+      "step": 17224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5868005752563477,
+      "learning_rate": 4.260782163742691e-06,
+      "loss": 1.4553,
+      "step": 17225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3682022094726562,
+      "learning_rate": 4.259868421052632e-06,
+      "loss": 1.3216,
+      "step": 17226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3818039894104004,
+      "learning_rate": 4.258954678362574e-06,
+      "loss": 1.2201,
+      "step": 17227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5669727325439453,
+      "learning_rate": 4.258040935672515e-06,
+      "loss": 1.7829,
+      "step": 17228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4993553161621094,
+      "learning_rate": 4.257127192982456e-06,
+      "loss": 1.4467,
+      "step": 17229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.844844341278076,
+      "learning_rate": 4.256213450292398e-06,
+      "loss": 1.7932,
+      "step": 17230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6983094215393066,
+      "learning_rate": 4.255299707602339e-06,
+      "loss": 1.4648,
+      "step": 17231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.323185443878174,
+      "learning_rate": 4.254385964912281e-06,
+      "loss": 1.2823,
+      "step": 17232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4703567028045654,
+      "learning_rate": 4.253472222222223e-06,
+      "loss": 1.4234,
+      "step": 17233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0226480960845947,
+      "learning_rate": 4.252558479532164e-06,
+      "loss": 1.3055,
+      "step": 17234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.770763397216797,
+      "learning_rate": 4.2516447368421055e-06,
+      "loss": 1.5792,
+      "step": 17235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.557551383972168,
+      "learning_rate": 4.250730994152047e-06,
+      "loss": 1.512,
+      "step": 17236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.638369560241699,
+      "learning_rate": 4.249817251461988e-06,
+      "loss": 1.5138,
+      "step": 17237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0647449493408203,
+      "learning_rate": 4.24890350877193e-06,
+      "loss": 1.2776,
+      "step": 17238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.782621383666992,
+      "learning_rate": 4.247989766081872e-06,
+      "loss": 1.7111,
+      "step": 17239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1781914234161377,
+      "learning_rate": 4.247076023391813e-06,
+      "loss": 1.3776,
+      "step": 17240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.618725538253784,
+      "learning_rate": 4.246162280701755e-06,
+      "loss": 1.2504,
+      "step": 17241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.583362579345703,
+      "learning_rate": 4.245248538011697e-06,
+      "loss": 1.6494,
+      "step": 17242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1980185508728027,
+      "learning_rate": 4.244334795321638e-06,
+      "loss": 1.2413,
+      "step": 17243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.161987066268921,
+      "learning_rate": 4.2434210526315796e-06,
+      "loss": 1.2808,
+      "step": 17244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6927719116210938,
+      "learning_rate": 4.242507309941521e-06,
+      "loss": 1.6446,
+      "step": 17245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4563260078430176,
+      "learning_rate": 4.241593567251462e-06,
+      "loss": 1.3481,
+      "step": 17246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3237104415893555,
+      "learning_rate": 4.240679824561404e-06,
+      "loss": 1.207,
+      "step": 17247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.400515079498291,
+      "learning_rate": 4.239766081871345e-06,
+      "loss": 1.4199,
+      "step": 17248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6573100090026855,
+      "learning_rate": 4.238852339181287e-06,
+      "loss": 2.0906,
+      "step": 17249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.395437002182007,
+      "learning_rate": 4.237938596491229e-06,
+      "loss": 1.2947,
+      "step": 17250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7429449558258057,
+      "learning_rate": 4.23702485380117e-06,
+      "loss": 1.0411,
+      "step": 17251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8768484592437744,
+      "learning_rate": 4.236111111111111e-06,
+      "loss": 1.6629,
+      "step": 17252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6337082386016846,
+      "learning_rate": 4.235197368421053e-06,
+      "loss": 1.701,
+      "step": 17253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2200403213500977,
+      "learning_rate": 4.234283625730994e-06,
+      "loss": 1.253,
+      "step": 17254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4621422290802,
+      "learning_rate": 4.2333698830409355e-06,
+      "loss": 1.7023,
+      "step": 17255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.536292552947998,
+      "learning_rate": 4.232456140350878e-06,
+      "loss": 1.449,
+      "step": 17256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3826329708099365,
+      "learning_rate": 4.231542397660819e-06,
+      "loss": 1.4497,
+      "step": 17257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4739739894866943,
+      "learning_rate": 4.2306286549707605e-06,
+      "loss": 1.2722,
+      "step": 17258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5353620052337646,
+      "learning_rate": 4.229714912280702e-06,
+      "loss": 1.4141,
+      "step": 17259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.654343843460083,
+      "learning_rate": 4.228801169590643e-06,
+      "loss": 1.4849,
+      "step": 17260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.651341199874878,
+      "learning_rate": 4.2278874269005855e-06,
+      "loss": 1.2548,
+      "step": 17261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.157599687576294,
+      "learning_rate": 4.226973684210527e-06,
+      "loss": 1.1174,
+      "step": 17262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.364820718765259,
+      "learning_rate": 4.226059941520468e-06,
+      "loss": 1.2597,
+      "step": 17263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.339205265045166,
+      "learning_rate": 4.22514619883041e-06,
+      "loss": 1.4529,
+      "step": 17264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.791624069213867,
+      "learning_rate": 4.224232456140351e-06,
+      "loss": 1.6877,
+      "step": 17265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7451226711273193,
+      "learning_rate": 4.223318713450293e-06,
+      "loss": 1.6595,
+      "step": 17266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.009669303894043,
+      "learning_rate": 4.2224049707602346e-06,
+      "loss": 1.108,
+      "step": 17267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2307541370391846,
+      "learning_rate": 4.221491228070176e-06,
+      "loss": 1.3913,
+      "step": 17268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9102954864501953,
+      "learning_rate": 4.220577485380117e-06,
+      "loss": 1.0949,
+      "step": 17269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.845156669616699,
+      "learning_rate": 4.219663742690059e-06,
+      "loss": 1.6627,
+      "step": 17270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.647907257080078,
+      "learning_rate": 4.21875e-06,
+      "loss": 1.6282,
+      "step": 17271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5480170249938965,
+      "learning_rate": 4.2178362573099414e-06,
+      "loss": 1.4417,
+      "step": 17272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5844497680664062,
+      "learning_rate": 4.216922514619884e-06,
+      "loss": 1.4958,
+      "step": 17273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7790443897247314,
+      "learning_rate": 4.216008771929825e-06,
+      "loss": 1.7223,
+      "step": 17274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6628217697143555,
+      "learning_rate": 4.215095029239766e-06,
+      "loss": 1.2203,
+      "step": 17275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5451760292053223,
+      "learning_rate": 4.214181286549708e-06,
+      "loss": 1.3643,
+      "step": 17276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2016472816467285,
+      "learning_rate": 4.213267543859649e-06,
+      "loss": 1.106,
+      "step": 17277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3096728324890137,
+      "learning_rate": 4.2123538011695905e-06,
+      "loss": 1.9408,
+      "step": 17278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.714339256286621,
+      "learning_rate": 4.211440058479532e-06,
+      "loss": 1.373,
+      "step": 17279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.559440851211548,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 1.6376,
+      "step": 17280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.247532367706299,
+      "learning_rate": 4.2096125730994155e-06,
+      "loss": 1.9581,
+      "step": 17281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9821951389312744,
+      "learning_rate": 4.208698830409358e-06,
+      "loss": 1.2673,
+      "step": 17282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.324303150177002,
+      "learning_rate": 4.207785087719299e-06,
+      "loss": 1.4445,
+      "step": 17283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1115005016326904,
+      "learning_rate": 4.2068713450292405e-06,
+      "loss": 1.5107,
+      "step": 17284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2830984592437744,
+      "learning_rate": 4.205957602339182e-06,
+      "loss": 1.2002,
+      "step": 17285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1640584468841553,
+      "learning_rate": 4.205043859649123e-06,
+      "loss": 1.233,
+      "step": 17286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8256802558898926,
+      "learning_rate": 4.204130116959065e-06,
+      "loss": 1.6662,
+      "step": 17287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5751688480377197,
+      "learning_rate": 4.203216374269006e-06,
+      "loss": 1.2909,
+      "step": 17288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.856717109680176,
+      "learning_rate": 4.202302631578948e-06,
+      "loss": 1.5618,
+      "step": 17289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.569103956222534,
+      "learning_rate": 4.2013888888888896e-06,
+      "loss": 1.4273,
+      "step": 17290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6582932472229004,
+      "learning_rate": 4.200475146198831e-06,
+      "loss": 1.2289,
+      "step": 17291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.226644992828369,
+      "learning_rate": 4.199561403508772e-06,
+      "loss": 1.3516,
+      "step": 17292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.862729549407959,
+      "learning_rate": 4.198647660818714e-06,
+      "loss": 1.9143,
+      "step": 17293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.007972002029419,
+      "learning_rate": 4.197733918128655e-06,
+      "loss": 1.3167,
+      "step": 17294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.221510171890259,
+      "learning_rate": 4.1968201754385964e-06,
+      "loss": 1.2105,
+      "step": 17295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0427308082580566,
+      "learning_rate": 4.195906432748539e-06,
+      "loss": 1.2578,
+      "step": 17296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4353456497192383,
+      "learning_rate": 4.19499269005848e-06,
+      "loss": 1.492,
+      "step": 17297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.639641523361206,
+      "learning_rate": 4.194078947368421e-06,
+      "loss": 1.6009,
+      "step": 17298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6302099227905273,
+      "learning_rate": 4.193165204678363e-06,
+      "loss": 1.5153,
+      "step": 17299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6094305515289307,
+      "learning_rate": 4.192251461988304e-06,
+      "loss": 1.794,
+      "step": 17300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.306945323944092,
+      "learning_rate": 4.1913377192982455e-06,
+      "loss": 1.4448,
+      "step": 17301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4893696308135986,
+      "learning_rate": 4.190423976608187e-06,
+      "loss": 1.4037,
+      "step": 17302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.247093677520752,
+      "learning_rate": 4.189510233918129e-06,
+      "loss": 1.2122,
+      "step": 17303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.528331995010376,
+      "learning_rate": 4.1885964912280705e-06,
+      "loss": 1.1989,
+      "step": 17304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.494074583053589,
+      "learning_rate": 4.187682748538012e-06,
+      "loss": 1.501,
+      "step": 17305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6599438190460205,
+      "learning_rate": 4.186769005847954e-06,
+      "loss": 1.6535,
+      "step": 17306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1459720134735107,
+      "learning_rate": 4.1858552631578955e-06,
+      "loss": 1.2488,
+      "step": 17307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7014286518096924,
+      "learning_rate": 4.184941520467837e-06,
+      "loss": 1.2829,
+      "step": 17308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3914263248443604,
+      "learning_rate": 4.184027777777778e-06,
+      "loss": 1.7198,
+      "step": 17309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1163125038146973,
+      "learning_rate": 4.18311403508772e-06,
+      "loss": 1.1168,
+      "step": 17310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3514623641967773,
+      "learning_rate": 4.182200292397661e-06,
+      "loss": 1.7064,
+      "step": 17311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4104621410369873,
+      "learning_rate": 4.181286549707602e-06,
+      "loss": 1.4829,
+      "step": 17312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2452392578125,
+      "learning_rate": 4.1803728070175446e-06,
+      "loss": 1.5043,
+      "step": 17313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6681957244873047,
+      "learning_rate": 4.179459064327486e-06,
+      "loss": 1.453,
+      "step": 17314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4433953762054443,
+      "learning_rate": 4.178545321637427e-06,
+      "loss": 1.1396,
+      "step": 17315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.608372211456299,
+      "learning_rate": 4.177631578947369e-06,
+      "loss": 1.5522,
+      "step": 17316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4941749572753906,
+      "learning_rate": 4.17671783625731e-06,
+      "loss": 1.3816,
+      "step": 17317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.875070095062256,
+      "learning_rate": 4.1758040935672514e-06,
+      "loss": 1.6252,
+      "step": 17318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7797229290008545,
+      "learning_rate": 4.174890350877193e-06,
+      "loss": 1.1027,
+      "step": 17319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5686771869659424,
+      "learning_rate": 4.173976608187135e-06,
+      "loss": 1.4642,
+      "step": 17320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3615987300872803,
+      "learning_rate": 4.173062865497076e-06,
+      "loss": 1.4332,
+      "step": 17321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.380052089691162,
+      "learning_rate": 4.172149122807018e-06,
+      "loss": 1.4001,
+      "step": 17322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1369941234588623,
+      "learning_rate": 4.171235380116959e-06,
+      "loss": 1.2795,
+      "step": 17323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6494460105895996,
+      "learning_rate": 4.1703216374269005e-06,
+      "loss": 1.5043,
+      "step": 17324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.007713556289673,
+      "learning_rate": 4.169407894736843e-06,
+      "loss": 1.9222,
+      "step": 17325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7745141983032227,
+      "learning_rate": 4.168494152046784e-06,
+      "loss": 1.4368,
+      "step": 17326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8756524324417114,
+      "learning_rate": 4.1675804093567255e-06,
+      "loss": 1.0647,
+      "step": 17327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.412043333053589,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.3418,
+      "step": 17328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4758665561676025,
+      "learning_rate": 4.165752923976609e-06,
+      "loss": 1.0136,
+      "step": 17329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2160708904266357,
+      "learning_rate": 4.1648391812865505e-06,
+      "loss": 1.2851,
+      "step": 17330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.766761302947998,
+      "learning_rate": 4.163925438596492e-06,
+      "loss": 1.6211,
+      "step": 17331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7956302165985107,
+      "learning_rate": 4.163011695906433e-06,
+      "loss": 1.3052,
+      "step": 17332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3610007762908936,
+      "learning_rate": 4.162097953216375e-06,
+      "loss": 1.4415,
+      "step": 17333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5480949878692627,
+      "learning_rate": 4.161184210526316e-06,
+      "loss": 1.5713,
+      "step": 17334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1721513271331787,
+      "learning_rate": 4.160270467836257e-06,
+      "loss": 1.3336,
+      "step": 17335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3873000144958496,
+      "learning_rate": 4.1593567251461996e-06,
+      "loss": 1.6222,
+      "step": 17336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.436415433883667,
+      "learning_rate": 4.158442982456141e-06,
+      "loss": 1.2475,
+      "step": 17337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7106473445892334,
+      "learning_rate": 4.157529239766082e-06,
+      "loss": 1.0816,
+      "step": 17338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0765016078948975,
+      "learning_rate": 4.156615497076024e-06,
+      "loss": 1.1247,
+      "step": 17339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.78225040435791,
+      "learning_rate": 4.155701754385965e-06,
+      "loss": 1.6209,
+      "step": 17340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.158803701400757,
+      "learning_rate": 4.1547880116959064e-06,
+      "loss": 1.0066,
+      "step": 17341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.354548692703247,
+      "learning_rate": 4.153874269005848e-06,
+      "loss": 1.2332,
+      "step": 17342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5611929893493652,
+      "learning_rate": 4.15296052631579e-06,
+      "loss": 1.4181,
+      "step": 17343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6585075855255127,
+      "learning_rate": 4.152046783625731e-06,
+      "loss": 1.6658,
+      "step": 17344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.609602689743042,
+      "learning_rate": 4.151133040935673e-06,
+      "loss": 1.5653,
+      "step": 17345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.773698091506958,
+      "learning_rate": 4.150219298245615e-06,
+      "loss": 1.5746,
+      "step": 17346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5928499698638916,
+      "learning_rate": 4.149305555555556e-06,
+      "loss": 1.5686,
+      "step": 17347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.257824420928955,
+      "learning_rate": 4.148391812865498e-06,
+      "loss": 1.3991,
+      "step": 17348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6853132247924805,
+      "learning_rate": 4.147478070175439e-06,
+      "loss": 1.334,
+      "step": 17349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2403786182403564,
+      "learning_rate": 4.1465643274853805e-06,
+      "loss": 1.4514,
+      "step": 17350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5947463512420654,
+      "learning_rate": 4.145650584795322e-06,
+      "loss": 1.5977,
+      "step": 17351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.474632978439331,
+      "learning_rate": 4.144736842105263e-06,
+      "loss": 1.3513,
+      "step": 17352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2278501987457275,
+      "learning_rate": 4.1438230994152055e-06,
+      "loss": 1.4716,
+      "step": 17353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5510241985321045,
+      "learning_rate": 4.142909356725147e-06,
+      "loss": 1.4643,
+      "step": 17354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6841423511505127,
+      "learning_rate": 4.141995614035088e-06,
+      "loss": 1.4487,
+      "step": 17355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.060284376144409,
+      "learning_rate": 4.14108187134503e-06,
+      "loss": 1.1929,
+      "step": 17356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.295260429382324,
+      "learning_rate": 4.140168128654971e-06,
+      "loss": 1.2476,
+      "step": 17357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4519670009613037,
+      "learning_rate": 4.139254385964912e-06,
+      "loss": 1.298,
+      "step": 17358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1922807693481445,
+      "learning_rate": 4.138340643274854e-06,
+      "loss": 1.4689,
+      "step": 17359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3871748447418213,
+      "learning_rate": 4.137426900584796e-06,
+      "loss": 1.4767,
+      "step": 17360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7007389068603516,
+      "learning_rate": 4.136513157894737e-06,
+      "loss": 1.7296,
+      "step": 17361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1301052570343018,
+      "learning_rate": 4.135599415204679e-06,
+      "loss": 1.6603,
+      "step": 17362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.37192440032959,
+      "learning_rate": 4.13468567251462e-06,
+      "loss": 1.5559,
+      "step": 17363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3953611850738525,
+      "learning_rate": 4.1337719298245614e-06,
+      "loss": 1.6007,
+      "step": 17364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0911483764648438,
+      "learning_rate": 4.132858187134503e-06,
+      "loss": 1.5442,
+      "step": 17365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.910648822784424,
+      "learning_rate": 4.131944444444444e-06,
+      "loss": 1.389,
+      "step": 17366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.065528154373169,
+      "learning_rate": 4.131030701754386e-06,
+      "loss": 1.0978,
+      "step": 17367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4701125621795654,
+      "learning_rate": 4.130116959064328e-06,
+      "loss": 1.2965,
+      "step": 17368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4469897747039795,
+      "learning_rate": 4.12920321637427e-06,
+      "loss": 1.4585,
+      "step": 17369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.40454363822937,
+      "learning_rate": 4.128289473684211e-06,
+      "loss": 1.5649,
+      "step": 17370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.340226650238037,
+      "learning_rate": 4.127375730994153e-06,
+      "loss": 1.4971,
+      "step": 17371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.533907651901245,
+      "learning_rate": 4.126461988304094e-06,
+      "loss": 1.4314,
+      "step": 17372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6630032062530518,
+      "learning_rate": 4.1255482456140355e-06,
+      "loss": 1.4936,
+      "step": 17373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9315632581710815,
+      "learning_rate": 4.124634502923977e-06,
+      "loss": 1.0101,
+      "step": 17374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3858137130737305,
+      "learning_rate": 4.123720760233918e-06,
+      "loss": 1.5644,
+      "step": 17375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.073251962661743,
+      "learning_rate": 4.12280701754386e-06,
+      "loss": 1.0892,
+      "step": 17376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1498055458068848,
+      "learning_rate": 4.121893274853802e-06,
+      "loss": 1.2321,
+      "step": 17377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8064117431640625,
+      "learning_rate": 4.120979532163743e-06,
+      "loss": 1.6474,
+      "step": 17378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4547817707061768,
+      "learning_rate": 4.120065789473685e-06,
+      "loss": 1.4672,
+      "step": 17379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6405189037323,
+      "learning_rate": 4.119152046783626e-06,
+      "loss": 1.393,
+      "step": 17380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.933758020401001,
+      "learning_rate": 4.118238304093567e-06,
+      "loss": 1.0638,
+      "step": 17381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.205498218536377,
+      "learning_rate": 4.117324561403509e-06,
+      "loss": 1.2665,
+      "step": 17382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0425124168395996,
+      "learning_rate": 4.11641081871345e-06,
+      "loss": 1.2101,
+      "step": 17383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.337554931640625,
+      "learning_rate": 4.115497076023392e-06,
+      "loss": 1.1134,
+      "step": 17384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3103041648864746,
+      "learning_rate": 4.114583333333334e-06,
+      "loss": 1.6223,
+      "step": 17385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7471702098846436,
+      "learning_rate": 4.113669590643275e-06,
+      "loss": 1.6974,
+      "step": 17386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7547519207000732,
+      "learning_rate": 4.1127558479532164e-06,
+      "loss": 1.6719,
+      "step": 17387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.601989984512329,
+      "learning_rate": 4.111842105263158e-06,
+      "loss": 1.4683,
+      "step": 17388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0756471157073975,
+      "learning_rate": 4.1109283625731e-06,
+      "loss": 1.0474,
+      "step": 17389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6268043518066406,
+      "learning_rate": 4.110014619883041e-06,
+      "loss": 1.3736,
+      "step": 17390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5363929271698,
+      "learning_rate": 4.109100877192983e-06,
+      "loss": 1.1216,
+      "step": 17391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.745560646057129,
+      "learning_rate": 4.108187134502924e-06,
+      "loss": 1.6244,
+      "step": 17392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7896101474761963,
+      "learning_rate": 4.107273391812866e-06,
+      "loss": 1.6492,
+      "step": 17393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3882501125335693,
+      "learning_rate": 4.106359649122808e-06,
+      "loss": 1.2587,
+      "step": 17394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6739721298217773,
+      "learning_rate": 4.105445906432749e-06,
+      "loss": 1.7152,
+      "step": 17395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6110379695892334,
+      "learning_rate": 4.1045321637426905e-06,
+      "loss": 1.303,
+      "step": 17396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.835855007171631,
+      "learning_rate": 4.103618421052632e-06,
+      "loss": 1.5317,
+      "step": 17397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4334545135498047,
+      "learning_rate": 4.102704678362573e-06,
+      "loss": 1.1012,
+      "step": 17398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2858834266662598,
+      "learning_rate": 4.101790935672515e-06,
+      "loss": 1.2097,
+      "step": 17399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3434574604034424,
+      "learning_rate": 4.100877192982457e-06,
+      "loss": 1.342,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7832772731781006,
+      "learning_rate": 4.099963450292398e-06,
+      "loss": 1.6113,
+      "step": 17401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8151003122329712,
+      "learning_rate": 4.09904970760234e-06,
+      "loss": 1.0396,
+      "step": 17402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.32627534866333,
+      "learning_rate": 4.098135964912281e-06,
+      "loss": 1.3863,
+      "step": 17403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7618377208709717,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 1.6103,
+      "step": 17404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.23791241645813,
+      "learning_rate": 4.096308479532164e-06,
+      "loss": 1.3342,
+      "step": 17405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8362274169921875,
+      "learning_rate": 4.095394736842105e-06,
+      "loss": 1.3226,
+      "step": 17406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.428877115249634,
+      "learning_rate": 4.094480994152047e-06,
+      "loss": 1.3778,
+      "step": 17407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4175238609313965,
+      "learning_rate": 4.093567251461989e-06,
+      "loss": 1.4452,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3585903644561768,
+      "learning_rate": 4.09265350877193e-06,
+      "loss": 1.6327,
+      "step": 17409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.608142137527466,
+      "learning_rate": 4.091739766081872e-06,
+      "loss": 1.4869,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2257211208343506,
+      "learning_rate": 4.090826023391814e-06,
+      "loss": 1.4007,
+      "step": 17411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4530093669891357,
+      "learning_rate": 4.089912280701755e-06,
+      "loss": 1.2375,
+      "step": 17412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5897903442382812,
+      "learning_rate": 4.088998538011696e-06,
+      "loss": 1.9719,
+      "step": 17413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.613407611846924,
+      "learning_rate": 4.088084795321638e-06,
+      "loss": 1.9052,
+      "step": 17414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.239844799041748,
+      "learning_rate": 4.087171052631579e-06,
+      "loss": 1.2897,
+      "step": 17415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.255505323410034,
+      "learning_rate": 4.0862573099415205e-06,
+      "loss": 1.4284,
+      "step": 17416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9262754917144775,
+      "learning_rate": 4.085343567251463e-06,
+      "loss": 1.173,
+      "step": 17417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0090837478637695,
+      "learning_rate": 4.084429824561404e-06,
+      "loss": 1.646,
+      "step": 17418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.175072431564331,
+      "learning_rate": 4.0835160818713455e-06,
+      "loss": 1.3836,
+      "step": 17419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.464660406112671,
+      "learning_rate": 4.082602339181287e-06,
+      "loss": 1.3784,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7737514972686768,
+      "learning_rate": 4.081688596491228e-06,
+      "loss": 1.4012,
+      "step": 17421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3919832706451416,
+      "learning_rate": 4.08077485380117e-06,
+      "loss": 1.471,
+      "step": 17422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.595918893814087,
+      "learning_rate": 4.079861111111111e-06,
+      "loss": 1.4573,
+      "step": 17423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3903965950012207,
+      "learning_rate": 4.078947368421053e-06,
+      "loss": 1.4806,
+      "step": 17424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.453321695327759,
+      "learning_rate": 4.078033625730995e-06,
+      "loss": 1.1477,
+      "step": 17425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3769171237945557,
+      "learning_rate": 4.077119883040936e-06,
+      "loss": 1.469,
+      "step": 17426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9677608013153076,
+      "learning_rate": 4.076206140350877e-06,
+      "loss": 1.8196,
+      "step": 17427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.772538900375366,
+      "learning_rate": 4.075292397660819e-06,
+      "loss": 1.5145,
+      "step": 17428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.57234787940979,
+      "learning_rate": 4.07437865497076e-06,
+      "loss": 1.4574,
+      "step": 17429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7289624214172363,
+      "learning_rate": 4.0734649122807015e-06,
+      "loss": 1.5783,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7169594764709473,
+      "learning_rate": 4.072551169590644e-06,
+      "loss": 1.5803,
+      "step": 17431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3538033962249756,
+      "learning_rate": 4.071637426900585e-06,
+      "loss": 1.5112,
+      "step": 17432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.736102342605591,
+      "learning_rate": 4.070723684210527e-06,
+      "loss": 1.5752,
+      "step": 17433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2948570251464844,
+      "learning_rate": 4.069809941520469e-06,
+      "loss": 1.5756,
+      "step": 17434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6147966384887695,
+      "learning_rate": 4.06889619883041e-06,
+      "loss": 1.3345,
+      "step": 17435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0026607513427734,
+      "learning_rate": 4.067982456140351e-06,
+      "loss": 1.1927,
+      "step": 17436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6038966178894043,
+      "learning_rate": 4.067068713450293e-06,
+      "loss": 1.5681,
+      "step": 17437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7272512912750244,
+      "learning_rate": 4.066154970760234e-06,
+      "loss": 1.4098,
+      "step": 17438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1898725032806396,
+      "learning_rate": 4.0652412280701755e-06,
+      "loss": 1.8005,
+      "step": 17439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.973653793334961,
+      "learning_rate": 4.064327485380118e-06,
+      "loss": 2.1029,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7854297161102295,
+      "learning_rate": 4.063413742690059e-06,
+      "loss": 1.4762,
+      "step": 17441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8676297664642334,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 1.8364,
+      "step": 17442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4686150550842285,
+      "learning_rate": 4.061586257309942e-06,
+      "loss": 1.3903,
+      "step": 17443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.398247241973877,
+      "learning_rate": 4.060672514619883e-06,
+      "loss": 1.4102,
+      "step": 17444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.786092758178711,
+      "learning_rate": 4.059758771929825e-06,
+      "loss": 1.5049,
+      "step": 17445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6294803619384766,
+      "learning_rate": 4.058845029239766e-06,
+      "loss": 1.568,
+      "step": 17446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.292344331741333,
+      "learning_rate": 4.057931286549708e-06,
+      "loss": 1.2457,
+      "step": 17447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9678590297698975,
+      "learning_rate": 4.05701754385965e-06,
+      "loss": 1.7651,
+      "step": 17448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6200692653656006,
+      "learning_rate": 4.056103801169591e-06,
+      "loss": 1.7315,
+      "step": 17449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.278639316558838,
+      "learning_rate": 4.055190058479532e-06,
+      "loss": 1.3175,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2466607093811035,
+      "learning_rate": 4.054276315789474e-06,
+      "loss": 1.0485,
+      "step": 17451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1204442977905273,
+      "learning_rate": 4.053362573099415e-06,
+      "loss": 1.7575,
+      "step": 17452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.05299711227417,
+      "learning_rate": 4.052448830409357e-06,
+      "loss": 1.7773,
+      "step": 17453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9485676288604736,
+      "learning_rate": 4.051535087719299e-06,
+      "loss": 1.8154,
+      "step": 17454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.24055814743042,
+      "learning_rate": 4.05062134502924e-06,
+      "loss": 1.2969,
+      "step": 17455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.399820327758789,
+      "learning_rate": 4.0497076023391814e-06,
+      "loss": 1.4717,
+      "step": 17456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.793576955795288,
+      "learning_rate": 4.048793859649124e-06,
+      "loss": 1.4622,
+      "step": 17457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.220515727996826,
+      "learning_rate": 4.047880116959065e-06,
+      "loss": 1.3856,
+      "step": 17458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2980971336364746,
+      "learning_rate": 4.046966374269006e-06,
+      "loss": 1.7641,
+      "step": 17459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.368964195251465,
+      "learning_rate": 4.046052631578948e-06,
+      "loss": 1.142,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4886879920959473,
+      "learning_rate": 4.045138888888889e-06,
+      "loss": 1.6323,
+      "step": 17461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.438204050064087,
+      "learning_rate": 4.0442251461988305e-06,
+      "loss": 1.5927,
+      "step": 17462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.533369541168213,
+      "learning_rate": 4.043311403508772e-06,
+      "loss": 1.2241,
+      "step": 17463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.266507625579834,
+      "learning_rate": 4.042397660818714e-06,
+      "loss": 1.212,
+      "step": 17464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2288320064544678,
+      "learning_rate": 4.0414839181286555e-06,
+      "loss": 1.3045,
+      "step": 17465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6330363750457764,
+      "learning_rate": 4.040570175438597e-06,
+      "loss": 1.6743,
+      "step": 17466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3553340435028076,
+      "learning_rate": 4.039656432748538e-06,
+      "loss": 1.2848,
+      "step": 17467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.978031635284424,
+      "learning_rate": 4.03874269005848e-06,
+      "loss": 1.4842,
+      "step": 17468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.518397331237793,
+      "learning_rate": 4.037828947368421e-06,
+      "loss": 1.4672,
+      "step": 17469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4057114124298096,
+      "learning_rate": 4.036915204678362e-06,
+      "loss": 1.5336,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.676546335220337,
+      "learning_rate": 4.036001461988305e-06,
+      "loss": 1.3048,
+      "step": 17471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0740015506744385,
+      "learning_rate": 4.035087719298246e-06,
+      "loss": 1.2768,
+      "step": 17472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.801661252975464,
+      "learning_rate": 4.034173976608187e-06,
+      "loss": 1.5062,
+      "step": 17473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3379039764404297,
+      "learning_rate": 4.033260233918129e-06,
+      "loss": 1.5886,
+      "step": 17474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.259586811065674,
+      "learning_rate": 4.032346491228071e-06,
+      "loss": 1.2596,
+      "step": 17475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8962574005126953,
+      "learning_rate": 4.031432748538012e-06,
+      "loss": 1.4174,
+      "step": 17476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.679429054260254,
+      "learning_rate": 4.030519005847954e-06,
+      "loss": 1.5775,
+      "step": 17477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8051297664642334,
+      "learning_rate": 4.029605263157895e-06,
+      "loss": 1.5075,
+      "step": 17478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3873002529144287,
+      "learning_rate": 4.0286915204678364e-06,
+      "loss": 1.4043,
+      "step": 17479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5557446479797363,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 1.3653,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3603262901306152,
+      "learning_rate": 4.02686403508772e-06,
+      "loss": 1.5123,
+      "step": 17481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5519471168518066,
+      "learning_rate": 4.025950292397661e-06,
+      "loss": 1.6118,
+      "step": 17482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2243552207946777,
+      "learning_rate": 4.025036549707603e-06,
+      "loss": 1.1439,
+      "step": 17483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2695119380950928,
+      "learning_rate": 4.024122807017544e-06,
+      "loss": 1.4131,
+      "step": 17484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2374157905578613,
+      "learning_rate": 4.0232090643274855e-06,
+      "loss": 1.0803,
+      "step": 17485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2773377895355225,
+      "learning_rate": 4.022295321637427e-06,
+      "loss": 1.2897,
+      "step": 17486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.748696804046631,
+      "learning_rate": 4.021381578947368e-06,
+      "loss": 1.3641,
+      "step": 17487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2595295906066895,
+      "learning_rate": 4.0204678362573105e-06,
+      "loss": 1.4295,
+      "step": 17488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5859124660491943,
+      "learning_rate": 4.019554093567252e-06,
+      "loss": 1.3495,
+      "step": 17489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8201634883880615,
+      "learning_rate": 4.018640350877193e-06,
+      "loss": 1.7837,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.797004222869873,
+      "learning_rate": 4.017726608187135e-06,
+      "loss": 1.5983,
+      "step": 17491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.534113645553589,
+      "learning_rate": 4.016812865497076e-06,
+      "loss": 1.213,
+      "step": 17492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9979562759399414,
+      "learning_rate": 4.015899122807017e-06,
+      "loss": 1.3522,
+      "step": 17493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.574605703353882,
+      "learning_rate": 4.014985380116959e-06,
+      "loss": 1.4948,
+      "step": 17494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3888304233551025,
+      "learning_rate": 4.014071637426901e-06,
+      "loss": 1.4759,
+      "step": 17495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.415215492248535,
+      "learning_rate": 4.013157894736842e-06,
+      "loss": 1.3379,
+      "step": 17496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3311119079589844,
+      "learning_rate": 4.0122441520467846e-06,
+      "loss": 1.1679,
+      "step": 17497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2181003093719482,
+      "learning_rate": 4.011330409356726e-06,
+      "loss": 1.3352,
+      "step": 17498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.522284507751465,
+      "learning_rate": 4.010416666666667e-06,
+      "loss": 1.3716,
+      "step": 17499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7589669227600098,
+      "learning_rate": 4.009502923976609e-06,
+      "loss": 1.5011,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2258918285369873,
+      "learning_rate": 4.00858918128655e-06,
+      "loss": 1.7007,
+      "step": 17501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4637112617492676,
+      "learning_rate": 4.0076754385964914e-06,
+      "loss": 1.3458,
+      "step": 17502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.398909330368042,
+      "learning_rate": 4.006761695906433e-06,
+      "loss": 1.4387,
+      "step": 17503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9820270538330078,
+      "learning_rate": 4.005847953216375e-06,
+      "loss": 1.41,
+      "step": 17504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3849635124206543,
+      "learning_rate": 4.004934210526316e-06,
+      "loss": 1.4824,
+      "step": 17505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.151820182800293,
+      "learning_rate": 4.004020467836258e-06,
+      "loss": 1.2383,
+      "step": 17506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9334170818328857,
+      "learning_rate": 4.003106725146199e-06,
+      "loss": 1.5811,
+      "step": 17507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0177624225616455,
+      "learning_rate": 4.0021929824561405e-06,
+      "loss": 1.1096,
+      "step": 17508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8549904823303223,
+      "learning_rate": 4.001279239766082e-06,
+      "loss": 1.666,
+      "step": 17509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.734651803970337,
+      "learning_rate": 4.000365497076023e-06,
+      "loss": 1.4776,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.190272331237793,
+      "learning_rate": 3.9994517543859655e-06,
+      "loss": 1.226,
+      "step": 17511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9105355739593506,
+      "learning_rate": 3.998538011695907e-06,
+      "loss": 1.6392,
+      "step": 17512
+    },
+    {
+      "epoch": 0.8,
+      "eval_train_bleu": 0.22838325580391647,
+      "eval_train_loss": 2.0962514877319336,
+      "eval_train_runtime": 24.2038,
+      "eval_train_samples_per_second": 41.357,
+      "eval_train_steps_per_second": 1.322,
+      "step": 17512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.770211696624756,
+      "learning_rate": 3.997624269005848e-06,
+      "loss": 1.4183,
+      "step": 17513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.284808397293091,
+      "learning_rate": 3.99671052631579e-06,
+      "loss": 1.2879,
+      "step": 17514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.036804676055908,
+      "learning_rate": 3.995796783625731e-06,
+      "loss": 1.9283,
+      "step": 17515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0609629154205322,
+      "learning_rate": 3.994883040935672e-06,
+      "loss": 1.2637,
+      "step": 17516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.714655637741089,
+      "learning_rate": 3.993969298245615e-06,
+      "loss": 1.3234,
+      "step": 17517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.220667839050293,
+      "learning_rate": 3.993055555555556e-06,
+      "loss": 1.4542,
+      "step": 17518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5184521675109863,
+      "learning_rate": 3.992141812865497e-06,
+      "loss": 1.5338,
+      "step": 17519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3792614936828613,
+      "learning_rate": 3.991228070175439e-06,
+      "loss": 1.5571,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.660184144973755,
+      "learning_rate": 3.990314327485381e-06,
+      "loss": 2.0622,
+      "step": 17521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.21036434173584,
+      "learning_rate": 3.989400584795322e-06,
+      "loss": 1.2123,
+      "step": 17522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4684455394744873,
+      "learning_rate": 3.988486842105264e-06,
+      "loss": 1.5468,
+      "step": 17523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.600809335708618,
+      "learning_rate": 3.987573099415205e-06,
+      "loss": 1.673,
+      "step": 17524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.419750213623047,
+      "learning_rate": 3.9866593567251464e-06,
+      "loss": 1.4647,
+      "step": 17525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.253300905227661,
+      "learning_rate": 3.985745614035088e-06,
+      "loss": 1.2907,
+      "step": 17526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9323643445968628,
+      "learning_rate": 3.984831871345029e-06,
+      "loss": 1.243,
+      "step": 17527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8878049850463867,
+      "learning_rate": 3.983918128654971e-06,
+      "loss": 1.7542,
+      "step": 17528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.827122688293457,
+      "learning_rate": 3.983004385964913e-06,
+      "loss": 1.7492,
+      "step": 17529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4546501636505127,
+      "learning_rate": 3.982090643274854e-06,
+      "loss": 1.6113,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.472358465194702,
+      "learning_rate": 3.9811769005847955e-06,
+      "loss": 1.3889,
+      "step": 17531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.479283571243286,
+      "learning_rate": 3.980263157894737e-06,
+      "loss": 1.3526,
+      "step": 17532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1326396465301514,
+      "learning_rate": 3.979349415204678e-06,
+      "loss": 1.089,
+      "step": 17533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3322224617004395,
+      "learning_rate": 3.97843567251462e-06,
+      "loss": 1.1666,
+      "step": 17534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1867258548736572,
+      "learning_rate": 3.977521929824562e-06,
+      "loss": 1.257,
+      "step": 17535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2888145446777344,
+      "learning_rate": 3.976608187134503e-06,
+      "loss": 1.3329,
+      "step": 17536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5894174575805664,
+      "learning_rate": 3.975694444444445e-06,
+      "loss": 1.352,
+      "step": 17537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2206854820251465,
+      "learning_rate": 3.974780701754386e-06,
+      "loss": 1.7344,
+      "step": 17538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7158005237579346,
+      "learning_rate": 3.973866959064328e-06,
+      "loss": 1.4655,
+      "step": 17539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7277309894561768,
+      "learning_rate": 3.97295321637427e-06,
+      "loss": 1.5443,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.110213041305542,
+      "learning_rate": 3.972039473684211e-06,
+      "loss": 1.1317,
+      "step": 17541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.78047776222229,
+      "learning_rate": 3.971125730994152e-06,
+      "loss": 1.5924,
+      "step": 17542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.489520311355591,
+      "learning_rate": 3.970211988304094e-06,
+      "loss": 1.4113,
+      "step": 17543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.918698787689209,
+      "learning_rate": 3.969298245614036e-06,
+      "loss": 1.6531,
+      "step": 17544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8599460124969482,
+      "learning_rate": 3.968384502923977e-06,
+      "loss": 1.2359,
+      "step": 17545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.615787982940674,
+      "learning_rate": 3.967470760233919e-06,
+      "loss": 1.6791,
+      "step": 17546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.188363552093506,
+      "learning_rate": 3.96655701754386e-06,
+      "loss": 1.6668,
+      "step": 17547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.853396415710449,
+      "learning_rate": 3.9656432748538014e-06,
+      "loss": 1.7123,
+      "step": 17548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8255457878112793,
+      "learning_rate": 3.964729532163743e-06,
+      "loss": 1.2687,
+      "step": 17549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1112937927246094,
+      "learning_rate": 3.963815789473684e-06,
+      "loss": 1.6069,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5222842693328857,
+      "learning_rate": 3.962902046783626e-06,
+      "loss": 1.572,
+      "step": 17551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.395634889602661,
+      "learning_rate": 3.961988304093568e-06,
+      "loss": 1.5832,
+      "step": 17552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.622469186782837,
+      "learning_rate": 3.961074561403509e-06,
+      "loss": 1.5279,
+      "step": 17553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6942086219787598,
+      "learning_rate": 3.9601608187134505e-06,
+      "loss": 1.4671,
+      "step": 17554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7680726051330566,
+      "learning_rate": 3.959247076023392e-06,
+      "loss": 1.5835,
+      "step": 17555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.334239959716797,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 1.4557,
+      "step": 17556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.619656801223755,
+      "learning_rate": 3.957419590643275e-06,
+      "loss": 1.7049,
+      "step": 17557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1098906993865967,
+      "learning_rate": 3.956505847953217e-06,
+      "loss": 1.1311,
+      "step": 17558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.115372657775879,
+      "learning_rate": 3.955592105263158e-06,
+      "loss": 1.4211,
+      "step": 17559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3072235584259033,
+      "learning_rate": 3.9546783625731e-06,
+      "loss": 1.3623,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.427293300628662,
+      "learning_rate": 3.953764619883042e-06,
+      "loss": 1.2119,
+      "step": 17561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.337369918823242,
+      "learning_rate": 3.952850877192983e-06,
+      "loss": 1.4009,
+      "step": 17562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.813882350921631,
+      "learning_rate": 3.951937134502925e-06,
+      "loss": 1.5423,
+      "step": 17563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2414581775665283,
+      "learning_rate": 3.951023391812866e-06,
+      "loss": 1.2433,
+      "step": 17564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.326169967651367,
+      "learning_rate": 3.950109649122807e-06,
+      "loss": 1.2884,
+      "step": 17565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.350955009460449,
+      "learning_rate": 3.949195906432749e-06,
+      "loss": 1.3018,
+      "step": 17566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.822211503982544,
+      "learning_rate": 3.94828216374269e-06,
+      "loss": 1.3753,
+      "step": 17567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3294036388397217,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 1.3449,
+      "step": 17568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6195013523101807,
+      "learning_rate": 3.946454678362574e-06,
+      "loss": 1.8677,
+      "step": 17569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.26849365234375,
+      "learning_rate": 3.945540935672515e-06,
+      "loss": 1.4497,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.069554090499878,
+      "learning_rate": 3.9446271929824564e-06,
+      "loss": 2.1465,
+      "step": 17571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4206576347351074,
+      "learning_rate": 3.943713450292398e-06,
+      "loss": 1.4439,
+      "step": 17572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9422367811203003,
+      "learning_rate": 3.942799707602339e-06,
+      "loss": 1.115,
+      "step": 17573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.310459852218628,
+      "learning_rate": 3.9418859649122805e-06,
+      "loss": 1.4827,
+      "step": 17574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1936514377593994,
+      "learning_rate": 3.940972222222223e-06,
+      "loss": 1.3459,
+      "step": 17575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4728293418884277,
+      "learning_rate": 3.940058479532164e-06,
+      "loss": 1.5516,
+      "step": 17576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6056458950042725,
+      "learning_rate": 3.9391447368421055e-06,
+      "loss": 1.5934,
+      "step": 17577
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.820091962814331,
+      "learning_rate": 3.938230994152047e-06,
+      "loss": 2.1252,
+      "step": 17578
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.416867733001709,
+      "learning_rate": 3.937317251461988e-06,
+      "loss": 1.5401,
+      "step": 17579
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0794498920440674,
+      "learning_rate": 3.93640350877193e-06,
+      "loss": 1.8143,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5290157794952393,
+      "learning_rate": 3.935489766081872e-06,
+      "loss": 1.7862,
+      "step": 17581
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0191895961761475,
+      "learning_rate": 3.934576023391813e-06,
+      "loss": 1.3673,
+      "step": 17582
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.647928237915039,
+      "learning_rate": 3.933662280701755e-06,
+      "loss": 1.516,
+      "step": 17583
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1498031616210938,
+      "learning_rate": 3.932748538011697e-06,
+      "loss": 1.3119,
+      "step": 17584
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0308334827423096,
+      "learning_rate": 3.931834795321638e-06,
+      "loss": 1.5545,
+      "step": 17585
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9048256874084473,
+      "learning_rate": 3.93092105263158e-06,
+      "loss": 1.5545,
+      "step": 17586
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9758315086364746,
+      "learning_rate": 3.930007309941521e-06,
+      "loss": 1.949,
+      "step": 17587
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.462322235107422,
+      "learning_rate": 3.929093567251462e-06,
+      "loss": 1.0922,
+      "step": 17588
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.795389413833618,
+      "learning_rate": 3.928179824561404e-06,
+      "loss": 1.7729,
+      "step": 17589
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2296924591064453,
+      "learning_rate": 3.927266081871345e-06,
+      "loss": 1.5423,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.698331832885742,
+      "learning_rate": 3.926352339181287e-06,
+      "loss": 1.3173,
+      "step": 17591
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.389615297317505,
+      "learning_rate": 3.925438596491229e-06,
+      "loss": 1.3733,
+      "step": 17592
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.558321237564087,
+      "learning_rate": 3.92452485380117e-06,
+      "loss": 1.4773,
+      "step": 17593
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.626460552215576,
+      "learning_rate": 3.9236111111111114e-06,
+      "loss": 1.5877,
+      "step": 17594
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4527454376220703,
+      "learning_rate": 3.922697368421053e-06,
+      "loss": 1.5206,
+      "step": 17595
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3669209480285645,
+      "learning_rate": 3.921783625730994e-06,
+      "loss": 1.0932,
+      "step": 17596
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4521403312683105,
+      "learning_rate": 3.9208698830409355e-06,
+      "loss": 1.555,
+      "step": 17597
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.812891721725464,
+      "learning_rate": 3.919956140350878e-06,
+      "loss": 1.6193,
+      "step": 17598
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.049056053161621,
+      "learning_rate": 3.919042397660819e-06,
+      "loss": 1.7215,
+      "step": 17599
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3060460090637207,
+      "learning_rate": 3.9181286549707605e-06,
+      "loss": 1.4069,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.486128330230713,
+      "learning_rate": 3.917214912280702e-06,
+      "loss": 1.5021,
+      "step": 17601
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3857638835906982,
+      "learning_rate": 3.916301169590643e-06,
+      "loss": 1.4591,
+      "step": 17602
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8712692260742188,
+      "learning_rate": 3.9153874269005855e-06,
+      "loss": 1.7786,
+      "step": 17603
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.076030969619751,
+      "learning_rate": 3.914473684210527e-06,
+      "loss": 1.2609,
+      "step": 17604
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.704188823699951,
+      "learning_rate": 3.913559941520468e-06,
+      "loss": 1.3786,
+      "step": 17605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7280423641204834,
+      "learning_rate": 3.91264619883041e-06,
+      "loss": 1.7462,
+      "step": 17606
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.46357798576355,
+      "learning_rate": 3.911732456140351e-06,
+      "loss": 1.3482,
+      "step": 17607
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5019054412841797,
+      "learning_rate": 3.910818713450293e-06,
+      "loss": 1.4179,
+      "step": 17608
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4381918907165527,
+      "learning_rate": 3.909904970760235e-06,
+      "loss": 1.5391,
+      "step": 17609
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.422875165939331,
+      "learning_rate": 3.908991228070176e-06,
+      "loss": 1.2977,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.46885347366333,
+      "learning_rate": 3.908077485380117e-06,
+      "loss": 1.4109,
+      "step": 17611
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8475797176361084,
+      "learning_rate": 3.907163742690059e-06,
+      "loss": 1.1865,
+      "step": 17612
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6484644412994385,
+      "learning_rate": 3.90625e-06,
+      "loss": 1.7422,
+      "step": 17613
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.571096897125244,
+      "learning_rate": 3.9053362573099415e-06,
+      "loss": 1.2761,
+      "step": 17614
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4320828914642334,
+      "learning_rate": 3.904422514619884e-06,
+      "loss": 1.4565,
+      "step": 17615
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3325302600860596,
+      "learning_rate": 3.903508771929825e-06,
+      "loss": 1.3868,
+      "step": 17616
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2982468605041504,
+      "learning_rate": 3.9025950292397664e-06,
+      "loss": 1.3254,
+      "step": 17617
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.787903070449829,
+      "learning_rate": 3.901681286549708e-06,
+      "loss": 1.6748,
+      "step": 17618
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.857667922973633,
+      "learning_rate": 3.900767543859649e-06,
+      "loss": 1.6037,
+      "step": 17619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2744765281677246,
+      "learning_rate": 3.8998538011695905e-06,
+      "loss": 1.4234,
+      "step": 17620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2722699642181396,
+      "learning_rate": 3.898940058479532e-06,
+      "loss": 1.297,
+      "step": 17621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5084540843963623,
+      "learning_rate": 3.898026315789474e-06,
+      "loss": 1.4175,
+      "step": 17622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5903613567352295,
+      "learning_rate": 3.8971125730994155e-06,
+      "loss": 1.6178,
+      "step": 17623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.957401990890503,
+      "learning_rate": 3.896198830409357e-06,
+      "loss": 1.2331,
+      "step": 17624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7177581787109375,
+      "learning_rate": 3.895285087719299e-06,
+      "loss": 1.5481,
+      "step": 17625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9758975505828857,
+      "learning_rate": 3.8943713450292405e-06,
+      "loss": 1.4813,
+      "step": 17626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.546334743499756,
+      "learning_rate": 3.893457602339182e-06,
+      "loss": 1.6952,
+      "step": 17627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.755336046218872,
+      "learning_rate": 3.892543859649123e-06,
+      "loss": 1.8001,
+      "step": 17628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.632582187652588,
+      "learning_rate": 3.891630116959065e-06,
+      "loss": 1.4008,
+      "step": 17629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9288667440414429,
+      "learning_rate": 3.890716374269006e-06,
+      "loss": 1.1769,
+      "step": 17630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2747862339019775,
+      "learning_rate": 3.889802631578947e-06,
+      "loss": 1.4482,
+      "step": 17631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.41652512550354,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 1.5226,
+      "step": 17632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2803478240966797,
+      "learning_rate": 3.887975146198831e-06,
+      "loss": 1.2496,
+      "step": 17633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.108377695083618,
+      "learning_rate": 3.887061403508772e-06,
+      "loss": 1.3242,
+      "step": 17634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3292713165283203,
+      "learning_rate": 3.886147660818714e-06,
+      "loss": 1.6668,
+      "step": 17635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3507189750671387,
+      "learning_rate": 3.885233918128655e-06,
+      "loss": 1.3544,
+      "step": 17636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4579200744628906,
+      "learning_rate": 3.8843201754385965e-06,
+      "loss": 1.5216,
+      "step": 17637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.139404296875,
+      "learning_rate": 3.883406432748538e-06,
+      "loss": 1.2216,
+      "step": 17638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.742274045944214,
+      "learning_rate": 3.88249269005848e-06,
+      "loss": 2.0428,
+      "step": 17639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5129446983337402,
+      "learning_rate": 3.8815789473684214e-06,
+      "loss": 1.5856,
+      "step": 17640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.302135944366455,
+      "learning_rate": 3.880665204678363e-06,
+      "loss": 1.4628,
+      "step": 17641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1630921363830566,
+      "learning_rate": 3.879751461988304e-06,
+      "loss": 1.1021,
+      "step": 17642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1220474243164062,
+      "learning_rate": 3.8788377192982455e-06,
+      "loss": 1.2159,
+      "step": 17643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1599087715148926,
+      "learning_rate": 3.877923976608187e-06,
+      "loss": 1.259,
+      "step": 17644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7921767234802246,
+      "learning_rate": 3.877010233918128e-06,
+      "loss": 1.7732,
+      "step": 17645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.48184871673584,
+      "learning_rate": 3.8760964912280705e-06,
+      "loss": 1.5243,
+      "step": 17646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1662189960479736,
+      "learning_rate": 3.875182748538012e-06,
+      "loss": 1.3337,
+      "step": 17647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.603468418121338,
+      "learning_rate": 3.874269005847954e-06,
+      "loss": 1.8136,
+      "step": 17648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.974977970123291,
+      "learning_rate": 3.8733552631578955e-06,
+      "loss": 1.6786,
+      "step": 17649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.637605905532837,
+      "learning_rate": 3.872441520467837e-06,
+      "loss": 1.3126,
+      "step": 17650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.070462465286255,
+      "learning_rate": 3.871527777777778e-06,
+      "loss": 1.4441,
+      "step": 17651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.01094126701355,
+      "learning_rate": 3.87061403508772e-06,
+      "loss": 1.8544,
+      "step": 17652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5309200286865234,
+      "learning_rate": 3.869700292397661e-06,
+      "loss": 1.6674,
+      "step": 17653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.252269983291626,
+      "learning_rate": 3.868786549707602e-06,
+      "loss": 1.3326,
+      "step": 17654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.223644495010376,
+      "learning_rate": 3.867872807017545e-06,
+      "loss": 1.1956,
+      "step": 17655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5846126079559326,
+      "learning_rate": 3.866959064327486e-06,
+      "loss": 1.6019,
+      "step": 17656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0255870819091797,
+      "learning_rate": 3.866045321637427e-06,
+      "loss": 1.678,
+      "step": 17657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.806856870651245,
+      "learning_rate": 3.865131578947369e-06,
+      "loss": 1.5101,
+      "step": 17658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.588794708251953,
+      "learning_rate": 3.86421783625731e-06,
+      "loss": 1.6185,
+      "step": 17659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3651316165924072,
+      "learning_rate": 3.8633040935672515e-06,
+      "loss": 1.3858,
+      "step": 17660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.085657835006714,
+      "learning_rate": 3.862390350877193e-06,
+      "loss": 1.1845,
+      "step": 17661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1657216548919678,
+      "learning_rate": 3.861476608187135e-06,
+      "loss": 1.667,
+      "step": 17662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5360798835754395,
+      "learning_rate": 3.8605628654970764e-06,
+      "loss": 1.5469,
+      "step": 17663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6157662868499756,
+      "learning_rate": 3.859649122807018e-06,
+      "loss": 1.469,
+      "step": 17664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0857040882110596,
+      "learning_rate": 3.858735380116959e-06,
+      "loss": 1.3153,
+      "step": 17665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4362285137176514,
+      "learning_rate": 3.8578216374269005e-06,
+      "loss": 1.5953,
+      "step": 17666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4749717712402344,
+      "learning_rate": 3.856907894736843e-06,
+      "loss": 1.2383,
+      "step": 17667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0994389057159424,
+      "learning_rate": 3.855994152046784e-06,
+      "loss": 1.2642,
+      "step": 17668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7683162689208984,
+      "learning_rate": 3.8550804093567255e-06,
+      "loss": 1.8402,
+      "step": 17669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6812801361083984,
+      "learning_rate": 3.854166666666667e-06,
+      "loss": 1.6144,
+      "step": 17670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.292085886001587,
+      "learning_rate": 3.853252923976608e-06,
+      "loss": 1.2519,
+      "step": 17671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.05483078956604,
+      "learning_rate": 3.8523391812865505e-06,
+      "loss": 1.238,
+      "step": 17672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.214745044708252,
+      "learning_rate": 3.851425438596492e-06,
+      "loss": 1.2512,
+      "step": 17673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6254971027374268,
+      "learning_rate": 3.850511695906433e-06,
+      "loss": 1.5295,
+      "step": 17674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0225861072540283,
+      "learning_rate": 3.849597953216375e-06,
+      "loss": 1.7914,
+      "step": 17675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9369258880615234,
+      "learning_rate": 3.848684210526316e-06,
+      "loss": 1.6364,
+      "step": 17676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.424163341522217,
+      "learning_rate": 3.847770467836257e-06,
+      "loss": 1.3834,
+      "step": 17677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2088820934295654,
+      "learning_rate": 3.846856725146199e-06,
+      "loss": 1.8788,
+      "step": 17678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4586169719696045,
+      "learning_rate": 3.845942982456141e-06,
+      "loss": 1.7317,
+      "step": 17679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4425721168518066,
+      "learning_rate": 3.845029239766082e-06,
+      "loss": 1.259,
+      "step": 17680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.399808645248413,
+      "learning_rate": 3.844115497076024e-06,
+      "loss": 1.2567,
+      "step": 17681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.550644636154175,
+      "learning_rate": 3.843201754385965e-06,
+      "loss": 1.3715,
+      "step": 17682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.666680335998535,
+      "learning_rate": 3.8422880116959065e-06,
+      "loss": 1.3213,
+      "step": 17683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.693293809890747,
+      "learning_rate": 3.841374269005848e-06,
+      "loss": 1.2636,
+      "step": 17684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2206969261169434,
+      "learning_rate": 3.840460526315789e-06,
+      "loss": 1.3332,
+      "step": 17685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.533127784729004,
+      "learning_rate": 3.8395467836257314e-06,
+      "loss": 1.4129,
+      "step": 17686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9947181940078735,
+      "learning_rate": 3.838633040935673e-06,
+      "loss": 1.1054,
+      "step": 17687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2866370677948,
+      "learning_rate": 3.837719298245615e-06,
+      "loss": 1.2545,
+      "step": 17688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4721126556396484,
+      "learning_rate": 3.836805555555556e-06,
+      "loss": 1.4883,
+      "step": 17689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.140324592590332,
+      "learning_rate": 3.835891812865498e-06,
+      "loss": 1.1532,
+      "step": 17690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1685750484466553,
+      "learning_rate": 3.834978070175439e-06,
+      "loss": 1.4467,
+      "step": 17691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1401360034942627,
+      "learning_rate": 3.8340643274853805e-06,
+      "loss": 1.2999,
+      "step": 17692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4595541954040527,
+      "learning_rate": 3.833150584795322e-06,
+      "loss": 1.4614,
+      "step": 17693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2605631351470947,
+      "learning_rate": 3.832236842105263e-06,
+      "loss": 1.3197,
+      "step": 17694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8678410053253174,
+      "learning_rate": 3.8313230994152055e-06,
+      "loss": 1.3843,
+      "step": 17695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9487698078155518,
+      "learning_rate": 3.830409356725147e-06,
+      "loss": 1.5923,
+      "step": 17696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4201009273529053,
+      "learning_rate": 3.829495614035088e-06,
+      "loss": 1.2731,
+      "step": 17697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6797635555267334,
+      "learning_rate": 3.82858187134503e-06,
+      "loss": 1.4638,
+      "step": 17698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.731693744659424,
+      "learning_rate": 3.827668128654971e-06,
+      "loss": 1.4971,
+      "step": 17699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4718539714813232,
+      "learning_rate": 3.826754385964912e-06,
+      "loss": 1.5384,
+      "step": 17700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5635695457458496,
+      "learning_rate": 3.825840643274854e-06,
+      "loss": 1.5584,
+      "step": 17701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.631802558898926,
+      "learning_rate": 3.824926900584796e-06,
+      "loss": 1.3225,
+      "step": 17702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.369792938232422,
+      "learning_rate": 3.824013157894737e-06,
+      "loss": 1.1776,
+      "step": 17703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2937214374542236,
+      "learning_rate": 3.823099415204679e-06,
+      "loss": 1.3446,
+      "step": 17704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4485723972320557,
+      "learning_rate": 3.82218567251462e-06,
+      "loss": 1.4653,
+      "step": 17705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2458338737487793,
+      "learning_rate": 3.8212719298245615e-06,
+      "loss": 1.191,
+      "step": 17706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8149843215942383,
+      "learning_rate": 3.820358187134503e-06,
+      "loss": 1.4758,
+      "step": 17707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.446645975112915,
+      "learning_rate": 3.819444444444444e-06,
+      "loss": 1.7452,
+      "step": 17708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4291322231292725,
+      "learning_rate": 3.8185307017543864e-06,
+      "loss": 1.5332,
+      "step": 17709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.118406295776367,
+      "learning_rate": 3.817616959064328e-06,
+      "loss": 1.3555,
+      "step": 17710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9728541374206543,
+      "learning_rate": 3.816703216374269e-06,
+      "loss": 2.0032,
+      "step": 17711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4355809688568115,
+      "learning_rate": 3.815789473684211e-06,
+      "loss": 1.4738,
+      "step": 17712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.765740156173706,
+      "learning_rate": 3.814875730994153e-06,
+      "loss": 1.6359,
+      "step": 17713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.00420880317688,
+      "learning_rate": 3.813961988304094e-06,
+      "loss": 1.795,
+      "step": 17714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.311685562133789,
+      "learning_rate": 3.8130482456140355e-06,
+      "loss": 1.3095,
+      "step": 17715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.819902181625366,
+      "learning_rate": 3.812134502923977e-06,
+      "loss": 1.5828,
+      "step": 17716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7802040576934814,
+      "learning_rate": 3.8112207602339187e-06,
+      "loss": 1.408,
+      "step": 17717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4995357990264893,
+      "learning_rate": 3.81030701754386e-06,
+      "loss": 1.4863,
+      "step": 17718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2094805240631104,
+      "learning_rate": 3.8093932748538014e-06,
+      "loss": 1.5296,
+      "step": 17719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4328436851501465,
+      "learning_rate": 3.808479532163743e-06,
+      "loss": 2.1288,
+      "step": 17720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7013959884643555,
+      "learning_rate": 3.8075657894736846e-06,
+      "loss": 1.4039,
+      "step": 17721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9125399589538574,
+      "learning_rate": 3.806652046783626e-06,
+      "loss": 1.6638,
+      "step": 17722
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0956625938415527,
+      "learning_rate": 3.8057383040935674e-06,
+      "loss": 1.3228,
+      "step": 17723
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4020025730133057,
+      "learning_rate": 3.804824561403509e-06,
+      "loss": 1.2357,
+      "step": 17724
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3679540157318115,
+      "learning_rate": 3.8039108187134505e-06,
+      "loss": 1.3762,
+      "step": 17725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1164321899414062,
+      "learning_rate": 3.802997076023392e-06,
+      "loss": 1.2862,
+      "step": 17726
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.564025640487671,
+      "learning_rate": 3.8020833333333333e-06,
+      "loss": 1.8541,
+      "step": 17727
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.435056447982788,
+      "learning_rate": 3.801169590643275e-06,
+      "loss": 1.3496,
+      "step": 17728
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.983454465866089,
+      "learning_rate": 3.8002558479532165e-06,
+      "loss": 1.6832,
+      "step": 17729
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6829893589019775,
+      "learning_rate": 3.799342105263158e-06,
+      "loss": 1.6581,
+      "step": 17730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0238358974456787,
+      "learning_rate": 3.7984283625731e-06,
+      "loss": 2.0095,
+      "step": 17731
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.204589366912842,
+      "learning_rate": 3.7975146198830414e-06,
+      "loss": 1.1446,
+      "step": 17732
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.479949951171875,
+      "learning_rate": 3.796600877192983e-06,
+      "loss": 1.4416,
+      "step": 17733
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.878047227859497,
+      "learning_rate": 3.7956871345029246e-06,
+      "loss": 1.796,
+      "step": 17734
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.260129451751709,
+      "learning_rate": 3.794773391812866e-06,
+      "loss": 1.4304,
+      "step": 17735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.832639694213867,
+      "learning_rate": 3.7938596491228073e-06,
+      "loss": 1.4488,
+      "step": 17736
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3666372299194336,
+      "learning_rate": 3.792945906432749e-06,
+      "loss": 1.593,
+      "step": 17737
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.713268518447876,
+      "learning_rate": 3.7920321637426905e-06,
+      "loss": 1.8317,
+      "step": 17738
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.472071886062622,
+      "learning_rate": 3.791118421052632e-06,
+      "loss": 1.2586,
+      "step": 17739
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6664857864379883,
+      "learning_rate": 3.7902046783625733e-06,
+      "loss": 1.1604,
+      "step": 17740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0101144313812256,
+      "learning_rate": 3.789290935672515e-06,
+      "loss": 1.9242,
+      "step": 17741
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.307924509048462,
+      "learning_rate": 3.7883771929824564e-06,
+      "loss": 1.3894,
+      "step": 17742
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.710031509399414,
+      "learning_rate": 3.787463450292398e-06,
+      "loss": 1.7001,
+      "step": 17743
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2024085521698,
+      "learning_rate": 3.7865497076023396e-06,
+      "loss": 1.2072,
+      "step": 17744
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6189863681793213,
+      "learning_rate": 3.785635964912281e-06,
+      "loss": 1.4279,
+      "step": 17745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6679434776306152,
+      "learning_rate": 3.7847222222222224e-06,
+      "loss": 1.3097,
+      "step": 17746
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.293205976486206,
+      "learning_rate": 3.7838084795321637e-06,
+      "loss": 1.577,
+      "step": 17747
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.588820457458496,
+      "learning_rate": 3.7828947368421055e-06,
+      "loss": 1.4351,
+      "step": 17748
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.435814380645752,
+      "learning_rate": 3.781980994152047e-06,
+      "loss": 1.5486,
+      "step": 17749
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.404649496078491,
+      "learning_rate": 3.7810672514619883e-06,
+      "loss": 1.3386,
+      "step": 17750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2188613414764404,
+      "learning_rate": 3.78015350877193e-06,
+      "loss": 1.9905,
+      "step": 17751
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5594427585601807,
+      "learning_rate": 3.779239766081872e-06,
+      "loss": 1.347,
+      "step": 17752
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.601151466369629,
+      "learning_rate": 3.7783260233918133e-06,
+      "loss": 1.371,
+      "step": 17753
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6791441440582275,
+      "learning_rate": 3.777412280701755e-06,
+      "loss": 1.4425,
+      "step": 17754
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1135079860687256,
+      "learning_rate": 3.7764985380116964e-06,
+      "loss": 1.6001,
+      "step": 17755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6239943504333496,
+      "learning_rate": 3.775584795321638e-06,
+      "loss": 1.3117,
+      "step": 17756
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5585694313049316,
+      "learning_rate": 3.7746710526315796e-06,
+      "loss": 1.5412,
+      "step": 17757
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1202380657196045,
+      "learning_rate": 3.773757309941521e-06,
+      "loss": 1.3726,
+      "step": 17758
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.83011531829834,
+      "learning_rate": 3.7728435672514623e-06,
+      "loss": 1.7332,
+      "step": 17759
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9407777786254883,
+      "learning_rate": 3.7719298245614037e-06,
+      "loss": 1.4386,
+      "step": 17760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7382466793060303,
+      "learning_rate": 3.7710160818713455e-06,
+      "loss": 1.6378,
+      "step": 17761
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2638015747070312,
+      "learning_rate": 3.770102339181287e-06,
+      "loss": 1.5258,
+      "step": 17762
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.346992015838623,
+      "learning_rate": 3.7691885964912283e-06,
+      "loss": 1.3015,
+      "step": 17763
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4500207901000977,
+      "learning_rate": 3.76827485380117e-06,
+      "loss": 1.2308,
+      "step": 17764
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.041978120803833,
+      "learning_rate": 3.7673611111111114e-06,
+      "loss": 1.1135,
+      "step": 17765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0251197814941406,
+      "learning_rate": 3.766447368421053e-06,
+      "loss": 1.9474,
+      "step": 17766
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5601515769958496,
+      "learning_rate": 3.765533625730994e-06,
+      "loss": 1.5802,
+      "step": 17767
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.149491548538208,
+      "learning_rate": 3.764619883040936e-06,
+      "loss": 0.9951,
+      "step": 17768
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9875884056091309,
+      "learning_rate": 3.7637061403508774e-06,
+      "loss": 1.2996,
+      "step": 17769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2923800945281982,
+      "learning_rate": 3.7627923976608187e-06,
+      "loss": 1.3681,
+      "step": 17770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5488345623016357,
+      "learning_rate": 3.7618786549707605e-06,
+      "loss": 1.3365,
+      "step": 17771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4107463359832764,
+      "learning_rate": 3.760964912280702e-06,
+      "loss": 1.3157,
+      "step": 17772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.489067792892456,
+      "learning_rate": 3.7600511695906433e-06,
+      "loss": 1.8228,
+      "step": 17773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4599735736846924,
+      "learning_rate": 3.7591374269005855e-06,
+      "loss": 1.5447,
+      "step": 17774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.544757843017578,
+      "learning_rate": 3.758223684210527e-06,
+      "loss": 1.5947,
+      "step": 17775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2575387954711914,
+      "learning_rate": 3.7573099415204683e-06,
+      "loss": 1.5076,
+      "step": 17776
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3977012634277344,
+      "learning_rate": 3.75639619883041e-06,
+      "loss": 1.5421,
+      "step": 17777
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4145162105560303,
+      "learning_rate": 3.7554824561403514e-06,
+      "loss": 1.28,
+      "step": 17778
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.831510066986084,
+      "learning_rate": 3.754568713450293e-06,
+      "loss": 1.6384,
+      "step": 17779
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6666691303253174,
+      "learning_rate": 3.753654970760234e-06,
+      "loss": 1.6925,
+      "step": 17780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9266135692596436,
+      "learning_rate": 3.752741228070176e-06,
+      "loss": 1.1182,
+      "step": 17781
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.070197343826294,
+      "learning_rate": 3.7518274853801173e-06,
+      "loss": 1.2892,
+      "step": 17782
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1483490467071533,
+      "learning_rate": 3.7509137426900587e-06,
+      "loss": 1.3314,
+      "step": 17783
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5931997299194336,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.5465,
+      "step": 17784
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0201120376586914,
+      "learning_rate": 3.749086257309942e-06,
+      "loss": 0.9144,
+      "step": 17785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5984299182891846,
+      "learning_rate": 3.7481725146198833e-06,
+      "loss": 1.4525,
+      "step": 17786
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9286599159240723,
+      "learning_rate": 3.7472587719298246e-06,
+      "loss": 1.7076,
+      "step": 17787
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2682759761810303,
+      "learning_rate": 3.7463450292397664e-06,
+      "loss": 1.5822,
+      "step": 17788
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3144466876983643,
+      "learning_rate": 3.745431286549708e-06,
+      "loss": 1.3763,
+      "step": 17789
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3160200119018555,
+      "learning_rate": 3.744517543859649e-06,
+      "loss": 1.3398,
+      "step": 17790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8019611835479736,
+      "learning_rate": 3.7436038011695906e-06,
+      "loss": 1.8158,
+      "step": 17791
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.719318389892578,
+      "learning_rate": 3.7426900584795324e-06,
+      "loss": 1.4337,
+      "step": 17792
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7376081943511963,
+      "learning_rate": 3.7417763157894737e-06,
+      "loss": 1.3997,
+      "step": 17793
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8447494506835938,
+      "learning_rate": 3.740862573099415e-06,
+      "loss": 1.4679,
+      "step": 17794
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5757761001586914,
+      "learning_rate": 3.7399488304093573e-06,
+      "loss": 1.388,
+      "step": 17795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.271085023880005,
+      "learning_rate": 3.7390350877192987e-06,
+      "loss": 1.2317,
+      "step": 17796
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.835724115371704,
+      "learning_rate": 3.73812134502924e-06,
+      "loss": 1.6876,
+      "step": 17797
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3454291820526123,
+      "learning_rate": 3.737207602339182e-06,
+      "loss": 0.9428,
+      "step": 17798
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.418931722640991,
+      "learning_rate": 3.7362938596491233e-06,
+      "loss": 1.6281,
+      "step": 17799
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.272094964981079,
+      "learning_rate": 3.7353801169590646e-06,
+      "loss": 1.8022,
+      "step": 17800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4801535606384277,
+      "learning_rate": 3.7344663742690064e-06,
+      "loss": 1.3813,
+      "step": 17801
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.671543598175049,
+      "learning_rate": 3.733552631578948e-06,
+      "loss": 1.474,
+      "step": 17802
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.571977376937866,
+      "learning_rate": 3.732638888888889e-06,
+      "loss": 1.7191,
+      "step": 17803
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1318061351776123,
+      "learning_rate": 3.7317251461988305e-06,
+      "loss": 1.3995,
+      "step": 17804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1351985931396484,
+      "learning_rate": 3.7308114035087723e-06,
+      "loss": 1.9965,
+      "step": 17805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5097622871398926,
+      "learning_rate": 3.7298976608187137e-06,
+      "loss": 1.4132,
+      "step": 17806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.828761577606201,
+      "learning_rate": 3.728983918128655e-06,
+      "loss": 1.9225,
+      "step": 17807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7246153354644775,
+      "learning_rate": 3.728070175438597e-06,
+      "loss": 1.4785,
+      "step": 17808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8383841514587402,
+      "learning_rate": 3.7271564327485383e-06,
+      "loss": 1.564,
+      "step": 17809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2366421222686768,
+      "learning_rate": 3.7262426900584796e-06,
+      "loss": 1.1716,
+      "step": 17810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5569586753845215,
+      "learning_rate": 3.725328947368421e-06,
+      "loss": 1.4793,
+      "step": 17811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1317334175109863,
+      "learning_rate": 3.724415204678363e-06,
+      "loss": 0.9975,
+      "step": 17812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7859113216400146,
+      "learning_rate": 3.723501461988304e-06,
+      "loss": 1.386,
+      "step": 17813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3506112098693848,
+      "learning_rate": 3.7225877192982456e-06,
+      "loss": 1.2017,
+      "step": 17814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3031089305877686,
+      "learning_rate": 3.7216739766081874e-06,
+      "loss": 1.1869,
+      "step": 17815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.409214496612549,
+      "learning_rate": 3.7207602339181287e-06,
+      "loss": 1.6134,
+      "step": 17816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5706372261047363,
+      "learning_rate": 3.7198464912280705e-06,
+      "loss": 1.2358,
+      "step": 17817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.319356918334961,
+      "learning_rate": 3.7189327485380123e-06,
+      "loss": 1.4688,
+      "step": 17818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.396548271179199,
+      "learning_rate": 3.7180190058479537e-06,
+      "loss": 1.3778,
+      "step": 17819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3560826778411865,
+      "learning_rate": 3.717105263157895e-06,
+      "loss": 1.3955,
+      "step": 17820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3157870769500732,
+      "learning_rate": 3.716191520467837e-06,
+      "loss": 1.4249,
+      "step": 17821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.011570453643799,
+      "learning_rate": 3.7152777777777783e-06,
+      "loss": 1.2434,
+      "step": 17822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8383331298828125,
+      "learning_rate": 3.7143640350877196e-06,
+      "loss": 1.5438,
+      "step": 17823
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2941877841949463,
+      "learning_rate": 3.713450292397661e-06,
+      "loss": 1.4815,
+      "step": 17824
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.777473211288452,
+      "learning_rate": 3.712536549707603e-06,
+      "loss": 1.7006,
+      "step": 17825
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.047034740447998,
+      "learning_rate": 3.711622807017544e-06,
+      "loss": 1.524,
+      "step": 17826
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6908183097839355,
+      "learning_rate": 3.7107090643274855e-06,
+      "loss": 1.4592,
+      "step": 17827
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.119475841522217,
+      "learning_rate": 3.7097953216374273e-06,
+      "loss": 1.3487,
+      "step": 17828
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4104061126708984,
+      "learning_rate": 3.7088815789473687e-06,
+      "loss": 1.274,
+      "step": 17829
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4756410121917725,
+      "learning_rate": 3.70796783625731e-06,
+      "loss": 1.3017,
+      "step": 17830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4178426265716553,
+      "learning_rate": 3.7070540935672515e-06,
+      "loss": 1.3689,
+      "step": 17831
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2112879753112793,
+      "learning_rate": 3.7061403508771933e-06,
+      "loss": 1.0563,
+      "step": 17832
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0089287757873535,
+      "learning_rate": 3.7052266081871346e-06,
+      "loss": 1.1703,
+      "step": 17833
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.835552215576172,
+      "learning_rate": 3.704312865497076e-06,
+      "loss": 1.6019,
+      "step": 17834
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0485622882843018,
+      "learning_rate": 3.703399122807018e-06,
+      "loss": 1.1631,
+      "step": 17835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2123475074768066,
+      "learning_rate": 3.702485380116959e-06,
+      "loss": 1.9225,
+      "step": 17836
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4474895000457764,
+      "learning_rate": 3.7015716374269006e-06,
+      "loss": 1.3787,
+      "step": 17837
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6296684741973877,
+      "learning_rate": 3.7006578947368428e-06,
+      "loss": 1.7008,
+      "step": 17838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.617755889892578,
+      "learning_rate": 3.699744152046784e-06,
+      "loss": 1.3544,
+      "step": 17839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.000478506088257,
+      "learning_rate": 3.6988304093567255e-06,
+      "loss": 1.2329,
+      "step": 17840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.216049909591675,
+      "learning_rate": 3.6979166666666673e-06,
+      "loss": 1.3861,
+      "step": 17841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0050172805786133,
+      "learning_rate": 3.6970029239766087e-06,
+      "loss": 1.1527,
+      "step": 17842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.366410732269287,
+      "learning_rate": 3.69608918128655e-06,
+      "loss": 1.4139,
+      "step": 17843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.261481523513794,
+      "learning_rate": 3.6951754385964915e-06,
+      "loss": 1.9612,
+      "step": 17844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.617600917816162,
+      "learning_rate": 3.6942616959064333e-06,
+      "loss": 1.5549,
+      "step": 17845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5292463302612305,
+      "learning_rate": 3.6933479532163746e-06,
+      "loss": 1.4432,
+      "step": 17846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1488595008850098,
+      "learning_rate": 3.692434210526316e-06,
+      "loss": 1.6358,
+      "step": 17847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.554777145385742,
+      "learning_rate": 3.691520467836258e-06,
+      "loss": 1.336,
+      "step": 17848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9916296005249023,
+      "learning_rate": 3.690606725146199e-06,
+      "loss": 1.9192,
+      "step": 17849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.580164909362793,
+      "learning_rate": 3.6896929824561405e-06,
+      "loss": 1.7295,
+      "step": 17850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0554537773132324,
+      "learning_rate": 3.688779239766082e-06,
+      "loss": 1.455,
+      "step": 17851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9848098754882812,
+      "learning_rate": 3.6878654970760237e-06,
+      "loss": 1.7574,
+      "step": 17852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.761256456375122,
+      "learning_rate": 3.686951754385965e-06,
+      "loss": 1.5313,
+      "step": 17853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0855181217193604,
+      "learning_rate": 3.6860380116959065e-06,
+      "loss": 1.6428,
+      "step": 17854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.103257179260254,
+      "learning_rate": 3.6851242690058483e-06,
+      "loss": 1.042,
+      "step": 17855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.820840358734131,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 1.5629,
+      "step": 17856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0142688751220703,
+      "learning_rate": 3.683296783625731e-06,
+      "loss": 2.1289,
+      "step": 17857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4801113605499268,
+      "learning_rate": 3.6823830409356724e-06,
+      "loss": 1.4387,
+      "step": 17858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.781144380569458,
+      "learning_rate": 3.6814692982456146e-06,
+      "loss": 1.7076,
+      "step": 17859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2518138885498047,
+      "learning_rate": 3.680555555555556e-06,
+      "loss": 1.2206,
+      "step": 17860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.217430830001831,
+      "learning_rate": 3.6796418128654978e-06,
+      "loss": 1.2293,
+      "step": 17861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1807916164398193,
+      "learning_rate": 3.678728070175439e-06,
+      "loss": 1.3624,
+      "step": 17862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2092409133911133,
+      "learning_rate": 3.6778143274853805e-06,
+      "loss": 1.1949,
+      "step": 17863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.606346368789673,
+      "learning_rate": 3.676900584795322e-06,
+      "loss": 1.4086,
+      "step": 17864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6416876316070557,
+      "learning_rate": 3.6759868421052637e-06,
+      "loss": 1.4519,
+      "step": 17865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8160550594329834,
+      "learning_rate": 3.675073099415205e-06,
+      "loss": 1.4795,
+      "step": 17866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.40183424949646,
+      "learning_rate": 3.6741593567251465e-06,
+      "loss": 1.4478,
+      "step": 17867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9947819709777832,
+      "learning_rate": 3.6732456140350883e-06,
+      "loss": 1.1215,
+      "step": 17868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.307992458343506,
+      "learning_rate": 3.6723318713450296e-06,
+      "loss": 1.4027,
+      "step": 17869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9331833124160767,
+      "learning_rate": 3.671418128654971e-06,
+      "loss": 0.9927,
+      "step": 17870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6481971740722656,
+      "learning_rate": 3.6705043859649124e-06,
+      "loss": 1.5836,
+      "step": 17871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0348565578460693,
+      "learning_rate": 3.669590643274854e-06,
+      "loss": 1.4318,
+      "step": 17872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3450043201446533,
+      "learning_rate": 3.6686769005847955e-06,
+      "loss": 1.4041,
+      "step": 17873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9846417903900146,
+      "learning_rate": 3.667763157894737e-06,
+      "loss": 1.7873,
+      "step": 17874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.152181386947632,
+      "learning_rate": 3.6668494152046787e-06,
+      "loss": 1.0936,
+      "step": 17875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8204758167266846,
+      "learning_rate": 3.66593567251462e-06,
+      "loss": 1.4153,
+      "step": 17876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.40289044380188,
+      "learning_rate": 3.6650219298245615e-06,
+      "loss": 1.5138,
+      "step": 17877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.093254566192627,
+      "learning_rate": 3.664108187134503e-06,
+      "loss": 2.0154,
+      "step": 17878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6538753509521484,
+      "learning_rate": 3.6631944444444446e-06,
+      "loss": 1.4445,
+      "step": 17879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7753190994262695,
+      "learning_rate": 3.662280701754386e-06,
+      "loss": 1.7601,
+      "step": 17880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2396368980407715,
+      "learning_rate": 3.6613669590643282e-06,
+      "loss": 1.1609,
+      "step": 17881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4587745666503906,
+      "learning_rate": 3.6604532163742696e-06,
+      "loss": 1.7151,
+      "step": 17882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.615402936935425,
+      "learning_rate": 3.659539473684211e-06,
+      "loss": 1.5677,
+      "step": 17883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4887032508850098,
+      "learning_rate": 3.6586257309941524e-06,
+      "loss": 1.1144,
+      "step": 17884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5316500663757324,
+      "learning_rate": 3.657711988304094e-06,
+      "loss": 1.4636,
+      "step": 17885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.789527416229248,
+      "learning_rate": 3.6567982456140355e-06,
+      "loss": 1.4138,
+      "step": 17886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5468246936798096,
+      "learning_rate": 3.655884502923977e-06,
+      "loss": 1.5849,
+      "step": 17887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8940041065216064,
+      "learning_rate": 3.6549707602339187e-06,
+      "loss": 1.8816,
+      "step": 17888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.432079792022705,
+      "learning_rate": 3.65405701754386e-06,
+      "loss": 1.5062,
+      "step": 17889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.212449312210083,
+      "learning_rate": 3.6531432748538015e-06,
+      "loss": 1.2158,
+      "step": 17890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.721452474594116,
+      "learning_rate": 3.652229532163743e-06,
+      "loss": 1.4934,
+      "step": 17891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1801986694335938,
+      "learning_rate": 3.6513157894736846e-06,
+      "loss": 1.2756,
+      "step": 17892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.05492901802063,
+      "learning_rate": 3.650402046783626e-06,
+      "loss": 1.2799,
+      "step": 17893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3628721237182617,
+      "learning_rate": 3.6494883040935674e-06,
+      "loss": 1.5589,
+      "step": 17894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1020989418029785,
+      "learning_rate": 3.648574561403509e-06,
+      "loss": 1.3109,
+      "step": 17895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.343531847000122,
+      "learning_rate": 3.6476608187134505e-06,
+      "loss": 1.6553,
+      "step": 17896
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1202242374420166,
+      "learning_rate": 3.646747076023392e-06,
+      "loss": 1.2203,
+      "step": 17897
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.890873670578003,
+      "learning_rate": 3.6458333333333333e-06,
+      "loss": 1.6602,
+      "step": 17898
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1312930583953857,
+      "learning_rate": 3.644919590643275e-06,
+      "loss": 1.1886,
+      "step": 17899
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.73787784576416,
+      "learning_rate": 3.6440058479532165e-06,
+      "loss": 1.5561,
+      "step": 17900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.13874888420105,
+      "learning_rate": 3.643092105263158e-06,
+      "loss": 1.1295,
+      "step": 17901
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.001577854156494,
+      "learning_rate": 3.6421783625731e-06,
+      "loss": 1.1459,
+      "step": 17902
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5099375247955322,
+      "learning_rate": 3.6412646198830414e-06,
+      "loss": 1.5555,
+      "step": 17903
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6416478157043457,
+      "learning_rate": 3.640350877192983e-06,
+      "loss": 1.2807,
+      "step": 17904
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.136983633041382,
+      "learning_rate": 3.6394371345029246e-06,
+      "loss": 1.2274,
+      "step": 17905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0718448162078857,
+      "learning_rate": 3.638523391812866e-06,
+      "loss": 1.1911,
+      "step": 17906
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.41609525680542,
+      "learning_rate": 3.6376096491228074e-06,
+      "loss": 1.306,
+      "step": 17907
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.95633065700531,
+      "learning_rate": 3.6366959064327487e-06,
+      "loss": 0.9535,
+      "step": 17908
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5868282318115234,
+      "learning_rate": 3.6357821637426905e-06,
+      "loss": 1.4512,
+      "step": 17909
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.640723943710327,
+      "learning_rate": 3.634868421052632e-06,
+      "loss": 1.5529,
+      "step": 17910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.098550796508789,
+      "learning_rate": 3.6339546783625733e-06,
+      "loss": 1.4937,
+      "step": 17911
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4506800174713135,
+      "learning_rate": 3.633040935672515e-06,
+      "loss": 1.2758,
+      "step": 17912
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.109570026397705,
+      "learning_rate": 3.6321271929824565e-06,
+      "loss": 1.2818,
+      "step": 17913
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.368960380554199,
+      "learning_rate": 3.631213450292398e-06,
+      "loss": 1.6686,
+      "step": 17914
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3466832637786865,
+      "learning_rate": 3.630299707602339e-06,
+      "loss": 1.3214,
+      "step": 17915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6918282508850098,
+      "learning_rate": 3.629385964912281e-06,
+      "loss": 1.3507,
+      "step": 17916
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9734920263290405,
+      "learning_rate": 3.6284722222222224e-06,
+      "loss": 1.204,
+      "step": 17917
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.641577959060669,
+      "learning_rate": 3.6275584795321637e-06,
+      "loss": 1.7696,
+      "step": 17918
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.95115065574646,
+      "learning_rate": 3.6266447368421055e-06,
+      "loss": 1.8293,
+      "step": 17919
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.635843515396118,
+      "learning_rate": 3.625730994152047e-06,
+      "loss": 1.351,
+      "step": 17920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.310044527053833,
+      "learning_rate": 3.6248172514619883e-06,
+      "loss": 1.2821,
+      "step": 17921
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8043558597564697,
+      "learning_rate": 3.6239035087719297e-06,
+      "loss": 1.1861,
+      "step": 17922
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7546582221984863,
+      "learning_rate": 3.622989766081872e-06,
+      "loss": 1.5002,
+      "step": 17923
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.386730670928955,
+      "learning_rate": 3.6220760233918133e-06,
+      "loss": 1.0833,
+      "step": 17924
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.483328104019165,
+      "learning_rate": 3.621162280701755e-06,
+      "loss": 1.0915,
+      "step": 17925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9823994636535645,
+      "learning_rate": 3.6202485380116964e-06,
+      "loss": 1.5457,
+      "step": 17926
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6578381061553955,
+      "learning_rate": 3.619334795321638e-06,
+      "loss": 1.5489,
+      "step": 17927
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4566752910614014,
+      "learning_rate": 3.618421052631579e-06,
+      "loss": 1.7144,
+      "step": 17928
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7616193294525146,
+      "learning_rate": 3.617507309941521e-06,
+      "loss": 1.5049,
+      "step": 17929
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2269954681396484,
+      "learning_rate": 3.6165935672514624e-06,
+      "loss": 1.5519,
+      "step": 17930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.64650821685791,
+      "learning_rate": 3.6156798245614037e-06,
+      "loss": 1.3962,
+      "step": 17931
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5529544353485107,
+      "learning_rate": 3.6147660818713455e-06,
+      "loss": 1.59,
+      "step": 17932
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.560716390609741,
+      "learning_rate": 3.613852339181287e-06,
+      "loss": 1.35,
+      "step": 17933
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4317209720611572,
+      "learning_rate": 3.6129385964912283e-06,
+      "loss": 1.1519,
+      "step": 17934
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.25911808013916,
+      "learning_rate": 3.6120248538011697e-06,
+      "loss": 1.3793,
+      "step": 17935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.540452480316162,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 1.6159,
+      "step": 17936
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1533443927764893,
+      "learning_rate": 3.610197368421053e-06,
+      "loss": 1.7002,
+      "step": 17937
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.390554904937744,
+      "learning_rate": 3.609283625730994e-06,
+      "loss": 1.3514,
+      "step": 17938
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9781203269958496,
+      "learning_rate": 3.608369883040936e-06,
+      "loss": 1.7912,
+      "step": 17939
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1676723957061768,
+      "learning_rate": 3.6074561403508774e-06,
+      "loss": 1.4146,
+      "step": 17940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.428964138031006,
+      "learning_rate": 3.6065423976608187e-06,
+      "loss": 1.6814,
+      "step": 17941
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8332996368408203,
+      "learning_rate": 3.60562865497076e-06,
+      "loss": 1.8475,
+      "step": 17942
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2790961265563965,
+      "learning_rate": 3.604714912280702e-06,
+      "loss": 1.1616,
+      "step": 17943
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3664157390594482,
+      "learning_rate": 3.6038011695906433e-06,
+      "loss": 2.0266,
+      "step": 17944
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3592967987060547,
+      "learning_rate": 3.6028874269005855e-06,
+      "loss": 1.3294,
+      "step": 17945
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.340668201446533,
+      "learning_rate": 3.601973684210527e-06,
+      "loss": 1.3045,
+      "step": 17946
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9437034130096436,
+      "learning_rate": 3.6010599415204683e-06,
+      "loss": 1.8191,
+      "step": 17947
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.036916494369507,
+      "learning_rate": 3.6001461988304096e-06,
+      "loss": 2.0985,
+      "step": 17948
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.428959369659424,
+      "learning_rate": 3.5992324561403514e-06,
+      "loss": 1.5464,
+      "step": 17949
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.951430320739746,
+      "learning_rate": 3.598318713450293e-06,
+      "loss": 1.5905,
+      "step": 17950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7639567852020264,
+      "learning_rate": 3.597404970760234e-06,
+      "loss": 1.5936,
+      "step": 17951
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6135733127593994,
+      "learning_rate": 3.596491228070176e-06,
+      "loss": 1.6797,
+      "step": 17952
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4454357624053955,
+      "learning_rate": 3.5955774853801174e-06,
+      "loss": 1.4084,
+      "step": 17953
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4784743785858154,
+      "learning_rate": 3.5946637426900587e-06,
+      "loss": 1.4228,
+      "step": 17954
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.269526481628418,
+      "learning_rate": 3.59375e-06,
+      "loss": 1.512,
+      "step": 17955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.790748953819275,
+      "learning_rate": 3.592836257309942e-06,
+      "loss": 1.1695,
+      "step": 17956
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.806627035140991,
+      "learning_rate": 3.5919225146198833e-06,
+      "loss": 1.7488,
+      "step": 17957
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8184781074523926,
+      "learning_rate": 3.5910087719298247e-06,
+      "loss": 1.1429,
+      "step": 17958
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0733230113983154,
+      "learning_rate": 3.5900950292397665e-06,
+      "loss": 1.2028,
+      "step": 17959
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2424685955047607,
+      "learning_rate": 3.589181286549708e-06,
+      "loss": 1.2382,
+      "step": 17960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.417410135269165,
+      "learning_rate": 3.588267543859649e-06,
+      "loss": 1.37,
+      "step": 17961
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3940675258636475,
+      "learning_rate": 3.5873538011695906e-06,
+      "loss": 1.4993,
+      "step": 17962
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1839370727539062,
+      "learning_rate": 3.5864400584795324e-06,
+      "loss": 1.6721,
+      "step": 17963
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0083138942718506,
+      "learning_rate": 3.5855263157894737e-06,
+      "loss": 1.1745,
+      "step": 17964
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4439098834991455,
+      "learning_rate": 3.584612573099415e-06,
+      "loss": 1.5769,
+      "step": 17965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.216132640838623,
+      "learning_rate": 3.5836988304093573e-06,
+      "loss": 1.2311,
+      "step": 17966
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.527576208114624,
+      "learning_rate": 3.5827850877192987e-06,
+      "loss": 1.4643,
+      "step": 17967
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1213200092315674,
+      "learning_rate": 3.58187134502924e-06,
+      "loss": 1.3997,
+      "step": 17968
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5744636058807373,
+      "learning_rate": 3.580957602339182e-06,
+      "loss": 1.6617,
+      "step": 17969
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3764467239379883,
+      "learning_rate": 3.5800438596491233e-06,
+      "loss": 1.2083,
+      "step": 17970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7929277420043945,
+      "learning_rate": 3.5791301169590646e-06,
+      "loss": 1.8064,
+      "step": 17971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.186544895172119,
+      "learning_rate": 3.5782163742690064e-06,
+      "loss": 1.3587,
+      "step": 17972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0160601139068604,
+      "learning_rate": 3.577302631578948e-06,
+      "loss": 1.8563,
+      "step": 17973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7876250743865967,
+      "learning_rate": 3.576388888888889e-06,
+      "loss": 1.3044,
+      "step": 17974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.410698890686035,
+      "learning_rate": 3.5754751461988306e-06,
+      "loss": 1.372,
+      "step": 17975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7951762676239014,
+      "learning_rate": 3.5745614035087724e-06,
+      "loss": 1.7835,
+      "step": 17976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.78690242767334,
+      "learning_rate": 3.5736476608187137e-06,
+      "loss": 1.4087,
+      "step": 17977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.27876615524292,
+      "learning_rate": 3.572733918128655e-06,
+      "loss": 1.4438,
+      "step": 17978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.493661642074585,
+      "learning_rate": 3.571820175438597e-06,
+      "loss": 1.4579,
+      "step": 17979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.598639726638794,
+      "learning_rate": 3.5709064327485383e-06,
+      "loss": 1.533,
+      "step": 17980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1811020374298096,
+      "learning_rate": 3.5699926900584797e-06,
+      "loss": 1.1525,
+      "step": 17981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0817058086395264,
+      "learning_rate": 3.569078947368421e-06,
+      "loss": 1.4031,
+      "step": 17982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.348313331604004,
+      "learning_rate": 3.568165204678363e-06,
+      "loss": 1.414,
+      "step": 17983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.316619396209717,
+      "learning_rate": 3.567251461988304e-06,
+      "loss": 1.4999,
+      "step": 17984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8300352096557617,
+      "learning_rate": 3.5663377192982456e-06,
+      "loss": 1.4447,
+      "step": 17985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5403192043304443,
+      "learning_rate": 3.5654239766081874e-06,
+      "loss": 1.4219,
+      "step": 17986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5027761459350586,
+      "learning_rate": 3.5645102339181287e-06,
+      "loss": 1.5238,
+      "step": 17987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5699832439422607,
+      "learning_rate": 3.5635964912280705e-06,
+      "loss": 1.3856,
+      "step": 17988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.721637487411499,
+      "learning_rate": 3.5626827485380123e-06,
+      "loss": 1.3615,
+      "step": 17989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.336254358291626,
+      "learning_rate": 3.5617690058479537e-06,
+      "loss": 1.277,
+      "step": 17990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1455111503601074,
+      "learning_rate": 3.560855263157895e-06,
+      "loss": 1.1978,
+      "step": 17991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5532383918762207,
+      "learning_rate": 3.559941520467837e-06,
+      "loss": 1.2902,
+      "step": 17992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3159899711608887,
+      "learning_rate": 3.5590277777777783e-06,
+      "loss": 1.5042,
+      "step": 17993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7688512802124023,
+      "learning_rate": 3.5581140350877196e-06,
+      "loss": 1.6042,
+      "step": 17994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.838778018951416,
+      "learning_rate": 3.557200292397661e-06,
+      "loss": 1.7519,
+      "step": 17995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.834190607070923,
+      "learning_rate": 3.556286549707603e-06,
+      "loss": 1.3107,
+      "step": 17996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3479292392730713,
+      "learning_rate": 3.555372807017544e-06,
+      "loss": 1.4964,
+      "step": 17997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4796507358551025,
+      "learning_rate": 3.5544590643274856e-06,
+      "loss": 1.6398,
+      "step": 17998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5846588611602783,
+      "learning_rate": 3.5535453216374274e-06,
+      "loss": 1.467,
+      "step": 17999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.681443691253662,
+      "learning_rate": 3.5526315789473687e-06,
+      "loss": 1.6564,
+      "step": 18000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7552108764648438,
+      "learning_rate": 3.55171783625731e-06,
+      "loss": 1.4975,
+      "step": 18001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4574177265167236,
+      "learning_rate": 3.5508040935672515e-06,
+      "loss": 1.5276,
+      "step": 18002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9424242973327637,
+      "learning_rate": 3.5498903508771933e-06,
+      "loss": 1.1891,
+      "step": 18003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8720924854278564,
+      "learning_rate": 3.5489766081871347e-06,
+      "loss": 1.6537,
+      "step": 18004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.654711961746216,
+      "learning_rate": 3.548062865497076e-06,
+      "loss": 1.3394,
+      "step": 18005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2609567642211914,
+      "learning_rate": 3.547149122807018e-06,
+      "loss": 1.1765,
+      "step": 18006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.159731864929199,
+      "learning_rate": 3.546235380116959e-06,
+      "loss": 1.1489,
+      "step": 18007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.390136957168579,
+      "learning_rate": 3.5453216374269006e-06,
+      "loss": 1.1581,
+      "step": 18008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8028992414474487,
+      "learning_rate": 3.544407894736843e-06,
+      "loss": 1.1732,
+      "step": 18009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3457226753234863,
+      "learning_rate": 3.543494152046784e-06,
+      "loss": 1.5409,
+      "step": 18010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.517822265625,
+      "learning_rate": 3.5425804093567255e-06,
+      "loss": 1.6151,
+      "step": 18011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2385597229003906,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 1.4312,
+      "step": 18012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.541809320449829,
+      "learning_rate": 3.5407529239766087e-06,
+      "loss": 1.4766,
+      "step": 18013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9988460540771484,
+      "learning_rate": 3.53983918128655e-06,
+      "loss": 1.7876,
+      "step": 18014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3089394569396973,
+      "learning_rate": 3.5389254385964915e-06,
+      "loss": 1.2903,
+      "step": 18015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2879128456115723,
+      "learning_rate": 3.5380116959064333e-06,
+      "loss": 1.3696,
+      "step": 18016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9662318229675293,
+      "learning_rate": 3.5370979532163746e-06,
+      "loss": 1.6472,
+      "step": 18017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.672673225402832,
+      "learning_rate": 3.536184210526316e-06,
+      "loss": 1.4504,
+      "step": 18018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9135822057724,
+      "learning_rate": 3.535270467836258e-06,
+      "loss": 0.9689,
+      "step": 18019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6561241149902344,
+      "learning_rate": 3.534356725146199e-06,
+      "loss": 1.4518,
+      "step": 18020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7202975749969482,
+      "learning_rate": 3.5334429824561406e-06,
+      "loss": 1.3585,
+      "step": 18021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2401061058044434,
+      "learning_rate": 3.532529239766082e-06,
+      "loss": 1.4121,
+      "step": 18022
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.579679012298584,
+      "learning_rate": 3.5316154970760237e-06,
+      "loss": 1.3877,
+      "step": 18023
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2837626934051514,
+      "learning_rate": 3.530701754385965e-06,
+      "loss": 1.4716,
+      "step": 18024
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.425504684448242,
+      "learning_rate": 3.5297880116959065e-06,
+      "loss": 1.3916,
+      "step": 18025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.549760580062866,
+      "learning_rate": 3.528874269005848e-06,
+      "loss": 1.6946,
+      "step": 18026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0089385509490967,
+      "learning_rate": 3.5279605263157897e-06,
+      "loss": 1.1284,
+      "step": 18027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.342877149581909,
+      "learning_rate": 3.527046783625731e-06,
+      "loss": 1.3686,
+      "step": 18028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.294996976852417,
+      "learning_rate": 3.5261330409356724e-06,
+      "loss": 1.3663,
+      "step": 18029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0141735076904297,
+      "learning_rate": 3.5252192982456146e-06,
+      "loss": 1.2805,
+      "step": 18030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0041449069976807,
+      "learning_rate": 3.524305555555556e-06,
+      "loss": 1.818,
+      "step": 18031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.383418560028076,
+      "learning_rate": 3.5233918128654974e-06,
+      "loss": 1.8262,
+      "step": 18032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4867873191833496,
+      "learning_rate": 3.522478070175439e-06,
+      "loss": 1.5353,
+      "step": 18033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.498861074447632,
+      "learning_rate": 3.5215643274853805e-06,
+      "loss": 1.6457,
+      "step": 18034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6149487495422363,
+      "learning_rate": 3.520650584795322e-06,
+      "loss": 1.6576,
+      "step": 18035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1516239643096924,
+      "learning_rate": 3.5197368421052637e-06,
+      "loss": 1.548,
+      "step": 18036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9143741130828857,
+      "learning_rate": 3.518823099415205e-06,
+      "loss": 1.5835,
+      "step": 18037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8018648624420166,
+      "learning_rate": 3.5179093567251465e-06,
+      "loss": 1.6972,
+      "step": 18038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7510783672332764,
+      "learning_rate": 3.516995614035088e-06,
+      "loss": 1.5332,
+      "step": 18039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.998234272003174,
+      "learning_rate": 3.5160818713450296e-06,
+      "loss": 1.5048,
+      "step": 18040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5281906127929688,
+      "learning_rate": 3.515168128654971e-06,
+      "loss": 1.386,
+      "step": 18041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7290854454040527,
+      "learning_rate": 3.5142543859649124e-06,
+      "loss": 1.5932,
+      "step": 18042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4953129291534424,
+      "learning_rate": 3.513340643274854e-06,
+      "loss": 1.3576,
+      "step": 18043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.359860897064209,
+      "learning_rate": 3.5124269005847956e-06,
+      "loss": 1.5263,
+      "step": 18044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.224630832672119,
+      "learning_rate": 3.511513157894737e-06,
+      "loss": 1.2394,
+      "step": 18045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.895735263824463,
+      "learning_rate": 3.5105994152046783e-06,
+      "loss": 1.5522,
+      "step": 18046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4269065856933594,
+      "learning_rate": 3.50968567251462e-06,
+      "loss": 1.0928,
+      "step": 18047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2289085388183594,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 1.1813,
+      "step": 18048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3990612030029297,
+      "learning_rate": 3.507858187134503e-06,
+      "loss": 1.7663,
+      "step": 18049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.554682731628418,
+      "learning_rate": 3.5069444444444447e-06,
+      "loss": 1.4637,
+      "step": 18050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1709046363830566,
+      "learning_rate": 3.506030701754386e-06,
+      "loss": 1.6169,
+      "step": 18051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0376129150390625,
+      "learning_rate": 3.505116959064328e-06,
+      "loss": 1.6549,
+      "step": 18052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1650547981262207,
+      "learning_rate": 3.5042032163742696e-06,
+      "loss": 1.1689,
+      "step": 18053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6266579627990723,
+      "learning_rate": 3.503289473684211e-06,
+      "loss": 1.6753,
+      "step": 18054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0752651691436768,
+      "learning_rate": 3.5023757309941524e-06,
+      "loss": 2.1746,
+      "step": 18055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2121121883392334,
+      "learning_rate": 3.501461988304094e-06,
+      "loss": 1.3049,
+      "step": 18056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6270971298217773,
+      "learning_rate": 3.5005482456140355e-06,
+      "loss": 1.4591,
+      "step": 18057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7446072101593018,
+      "learning_rate": 3.499634502923977e-06,
+      "loss": 1.487,
+      "step": 18058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8067033290863037,
+      "learning_rate": 3.4987207602339183e-06,
+      "loss": 1.8071,
+      "step": 18059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6905596256256104,
+      "learning_rate": 3.49780701754386e-06,
+      "loss": 1.4494,
+      "step": 18060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6634597778320312,
+      "learning_rate": 3.4968932748538015e-06,
+      "loss": 1.5905,
+      "step": 18061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3080005645751953,
+      "learning_rate": 3.495979532163743e-06,
+      "loss": 1.2575,
+      "step": 18062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.583570718765259,
+      "learning_rate": 3.4950657894736846e-06,
+      "loss": 1.5822,
+      "step": 18063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.297964096069336,
+      "learning_rate": 3.494152046783626e-06,
+      "loss": 1.5538,
+      "step": 18064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5317280292510986,
+      "learning_rate": 3.4932383040935674e-06,
+      "loss": 1.4428,
+      "step": 18065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0073139667510986,
+      "learning_rate": 3.4923245614035088e-06,
+      "loss": 1.7409,
+      "step": 18066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1972475051879883,
+      "learning_rate": 3.4914108187134506e-06,
+      "loss": 1.2852,
+      "step": 18067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2729222774505615,
+      "learning_rate": 3.490497076023392e-06,
+      "loss": 1.2349,
+      "step": 18068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6699953079223633,
+      "learning_rate": 3.4895833333333333e-06,
+      "loss": 1.6546,
+      "step": 18069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.371962070465088,
+      "learning_rate": 3.488669590643275e-06,
+      "loss": 1.2977,
+      "step": 18070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0476129055023193,
+      "learning_rate": 3.4877558479532165e-06,
+      "loss": 1.649,
+      "step": 18071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.614403009414673,
+      "learning_rate": 3.486842105263158e-06,
+      "loss": 1.3294,
+      "step": 18072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.883833885192871,
+      "learning_rate": 3.4859283625731e-06,
+      "loss": 1.9572,
+      "step": 18073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.54665207862854,
+      "learning_rate": 3.4850146198830415e-06,
+      "loss": 1.3931,
+      "step": 18074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0326693058013916,
+      "learning_rate": 3.484100877192983e-06,
+      "loss": 1.07,
+      "step": 18075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3125741481781006,
+      "learning_rate": 3.4831871345029246e-06,
+      "loss": 1.6994,
+      "step": 18076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.597074270248413,
+      "learning_rate": 3.482273391812866e-06,
+      "loss": 1.7859,
+      "step": 18077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.47308611869812,
+      "learning_rate": 3.4813596491228074e-06,
+      "loss": 1.2887,
+      "step": 18078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7692177295684814,
+      "learning_rate": 3.4804459064327487e-06,
+      "loss": 1.159,
+      "step": 18079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2209465503692627,
+      "learning_rate": 3.4795321637426905e-06,
+      "loss": 1.2657,
+      "step": 18080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.677825450897217,
+      "learning_rate": 3.478618421052632e-06,
+      "loss": 1.4678,
+      "step": 18081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5485947132110596,
+      "learning_rate": 3.4777046783625733e-06,
+      "loss": 1.366,
+      "step": 18082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.267491102218628,
+      "learning_rate": 3.476790935672515e-06,
+      "loss": 1.4393,
+      "step": 18083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.041597366333008,
+      "learning_rate": 3.4758771929824565e-06,
+      "loss": 1.8108,
+      "step": 18084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4590394496917725,
+      "learning_rate": 3.474963450292398e-06,
+      "loss": 1.42,
+      "step": 18085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.365548610687256,
+      "learning_rate": 3.4740497076023392e-06,
+      "loss": 1.3288,
+      "step": 18086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.678534984588623,
+      "learning_rate": 3.473135964912281e-06,
+      "loss": 1.6777,
+      "step": 18087
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5344157218933105,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 1.5732,
+      "step": 18088
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6942481994628906,
+      "learning_rate": 3.4713084795321638e-06,
+      "loss": 1.3423,
+      "step": 18089
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7366738319396973,
+      "learning_rate": 3.4703947368421056e-06,
+      "loss": 1.7307,
+      "step": 18090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.973594903945923,
+      "learning_rate": 3.469480994152047e-06,
+      "loss": 1.7891,
+      "step": 18091
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3365933895111084,
+      "learning_rate": 3.4685672514619883e-06,
+      "loss": 1.2939,
+      "step": 18092
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6340062618255615,
+      "learning_rate": 3.4676535087719297e-06,
+      "loss": 1.4024,
+      "step": 18093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7333078384399414,
+      "learning_rate": 3.466739766081872e-06,
+      "loss": 1.306,
+      "step": 18094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.58392333984375,
+      "learning_rate": 3.4658260233918133e-06,
+      "loss": 1.5666,
+      "step": 18095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.79445481300354,
+      "learning_rate": 3.464912280701755e-06,
+      "loss": 1.5509,
+      "step": 18096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4196724891662598,
+      "learning_rate": 3.4639985380116965e-06,
+      "loss": 1.5045,
+      "step": 18097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8093693256378174,
+      "learning_rate": 3.463084795321638e-06,
+      "loss": 1.4845,
+      "step": 18098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2857704162597656,
+      "learning_rate": 3.462171052631579e-06,
+      "loss": 1.2113,
+      "step": 18099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1409718990325928,
+      "learning_rate": 3.461257309941521e-06,
+      "loss": 1.1207,
+      "step": 18100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.156175136566162,
+      "learning_rate": 3.4603435672514624e-06,
+      "loss": 1.1893,
+      "step": 18101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0753517150878906,
+      "learning_rate": 3.4594298245614037e-06,
+      "loss": 1.3273,
+      "step": 18102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5742475986480713,
+      "learning_rate": 3.4585160818713455e-06,
+      "loss": 1.6956,
+      "step": 18103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.742271900177002,
+      "learning_rate": 3.457602339181287e-06,
+      "loss": 1.7985,
+      "step": 18104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.08829402923584,
+      "learning_rate": 3.4566885964912283e-06,
+      "loss": 1.4469,
+      "step": 18105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.431291341781616,
+      "learning_rate": 3.4557748538011697e-06,
+      "loss": 1.1416,
+      "step": 18106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.518606662750244,
+      "learning_rate": 3.4548611111111115e-06,
+      "loss": 1.4336,
+      "step": 18107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.402022361755371,
+      "learning_rate": 3.453947368421053e-06,
+      "loss": 1.2347,
+      "step": 18108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4396209716796875,
+      "learning_rate": 3.4530336257309942e-06,
+      "loss": 1.2957,
+      "step": 18109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.68603515625,
+      "learning_rate": 3.452119883040936e-06,
+      "loss": 1.6077,
+      "step": 18110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5862908363342285,
+      "learning_rate": 3.4512061403508774e-06,
+      "loss": 1.4307,
+      "step": 18111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7805168628692627,
+      "learning_rate": 3.4502923976608188e-06,
+      "loss": 1.342,
+      "step": 18112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9669189453125,
+      "learning_rate": 3.44937865497076e-06,
+      "loss": 1.5741,
+      "step": 18113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.589725971221924,
+      "learning_rate": 3.448464912280702e-06,
+      "loss": 1.1535,
+      "step": 18114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9699840545654297,
+      "learning_rate": 3.4475511695906433e-06,
+      "loss": 1.1589,
+      "step": 18115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.751133680343628,
+      "learning_rate": 3.4466374269005855e-06,
+      "loss": 1.5453,
+      "step": 18116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8520772457122803,
+      "learning_rate": 3.445723684210527e-06,
+      "loss": 1.6223,
+      "step": 18117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.222443103790283,
+      "learning_rate": 3.4448099415204683e-06,
+      "loss": 1.3137,
+      "step": 18118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.674596071243286,
+      "learning_rate": 3.4438961988304097e-06,
+      "loss": 1.2825,
+      "step": 18119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8862297534942627,
+      "learning_rate": 3.4429824561403515e-06,
+      "loss": 1.8896,
+      "step": 18120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.262750148773193,
+      "learning_rate": 3.442068713450293e-06,
+      "loss": 1.2372,
+      "step": 18121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9442239999771118,
+      "learning_rate": 3.441154970760234e-06,
+      "loss": 1.0619,
+      "step": 18122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4901299476623535,
+      "learning_rate": 3.440241228070176e-06,
+      "loss": 1.5371,
+      "step": 18123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3322737216949463,
+      "learning_rate": 3.4393274853801174e-06,
+      "loss": 1.4105,
+      "step": 18124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6023154258728027,
+      "learning_rate": 3.4384137426900587e-06,
+      "loss": 1.4702,
+      "step": 18125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.370182752609253,
+      "learning_rate": 3.4375e-06,
+      "loss": 1.3784,
+      "step": 18126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0867128372192383,
+      "learning_rate": 3.436586257309942e-06,
+      "loss": 1.1214,
+      "step": 18127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2437326908111572,
+      "learning_rate": 3.4356725146198833e-06,
+      "loss": 1.2462,
+      "step": 18128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.217395782470703,
+      "learning_rate": 3.4347587719298247e-06,
+      "loss": 1.4167,
+      "step": 18129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3846487998962402,
+      "learning_rate": 3.4338450292397665e-06,
+      "loss": 1.4714,
+      "step": 18130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1026601791381836,
+      "learning_rate": 3.432931286549708e-06,
+      "loss": 1.3594,
+      "step": 18131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.923281669616699,
+      "learning_rate": 3.4320175438596492e-06,
+      "loss": 1.541,
+      "step": 18132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.544992208480835,
+      "learning_rate": 3.4311038011695906e-06,
+      "loss": 1.1449,
+      "step": 18133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.592073917388916,
+      "learning_rate": 3.4301900584795324e-06,
+      "loss": 1.6353,
+      "step": 18134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.298081159591675,
+      "learning_rate": 3.4292763157894738e-06,
+      "loss": 1.406,
+      "step": 18135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.462599992752075,
+      "learning_rate": 3.428362573099415e-06,
+      "loss": 1.3159,
+      "step": 18136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7987313270568848,
+      "learning_rate": 3.4274488304093574e-06,
+      "loss": 1.5311,
+      "step": 18137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.496042013168335,
+      "learning_rate": 3.4265350877192987e-06,
+      "loss": 1.5941,
+      "step": 18138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.386136770248413,
+      "learning_rate": 3.42562134502924e-06,
+      "loss": 1.3172,
+      "step": 18139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.358572006225586,
+      "learning_rate": 3.424707602339182e-06,
+      "loss": 1.6138,
+      "step": 18140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.037036657333374,
+      "learning_rate": 3.4237938596491233e-06,
+      "loss": 1.1727,
+      "step": 18141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3683927059173584,
+      "learning_rate": 3.4228801169590647e-06,
+      "loss": 1.6169,
+      "step": 18142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5856847763061523,
+      "learning_rate": 3.4219663742690065e-06,
+      "loss": 1.5392,
+      "step": 18143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.416808605194092,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 1.4727,
+      "step": 18144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6256394386291504,
+      "learning_rate": 3.420138888888889e-06,
+      "loss": 1.8203,
+      "step": 18145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.888120412826538,
+      "learning_rate": 3.4192251461988306e-06,
+      "loss": 1.4962,
+      "step": 18146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.658315658569336,
+      "learning_rate": 3.4183114035087724e-06,
+      "loss": 1.6778,
+      "step": 18147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.525705575942993,
+      "learning_rate": 3.4173976608187137e-06,
+      "loss": 1.5565,
+      "step": 18148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0956854820251465,
+      "learning_rate": 3.416483918128655e-06,
+      "loss": 1.5807,
+      "step": 18149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4089245796203613,
+      "learning_rate": 3.4155701754385965e-06,
+      "loss": 1.4623,
+      "step": 18150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.899510622024536,
+      "learning_rate": 3.4146564327485383e-06,
+      "loss": 1.5171,
+      "step": 18151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.098508596420288,
+      "learning_rate": 3.4137426900584797e-06,
+      "loss": 1.8167,
+      "step": 18152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2295796871185303,
+      "learning_rate": 3.412828947368421e-06,
+      "loss": 1.3018,
+      "step": 18153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.571399688720703,
+      "learning_rate": 3.411915204678363e-06,
+      "loss": 1.3567,
+      "step": 18154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5191543102264404,
+      "learning_rate": 3.4110014619883042e-06,
+      "loss": 1.393,
+      "step": 18155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.368344783782959,
+      "learning_rate": 3.4100877192982456e-06,
+      "loss": 1.4166,
+      "step": 18156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3973615169525146,
+      "learning_rate": 3.409173976608187e-06,
+      "loss": 1.4202,
+      "step": 18157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3573172092437744,
+      "learning_rate": 3.4082602339181288e-06,
+      "loss": 1.4311,
+      "step": 18158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.176262140274048,
+      "learning_rate": 3.4073464912280706e-06,
+      "loss": 1.4762,
+      "step": 18159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6983802318573,
+      "learning_rate": 3.4064327485380124e-06,
+      "loss": 1.5186,
+      "step": 18160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0600688457489014,
+      "learning_rate": 3.4055190058479537e-06,
+      "loss": 1.7874,
+      "step": 18161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.937227487564087,
+      "learning_rate": 3.404605263157895e-06,
+      "loss": 1.759,
+      "step": 18162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3792080879211426,
+      "learning_rate": 3.4036915204678365e-06,
+      "loss": 1.2914,
+      "step": 18163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.802086591720581,
+      "learning_rate": 3.4027777777777783e-06,
+      "loss": 1.6859,
+      "step": 18164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6130974292755127,
+      "learning_rate": 3.4018640350877197e-06,
+      "loss": 1.4917,
+      "step": 18165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.527130603790283,
+      "learning_rate": 3.400950292397661e-06,
+      "loss": 1.4154,
+      "step": 18166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5547232627868652,
+      "learning_rate": 3.400036549707603e-06,
+      "loss": 1.5219,
+      "step": 18167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.454540491104126,
+      "learning_rate": 3.399122807017544e-06,
+      "loss": 1.539,
+      "step": 18168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0393450260162354,
+      "learning_rate": 3.3982090643274856e-06,
+      "loss": 1.2594,
+      "step": 18169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1987481117248535,
+      "learning_rate": 3.397295321637427e-06,
+      "loss": 1.2096,
+      "step": 18170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4579715728759766,
+      "learning_rate": 3.3963815789473687e-06,
+      "loss": 1.3385,
+      "step": 18171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.123339891433716,
+      "learning_rate": 3.39546783625731e-06,
+      "loss": 1.067,
+      "step": 18172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4995012283325195,
+      "learning_rate": 3.3945540935672515e-06,
+      "loss": 1.3087,
+      "step": 18173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1024820804595947,
+      "learning_rate": 3.3936403508771933e-06,
+      "loss": 1.5212,
+      "step": 18174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4643523693084717,
+      "learning_rate": 3.3927266081871347e-06,
+      "loss": 1.3846,
+      "step": 18175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3812339305877686,
+      "learning_rate": 3.391812865497076e-06,
+      "loss": 1.4791,
+      "step": 18176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.118875026702881,
+      "learning_rate": 3.3908991228070174e-06,
+      "loss": 1.8784,
+      "step": 18177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.73099422454834,
+      "learning_rate": 3.3899853801169592e-06,
+      "loss": 1.5811,
+      "step": 18178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5252223014831543,
+      "learning_rate": 3.3890716374269006e-06,
+      "loss": 1.3733,
+      "step": 18179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.56828236579895,
+      "learning_rate": 3.388157894736843e-06,
+      "loss": 1.5769,
+      "step": 18180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5079739093780518,
+      "learning_rate": 3.387244152046784e-06,
+      "loss": 1.5424,
+      "step": 18181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.314441680908203,
+      "learning_rate": 3.3863304093567256e-06,
+      "loss": 1.2273,
+      "step": 18182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9319703578948975,
+      "learning_rate": 3.385416666666667e-06,
+      "loss": 1.7839,
+      "step": 18183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1009953022003174,
+      "learning_rate": 3.3845029239766087e-06,
+      "loss": 1.7607,
+      "step": 18184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.024087905883789,
+      "learning_rate": 3.38358918128655e-06,
+      "loss": 1.6033,
+      "step": 18185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.18501877784729,
+      "learning_rate": 3.3826754385964915e-06,
+      "loss": 1.3362,
+      "step": 18186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.655654191970825,
+      "learning_rate": 3.3817616959064333e-06,
+      "loss": 1.8736,
+      "step": 18187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7992632389068604,
+      "learning_rate": 3.3808479532163747e-06,
+      "loss": 1.4296,
+      "step": 18188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3592543601989746,
+      "learning_rate": 3.379934210526316e-06,
+      "loss": 1.8984,
+      "step": 18189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6149210929870605,
+      "learning_rate": 3.3790204678362574e-06,
+      "loss": 1.5869,
+      "step": 18190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.229980707168579,
+      "learning_rate": 3.378106725146199e-06,
+      "loss": 1.2444,
+      "step": 18191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7521262168884277,
+      "learning_rate": 3.3771929824561406e-06,
+      "loss": 1.5776,
+      "step": 18192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.269200325012207,
+      "learning_rate": 3.376279239766082e-06,
+      "loss": 1.4669,
+      "step": 18193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8739278316497803,
+      "learning_rate": 3.3753654970760237e-06,
+      "loss": 1.8945,
+      "step": 18194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7826433181762695,
+      "learning_rate": 3.374451754385965e-06,
+      "loss": 1.3884,
+      "step": 18195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7561254501342773,
+      "learning_rate": 3.3735380116959065e-06,
+      "loss": 1.6213,
+      "step": 18196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.649045944213867,
+      "learning_rate": 3.372624269005848e-06,
+      "loss": 1.124,
+      "step": 18197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7326815128326416,
+      "learning_rate": 3.3717105263157897e-06,
+      "loss": 1.6134,
+      "step": 18198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9408836364746094,
+      "learning_rate": 3.370796783625731e-06,
+      "loss": 1.3842,
+      "step": 18199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.494658946990967,
+      "learning_rate": 3.3698830409356724e-06,
+      "loss": 1.4656,
+      "step": 18200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.699317693710327,
+      "learning_rate": 3.3689692982456146e-06,
+      "loss": 1.3921,
+      "step": 18201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7291271686553955,
+      "learning_rate": 3.368055555555556e-06,
+      "loss": 1.2491,
+      "step": 18202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6122140884399414,
+      "learning_rate": 3.3671418128654974e-06,
+      "loss": 1.4112,
+      "step": 18203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7966320514678955,
+      "learning_rate": 3.366228070175439e-06,
+      "loss": 1.5303,
+      "step": 18204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5432066917419434,
+      "learning_rate": 3.3653143274853806e-06,
+      "loss": 1.3115,
+      "step": 18205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.928297758102417,
+      "learning_rate": 3.364400584795322e-06,
+      "loss": 1.7631,
+      "step": 18206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.313345432281494,
+      "learning_rate": 3.3634868421052637e-06,
+      "loss": 1.2903,
+      "step": 18207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.048206090927124,
+      "learning_rate": 3.362573099415205e-06,
+      "loss": 1.229,
+      "step": 18208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6522228717803955,
+      "learning_rate": 3.3616593567251465e-06,
+      "loss": 1.7027,
+      "step": 18209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.057100534439087,
+      "learning_rate": 3.360745614035088e-06,
+      "loss": 1.6983,
+      "step": 18210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7936153411865234,
+      "learning_rate": 3.3598318713450297e-06,
+      "loss": 1.8681,
+      "step": 18211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3369462490081787,
+      "learning_rate": 3.358918128654971e-06,
+      "loss": 1.7859,
+      "step": 18212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.100421905517578,
+      "learning_rate": 3.3580043859649124e-06,
+      "loss": 1.5055,
+      "step": 18213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0624454021453857,
+      "learning_rate": 3.357090643274854e-06,
+      "loss": 1.2189,
+      "step": 18214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3100879192352295,
+      "learning_rate": 3.3561769005847956e-06,
+      "loss": 1.0984,
+      "step": 18215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.432317018508911,
+      "learning_rate": 3.355263157894737e-06,
+      "loss": 1.3155,
+      "step": 18216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4312186241149902,
+      "learning_rate": 3.3543494152046783e-06,
+      "loss": 1.3043,
+      "step": 18217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.547093152999878,
+      "learning_rate": 3.35343567251462e-06,
+      "loss": 1.2003,
+      "step": 18218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5746538639068604,
+      "learning_rate": 3.3525219298245615e-06,
+      "loss": 1.5811,
+      "step": 18219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.513962984085083,
+      "learning_rate": 3.351608187134503e-06,
+      "loss": 1.5754,
+      "step": 18220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8790524005889893,
+      "learning_rate": 3.3506944444444447e-06,
+      "loss": 1.8104,
+      "step": 18221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6198642253875732,
+      "learning_rate": 3.349780701754386e-06,
+      "loss": 1.5384,
+      "step": 18222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.405829429626465,
+      "learning_rate": 3.348866959064328e-06,
+      "loss": 1.4902,
+      "step": 18223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0579311847686768,
+      "learning_rate": 3.3479532163742696e-06,
+      "loss": 1.436,
+      "step": 18224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.393110990524292,
+      "learning_rate": 3.347039473684211e-06,
+      "loss": 1.0432,
+      "step": 18225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4655165672302246,
+      "learning_rate": 3.3461257309941524e-06,
+      "loss": 1.6079,
+      "step": 18226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.767259120941162,
+      "learning_rate": 3.345211988304094e-06,
+      "loss": 1.5755,
+      "step": 18227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4124348163604736,
+      "learning_rate": 3.3442982456140356e-06,
+      "loss": 1.4321,
+      "step": 18228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8686952590942383,
+      "learning_rate": 3.343384502923977e-06,
+      "loss": 1.3828,
+      "step": 18229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7216882705688477,
+      "learning_rate": 3.3424707602339183e-06,
+      "loss": 1.7335,
+      "step": 18230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3062431812286377,
+      "learning_rate": 3.34155701754386e-06,
+      "loss": 1.2969,
+      "step": 18231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.787623167037964,
+      "learning_rate": 3.3406432748538015e-06,
+      "loss": 1.8223,
+      "step": 18232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.623361825942993,
+      "learning_rate": 3.339729532163743e-06,
+      "loss": 1.4317,
+      "step": 18233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.197566270828247,
+      "learning_rate": 3.3388157894736847e-06,
+      "loss": 1.1637,
+      "step": 18234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.698647975921631,
+      "learning_rate": 3.337902046783626e-06,
+      "loss": 1.4234,
+      "step": 18235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5764925479888916,
+      "learning_rate": 3.3369883040935674e-06,
+      "loss": 1.4913,
+      "step": 18236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.060835123062134,
+      "learning_rate": 3.3360745614035088e-06,
+      "loss": 1.6065,
+      "step": 18237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.449448585510254,
+      "learning_rate": 3.3351608187134506e-06,
+      "loss": 1.2579,
+      "step": 18238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.400207757949829,
+      "learning_rate": 3.334247076023392e-06,
+      "loss": 1.4618,
+      "step": 18239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.678635358810425,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.4786,
+      "step": 18240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.37190318107605,
+      "learning_rate": 3.332419590643275e-06,
+      "loss": 1.4968,
+      "step": 18241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.471010684967041,
+      "learning_rate": 3.3315058479532165e-06,
+      "loss": 1.3368,
+      "step": 18242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1743319034576416,
+      "learning_rate": 3.330592105263158e-06,
+      "loss": 1.1859,
+      "step": 18243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.742607593536377,
+      "learning_rate": 3.3296783625731e-06,
+      "loss": 1.4939,
+      "step": 18244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3316078186035156,
+      "learning_rate": 3.3287646198830415e-06,
+      "loss": 1.6257,
+      "step": 18245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1988067626953125,
+      "learning_rate": 3.327850877192983e-06,
+      "loss": 1.2975,
+      "step": 18246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.374208927154541,
+      "learning_rate": 3.3269371345029246e-06,
+      "loss": 1.5409,
+      "step": 18247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3092596530914307,
+      "learning_rate": 3.326023391812866e-06,
+      "loss": 1.1513,
+      "step": 18248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5573484897613525,
+      "learning_rate": 3.3251096491228074e-06,
+      "loss": 1.3258,
+      "step": 18249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.559309244155884,
+      "learning_rate": 3.3241959064327488e-06,
+      "loss": 1.2635,
+      "step": 18250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3387434482574463,
+      "learning_rate": 3.3232821637426906e-06,
+      "loss": 1.3022,
+      "step": 18251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.266896963119507,
+      "learning_rate": 3.322368421052632e-06,
+      "loss": 1.3305,
+      "step": 18252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.654442310333252,
+      "learning_rate": 3.3214546783625733e-06,
+      "loss": 1.4903,
+      "step": 18253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8307952880859375,
+      "learning_rate": 3.320540935672515e-06,
+      "loss": 1.6493,
+      "step": 18254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7165114879608154,
+      "learning_rate": 3.3196271929824565e-06,
+      "loss": 1.558,
+      "step": 18255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.99122953414917,
+      "learning_rate": 3.318713450292398e-06,
+      "loss": 1.9199,
+      "step": 18256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0829849243164062,
+      "learning_rate": 3.3177997076023392e-06,
+      "loss": 1.3908,
+      "step": 18257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8250536918640137,
+      "learning_rate": 3.316885964912281e-06,
+      "loss": 1.6199,
+      "step": 18258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0295701026916504,
+      "learning_rate": 3.3159722222222224e-06,
+      "loss": 2.0065,
+      "step": 18259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7897701263427734,
+      "learning_rate": 3.3150584795321638e-06,
+      "loss": 1.675,
+      "step": 18260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.114990234375,
+      "learning_rate": 3.314144736842105e-06,
+      "loss": 1.1087,
+      "step": 18261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.439253330230713,
+      "learning_rate": 3.313230994152047e-06,
+      "loss": 1.5564,
+      "step": 18262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9399268627166748,
+      "learning_rate": 3.3123172514619883e-06,
+      "loss": 0.9501,
+      "step": 18263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.076608419418335,
+      "learning_rate": 3.3114035087719297e-06,
+      "loss": 1.484,
+      "step": 18264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.257997751235962,
+      "learning_rate": 3.310489766081872e-06,
+      "loss": 1.4176,
+      "step": 18265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0593953132629395,
+      "learning_rate": 3.3095760233918133e-06,
+      "loss": 1.5994,
+      "step": 18266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3550055027008057,
+      "learning_rate": 3.308662280701755e-06,
+      "loss": 1.3697,
+      "step": 18267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.687959671020508,
+      "learning_rate": 3.3077485380116965e-06,
+      "loss": 1.5805,
+      "step": 18268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5062899589538574,
+      "learning_rate": 3.306834795321638e-06,
+      "loss": 1.4125,
+      "step": 18269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.166189432144165,
+      "learning_rate": 3.3059210526315792e-06,
+      "loss": 1.1803,
+      "step": 18270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.305612325668335,
+      "learning_rate": 3.305007309941521e-06,
+      "loss": 1.3139,
+      "step": 18271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.335853099822998,
+      "learning_rate": 3.3040935672514624e-06,
+      "loss": 1.1377,
+      "step": 18272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.619642972946167,
+      "learning_rate": 3.3031798245614038e-06,
+      "loss": 1.4677,
+      "step": 18273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.190866470336914,
+      "learning_rate": 3.302266081871345e-06,
+      "loss": 1.369,
+      "step": 18274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.120466470718384,
+      "learning_rate": 3.301352339181287e-06,
+      "loss": 1.8173,
+      "step": 18275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8588969707489014,
+      "learning_rate": 3.3004385964912283e-06,
+      "loss": 1.7822,
+      "step": 18276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.017352819442749,
+      "learning_rate": 3.2995248538011697e-06,
+      "loss": 1.6831,
+      "step": 18277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2887043952941895,
+      "learning_rate": 3.2986111111111115e-06,
+      "loss": 1.1167,
+      "step": 18278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4275765419006348,
+      "learning_rate": 3.297697368421053e-06,
+      "loss": 1.4608,
+      "step": 18279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.332714796066284,
+      "learning_rate": 3.2967836257309942e-06,
+      "loss": 1.3664,
+      "step": 18280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4758431911468506,
+      "learning_rate": 3.2958698830409356e-06,
+      "loss": 1.2744,
+      "step": 18281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.432974338531494,
+      "learning_rate": 3.2949561403508774e-06,
+      "loss": 1.3485,
+      "step": 18282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.623537540435791,
+      "learning_rate": 3.2940423976608188e-06,
+      "loss": 1.6894,
+      "step": 18283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7002549171447754,
+      "learning_rate": 3.29312865497076e-06,
+      "loss": 1.4062,
+      "step": 18284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5087032318115234,
+      "learning_rate": 3.292214912280702e-06,
+      "loss": 1.5314,
+      "step": 18285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.36757493019104,
+      "learning_rate": 3.2913011695906433e-06,
+      "loss": 1.5129,
+      "step": 18286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.310800075531006,
+      "learning_rate": 3.290387426900585e-06,
+      "loss": 1.0378,
+      "step": 18287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5911037921905518,
+      "learning_rate": 3.289473684210527e-06,
+      "loss": 1.4223,
+      "step": 18288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3637325763702393,
+      "learning_rate": 3.2885599415204683e-06,
+      "loss": 1.2182,
+      "step": 18289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.573086977005005,
+      "learning_rate": 3.2876461988304097e-06,
+      "loss": 1.6134,
+      "step": 18290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.490718126296997,
+      "learning_rate": 3.2867324561403515e-06,
+      "loss": 1.0972,
+      "step": 18291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8718643188476562,
+      "learning_rate": 3.285818713450293e-06,
+      "loss": 1.7671,
+      "step": 18292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7276289463043213,
+      "learning_rate": 3.2849049707602342e-06,
+      "loss": 1.3967,
+      "step": 18293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.024547576904297,
+      "learning_rate": 3.2839912280701756e-06,
+      "loss": 1.4149,
+      "step": 18294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3759078979492188,
+      "learning_rate": 3.2830774853801174e-06,
+      "loss": 1.4889,
+      "step": 18295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.332062005996704,
+      "learning_rate": 3.2821637426900588e-06,
+      "loss": 1.3862,
+      "step": 18296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4581799507141113,
+      "learning_rate": 3.28125e-06,
+      "loss": 1.6341,
+      "step": 18297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5732109546661377,
+      "learning_rate": 3.280336257309942e-06,
+      "loss": 1.1909,
+      "step": 18298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6089794635772705,
+      "learning_rate": 3.2794225146198833e-06,
+      "loss": 1.5168,
+      "step": 18299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2342798709869385,
+      "learning_rate": 3.2785087719298247e-06,
+      "loss": 1.1849,
+      "step": 18300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.272742509841919,
+      "learning_rate": 3.277595029239766e-06,
+      "loss": 1.3727,
+      "step": 18301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5194647312164307,
+      "learning_rate": 3.276681286549708e-06,
+      "loss": 1.4048,
+      "step": 18302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3416199684143066,
+      "learning_rate": 3.2757675438596492e-06,
+      "loss": 1.4019,
+      "step": 18303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.123992681503296,
+      "learning_rate": 3.2748538011695906e-06,
+      "loss": 1.5404,
+      "step": 18304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.253577709197998,
+      "learning_rate": 3.2739400584795324e-06,
+      "loss": 1.1666,
+      "step": 18305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7302491664886475,
+      "learning_rate": 3.2730263157894738e-06,
+      "loss": 1.7063,
+      "step": 18306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.886044979095459,
+      "learning_rate": 3.272112573099415e-06,
+      "loss": 1.5968,
+      "step": 18307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1132967472076416,
+      "learning_rate": 3.2711988304093574e-06,
+      "loss": 1.25,
+      "step": 18308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.397508144378662,
+      "learning_rate": 3.2702850877192987e-06,
+      "loss": 1.2537,
+      "step": 18309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.863757371902466,
+      "learning_rate": 3.26937134502924e-06,
+      "loss": 1.7595,
+      "step": 18310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6654462814331055,
+      "learning_rate": 3.268457602339182e-06,
+      "loss": 1.6802,
+      "step": 18311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.863661527633667,
+      "learning_rate": 3.2675438596491233e-06,
+      "loss": 2.0026,
+      "step": 18312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4374420642852783,
+      "learning_rate": 3.2666301169590647e-06,
+      "loss": 1.37,
+      "step": 18313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6339657306671143,
+      "learning_rate": 3.265716374269006e-06,
+      "loss": 1.8039,
+      "step": 18314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7771356105804443,
+      "learning_rate": 3.264802631578948e-06,
+      "loss": 1.2759,
+      "step": 18315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3823132514953613,
+      "learning_rate": 3.2638888888888892e-06,
+      "loss": 1.1999,
+      "step": 18316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.868321657180786,
+      "learning_rate": 3.2629751461988306e-06,
+      "loss": 1.6913,
+      "step": 18317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1849446296691895,
+      "learning_rate": 3.2620614035087724e-06,
+      "loss": 1.0071,
+      "step": 18318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0051445960998535,
+      "learning_rate": 3.2611476608187138e-06,
+      "loss": 1.9428,
+      "step": 18319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1463561058044434,
+      "learning_rate": 3.260233918128655e-06,
+      "loss": 1.4649,
+      "step": 18320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.430967092514038,
+      "learning_rate": 3.2593201754385965e-06,
+      "loss": 1.4324,
+      "step": 18321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7410099506378174,
+      "learning_rate": 3.2584064327485383e-06,
+      "loss": 1.7248,
+      "step": 18322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.26017689704895,
+      "learning_rate": 3.2574926900584797e-06,
+      "loss": 1.4463,
+      "step": 18323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9964964389801025,
+      "learning_rate": 3.256578947368421e-06,
+      "loss": 1.6321,
+      "step": 18324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.863543748855591,
+      "learning_rate": 3.255665204678363e-06,
+      "loss": 1.642,
+      "step": 18325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.512467384338379,
+      "learning_rate": 3.2547514619883042e-06,
+      "loss": 1.3489,
+      "step": 18326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4015462398529053,
+      "learning_rate": 3.2538377192982456e-06,
+      "loss": 1.2419,
+      "step": 18327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.282521963119507,
+      "learning_rate": 3.252923976608187e-06,
+      "loss": 1.2748,
+      "step": 18328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.226712942123413,
+      "learning_rate": 3.2520102339181288e-06,
+      "loss": 1.3202,
+      "step": 18329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4622228145599365,
+      "learning_rate": 3.2510964912280706e-06,
+      "loss": 1.3482,
+      "step": 18330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8850553035736084,
+      "learning_rate": 3.2501827485380124e-06,
+      "loss": 1.635,
+      "step": 18331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4559102058410645,
+      "learning_rate": 3.2492690058479537e-06,
+      "loss": 1.3985,
+      "step": 18332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.699960708618164,
+      "learning_rate": 3.248355263157895e-06,
+      "loss": 1.3164,
+      "step": 18333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.052884101867676,
+      "learning_rate": 3.2474415204678365e-06,
+      "loss": 1.3473,
+      "step": 18334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1897900104522705,
+      "learning_rate": 3.2465277777777783e-06,
+      "loss": 1.2461,
+      "step": 18335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1073644161224365,
+      "learning_rate": 3.2456140350877197e-06,
+      "loss": 1.6314,
+      "step": 18336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4494335651397705,
+      "learning_rate": 3.244700292397661e-06,
+      "loss": 1.3491,
+      "step": 18337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.25053596496582,
+      "learning_rate": 3.243786549707603e-06,
+      "loss": 1.4226,
+      "step": 18338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5666043758392334,
+      "learning_rate": 3.2428728070175442e-06,
+      "loss": 1.3257,
+      "step": 18339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6319339275360107,
+      "learning_rate": 3.2419590643274856e-06,
+      "loss": 1.3521,
+      "step": 18340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.634054183959961,
+      "learning_rate": 3.241045321637427e-06,
+      "loss": 1.5434,
+      "step": 18341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4540741443634033,
+      "learning_rate": 3.2401315789473688e-06,
+      "loss": 1.3641,
+      "step": 18342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4363837242126465,
+      "learning_rate": 3.23921783625731e-06,
+      "loss": 1.4277,
+      "step": 18343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4909017086029053,
+      "learning_rate": 3.2383040935672515e-06,
+      "loss": 1.3279,
+      "step": 18344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4847073554992676,
+      "learning_rate": 3.2373903508771933e-06,
+      "loss": 1.4391,
+      "step": 18345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.10807466506958,
+      "learning_rate": 3.2364766081871347e-06,
+      "loss": 1.2203,
+      "step": 18346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.775376558303833,
+      "learning_rate": 3.235562865497076e-06,
+      "loss": 1.7505,
+      "step": 18347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.655834913253784,
+      "learning_rate": 3.2346491228070174e-06,
+      "loss": 1.6112,
+      "step": 18348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2765989303588867,
+      "learning_rate": 3.2337353801169592e-06,
+      "loss": 1.3293,
+      "step": 18349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3387835025787354,
+      "learning_rate": 3.2328216374269006e-06,
+      "loss": 1.3868,
+      "step": 18350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.448444128036499,
+      "learning_rate": 3.231907894736843e-06,
+      "loss": 1.0349,
+      "step": 18351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3831655979156494,
+      "learning_rate": 3.230994152046784e-06,
+      "loss": 1.4693,
+      "step": 18352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.403263568878174,
+      "learning_rate": 3.2300804093567256e-06,
+      "loss": 1.4377,
+      "step": 18353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.860690116882324,
+      "learning_rate": 3.229166666666667e-06,
+      "loss": 1.9028,
+      "step": 18354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.410475254058838,
+      "learning_rate": 3.2282529239766087e-06,
+      "loss": 1.4252,
+      "step": 18355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8575360774993896,
+      "learning_rate": 3.22733918128655e-06,
+      "loss": 1.5392,
+      "step": 18356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6451313495635986,
+      "learning_rate": 3.2264254385964915e-06,
+      "loss": 1.4638,
+      "step": 18357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.004290580749512,
+      "learning_rate": 3.2255116959064333e-06,
+      "loss": 1.6514,
+      "step": 18358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6404969692230225,
+      "learning_rate": 3.2245979532163747e-06,
+      "loss": 1.5384,
+      "step": 18359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.449672222137451,
+      "learning_rate": 3.223684210526316e-06,
+      "loss": 1.1868,
+      "step": 18360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.193222761154175,
+      "learning_rate": 3.2227704678362574e-06,
+      "loss": 1.5111,
+      "step": 18361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7578470706939697,
+      "learning_rate": 3.2218567251461992e-06,
+      "loss": 1.5328,
+      "step": 18362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.388410806655884,
+      "learning_rate": 3.2209429824561406e-06,
+      "loss": 1.3431,
+      "step": 18363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.562081813812256,
+      "learning_rate": 3.220029239766082e-06,
+      "loss": 1.4647,
+      "step": 18364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7505528926849365,
+      "learning_rate": 3.2191154970760238e-06,
+      "loss": 1.2783,
+      "step": 18365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.582204580307007,
+      "learning_rate": 3.218201754385965e-06,
+      "loss": 1.4509,
+      "step": 18366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6108274459838867,
+      "learning_rate": 3.2172880116959065e-06,
+      "loss": 1.3388,
+      "step": 18367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.469604969024658,
+      "learning_rate": 3.216374269005848e-06,
+      "loss": 1.2294,
+      "step": 18368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.731064796447754,
+      "learning_rate": 3.2154605263157897e-06,
+      "loss": 2.0361,
+      "step": 18369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7260019779205322,
+      "learning_rate": 3.214546783625731e-06,
+      "loss": 1.7126,
+      "step": 18370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5362799167633057,
+      "learning_rate": 3.2136330409356724e-06,
+      "loss": 1.8755,
+      "step": 18371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.567321538925171,
+      "learning_rate": 3.2127192982456147e-06,
+      "loss": 1.4177,
+      "step": 18372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.358955144882202,
+      "learning_rate": 3.211805555555556e-06,
+      "loss": 1.6623,
+      "step": 18373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.529008150100708,
+      "learning_rate": 3.2108918128654974e-06,
+      "loss": 1.6457,
+      "step": 18374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.316432237625122,
+      "learning_rate": 3.209978070175439e-06,
+      "loss": 1.1828,
+      "step": 18375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9076449871063232,
+      "learning_rate": 3.2090643274853806e-06,
+      "loss": 0.9856,
+      "step": 18376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5341813564300537,
+      "learning_rate": 3.208150584795322e-06,
+      "loss": 1.5059,
+      "step": 18377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1846885681152344,
+      "learning_rate": 3.2072368421052637e-06,
+      "loss": 1.2457,
+      "step": 18378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.235093593597412,
+      "learning_rate": 3.206323099415205e-06,
+      "loss": 1.1858,
+      "step": 18379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.106005907058716,
+      "learning_rate": 3.2054093567251465e-06,
+      "loss": 1.0677,
+      "step": 18380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5849287509918213,
+      "learning_rate": 3.204495614035088e-06,
+      "loss": 1.4561,
+      "step": 18381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2819149494171143,
+      "learning_rate": 3.2035818713450297e-06,
+      "loss": 1.4582,
+      "step": 18382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4864795207977295,
+      "learning_rate": 3.202668128654971e-06,
+      "loss": 1.4093,
+      "step": 18383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8021976947784424,
+      "learning_rate": 3.2017543859649124e-06,
+      "loss": 1.6462,
+      "step": 18384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3387608528137207,
+      "learning_rate": 3.200840643274854e-06,
+      "loss": 2.0605,
+      "step": 18385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.439007520675659,
+      "learning_rate": 3.1999269005847956e-06,
+      "loss": 1.3306,
+      "step": 18386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1465346813201904,
+      "learning_rate": 3.199013157894737e-06,
+      "loss": 1.2635,
+      "step": 18387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.890625238418579,
+      "learning_rate": 3.1980994152046783e-06,
+      "loss": 1.5544,
+      "step": 18388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.604470729827881,
+      "learning_rate": 3.19718567251462e-06,
+      "loss": 1.5066,
+      "step": 18389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4651637077331543,
+      "learning_rate": 3.1962719298245615e-06,
+      "loss": 1.2385,
+      "step": 18390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0669846534729004,
+      "learning_rate": 3.195358187134503e-06,
+      "loss": 1.9178,
+      "step": 18391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4900567531585693,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 1.3631,
+      "step": 18392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8859777450561523,
+      "learning_rate": 3.193530701754386e-06,
+      "loss": 1.9702,
+      "step": 18393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1861488819122314,
+      "learning_rate": 3.192616959064328e-06,
+      "loss": 1.3159,
+      "step": 18394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.612917423248291,
+      "learning_rate": 3.1917032163742697e-06,
+      "loss": 1.3891,
+      "step": 18395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1481645107269287,
+      "learning_rate": 3.190789473684211e-06,
+      "loss": 1.2063,
+      "step": 18396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.076937198638916,
+      "learning_rate": 3.1898757309941524e-06,
+      "loss": 1.2887,
+      "step": 18397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1265273094177246,
+      "learning_rate": 3.1889619883040938e-06,
+      "loss": 1.0416,
+      "step": 18398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3490195274353027,
+      "learning_rate": 3.1880482456140356e-06,
+      "loss": 1.4541,
+      "step": 18399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4529101848602295,
+      "learning_rate": 3.187134502923977e-06,
+      "loss": 1.2448,
+      "step": 18400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8566086292266846,
+      "learning_rate": 3.1862207602339183e-06,
+      "loss": 1.6092,
+      "step": 18401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2943198680877686,
+      "learning_rate": 3.18530701754386e-06,
+      "loss": 1.1567,
+      "step": 18402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.455723285675049,
+      "learning_rate": 3.1843932748538015e-06,
+      "loss": 1.7083,
+      "step": 18403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6638450622558594,
+      "learning_rate": 3.183479532163743e-06,
+      "loss": 1.6197,
+      "step": 18404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9018949270248413,
+      "learning_rate": 3.1825657894736842e-06,
+      "loss": 1.0878,
+      "step": 18405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.159191370010376,
+      "learning_rate": 3.181652046783626e-06,
+      "loss": 1.4467,
+      "step": 18406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4827167987823486,
+      "learning_rate": 3.1807383040935674e-06,
+      "loss": 1.4395,
+      "step": 18407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.47652530670166,
+      "learning_rate": 3.179824561403509e-06,
+      "loss": 1.4904,
+      "step": 18408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.677638053894043,
+      "learning_rate": 3.1789108187134506e-06,
+      "loss": 1.8519,
+      "step": 18409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3164761066436768,
+      "learning_rate": 3.177997076023392e-06,
+      "loss": 1.2148,
+      "step": 18410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.147804021835327,
+      "learning_rate": 3.1770833333333333e-06,
+      "loss": 1.2916,
+      "step": 18411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4482405185699463,
+      "learning_rate": 3.1761695906432747e-06,
+      "loss": 1.3473,
+      "step": 18412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.640162467956543,
+      "learning_rate": 3.1752558479532165e-06,
+      "loss": 1.5279,
+      "step": 18413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3313803672790527,
+      "learning_rate": 3.174342105263158e-06,
+      "loss": 1.488,
+      "step": 18414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9356260299682617,
+      "learning_rate": 3.1734283625731e-06,
+      "loss": 0.97,
+      "step": 18415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0752336978912354,
+      "learning_rate": 3.1725146198830415e-06,
+      "loss": 1.2091,
+      "step": 18416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2619848251342773,
+      "learning_rate": 3.171600877192983e-06,
+      "loss": 1.4303,
+      "step": 18417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0125017166137695,
+      "learning_rate": 3.1706871345029242e-06,
+      "loss": 1.1683,
+      "step": 18418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.740928888320923,
+      "learning_rate": 3.169773391812866e-06,
+      "loss": 1.4651,
+      "step": 18419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.575800895690918,
+      "learning_rate": 3.1688596491228074e-06,
+      "loss": 1.4262,
+      "step": 18420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.590374231338501,
+      "learning_rate": 3.1679459064327488e-06,
+      "loss": 1.6095,
+      "step": 18421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.856086492538452,
+      "learning_rate": 3.1670321637426906e-06,
+      "loss": 1.704,
+      "step": 18422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3424293994903564,
+      "learning_rate": 3.166118421052632e-06,
+      "loss": 1.0074,
+      "step": 18423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.546382188796997,
+      "learning_rate": 3.1652046783625733e-06,
+      "loss": 1.3337,
+      "step": 18424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.397296667098999,
+      "learning_rate": 3.1642909356725147e-06,
+      "loss": 1.1774,
+      "step": 18425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.563333511352539,
+      "learning_rate": 3.1633771929824565e-06,
+      "loss": 1.6955,
+      "step": 18426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.196507215499878,
+      "learning_rate": 3.162463450292398e-06,
+      "loss": 1.1823,
+      "step": 18427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3280699253082275,
+      "learning_rate": 3.1615497076023392e-06,
+      "loss": 1.2376,
+      "step": 18428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.968697428703308,
+      "learning_rate": 3.160635964912281e-06,
+      "loss": 1.0711,
+      "step": 18429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.858379602432251,
+      "learning_rate": 3.1597222222222224e-06,
+      "loss": 1.7884,
+      "step": 18430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1836307048797607,
+      "learning_rate": 3.158808479532164e-06,
+      "loss": 1.4115,
+      "step": 18431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2703986167907715,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 1.4684,
+      "step": 18432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.182772397994995,
+      "learning_rate": 3.156980994152047e-06,
+      "loss": 1.2691,
+      "step": 18433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6524746417999268,
+      "learning_rate": 3.1560672514619883e-06,
+      "loss": 1.4078,
+      "step": 18434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1085267066955566,
+      "learning_rate": 3.1551535087719297e-06,
+      "loss": 1.0806,
+      "step": 18435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2592949867248535,
+      "learning_rate": 3.154239766081872e-06,
+      "loss": 1.4453,
+      "step": 18436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5497329235076904,
+      "learning_rate": 3.1533260233918133e-06,
+      "loss": 1.364,
+      "step": 18437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.754209518432617,
+      "learning_rate": 3.1524122807017547e-06,
+      "loss": 1.5854,
+      "step": 18438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3641374111175537,
+      "learning_rate": 3.1514985380116965e-06,
+      "loss": 1.416,
+      "step": 18439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.478865146636963,
+      "learning_rate": 3.150584795321638e-06,
+      "loss": 1.3556,
+      "step": 18440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.847041368484497,
+      "learning_rate": 3.1496710526315792e-06,
+      "loss": 1.4083,
+      "step": 18441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1258933544158936,
+      "learning_rate": 3.148757309941521e-06,
+      "loss": 1.0954,
+      "step": 18442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.256162166595459,
+      "learning_rate": 3.1478435672514624e-06,
+      "loss": 1.4485,
+      "step": 18443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9922716617584229,
+      "learning_rate": 3.1469298245614038e-06,
+      "loss": 1.0927,
+      "step": 18444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.906428575515747,
+      "learning_rate": 3.146016081871345e-06,
+      "loss": 1.5996,
+      "step": 18445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2691867351531982,
+      "learning_rate": 3.145102339181287e-06,
+      "loss": 1.4181,
+      "step": 18446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5755035877227783,
+      "learning_rate": 3.1441885964912283e-06,
+      "loss": 1.3856,
+      "step": 18447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3706705570220947,
+      "learning_rate": 3.1432748538011697e-06,
+      "loss": 1.3853,
+      "step": 18448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.257507085800171,
+      "learning_rate": 3.1423611111111115e-06,
+      "loss": 1.18,
+      "step": 18449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.319310426712036,
+      "learning_rate": 3.141447368421053e-06,
+      "loss": 1.4124,
+      "step": 18450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.899818181991577,
+      "learning_rate": 3.1405336257309942e-06,
+      "loss": 1.3431,
+      "step": 18451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2656726837158203,
+      "learning_rate": 3.1396198830409356e-06,
+      "loss": 1.3941,
+      "step": 18452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2530853748321533,
+      "learning_rate": 3.1387061403508774e-06,
+      "loss": 1.0655,
+      "step": 18453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7332935333251953,
+      "learning_rate": 3.137792397660819e-06,
+      "loss": 1.4717,
+      "step": 18454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9629874229431152,
+      "learning_rate": 3.13687865497076e-06,
+      "loss": 1.7337,
+      "step": 18455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5327115058898926,
+      "learning_rate": 3.135964912280702e-06,
+      "loss": 1.3618,
+      "step": 18456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.639871835708618,
+      "learning_rate": 3.1350511695906433e-06,
+      "loss": 1.6651,
+      "step": 18457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4462990760803223,
+      "learning_rate": 3.134137426900585e-06,
+      "loss": 1.2541,
+      "step": 18458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3628101348876953,
+      "learning_rate": 3.133223684210527e-06,
+      "loss": 1.3522,
+      "step": 18459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.016279935836792,
+      "learning_rate": 3.1323099415204683e-06,
+      "loss": 1.4027,
+      "step": 18460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.560450792312622,
+      "learning_rate": 3.1313961988304097e-06,
+      "loss": 1.4733,
+      "step": 18461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4598395824432373,
+      "learning_rate": 3.1304824561403515e-06,
+      "loss": 1.7307,
+      "step": 18462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2689759731292725,
+      "learning_rate": 3.129568713450293e-06,
+      "loss": 1.5026,
+      "step": 18463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.266172170639038,
+      "learning_rate": 3.1286549707602342e-06,
+      "loss": 1.4418,
+      "step": 18464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8930500745773315,
+      "learning_rate": 3.1277412280701756e-06,
+      "loss": 0.9308,
+      "step": 18465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.594153881072998,
+      "learning_rate": 3.1268274853801174e-06,
+      "loss": 1.609,
+      "step": 18466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4865009784698486,
+      "learning_rate": 3.1259137426900588e-06,
+      "loss": 1.4241,
+      "step": 18467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6401877403259277,
+      "learning_rate": 3.125e-06,
+      "loss": 1.2663,
+      "step": 18468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1299657821655273,
+      "learning_rate": 3.124086257309942e-06,
+      "loss": 1.2307,
+      "step": 18469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.198387861251831,
+      "learning_rate": 3.1231725146198833e-06,
+      "loss": 1.2902,
+      "step": 18470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1108357906341553,
+      "learning_rate": 3.1222587719298247e-06,
+      "loss": 1.2632,
+      "step": 18471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5497210025787354,
+      "learning_rate": 3.121345029239766e-06,
+      "loss": 1.4599,
+      "step": 18472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.437026262283325,
+      "learning_rate": 3.120431286549708e-06,
+      "loss": 1.0797,
+      "step": 18473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.013807535171509,
+      "learning_rate": 3.1195175438596492e-06,
+      "loss": 1.0744,
+      "step": 18474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3161206245422363,
+      "learning_rate": 3.1186038011695906e-06,
+      "loss": 1.0984,
+      "step": 18475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6266112327575684,
+      "learning_rate": 3.1176900584795324e-06,
+      "loss": 1.5822,
+      "step": 18476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.663022518157959,
+      "learning_rate": 3.116776315789474e-06,
+      "loss": 1.2236,
+      "step": 18477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.133545160293579,
+      "learning_rate": 3.115862573099415e-06,
+      "loss": 1.8959,
+      "step": 18478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3203654289245605,
+      "learning_rate": 3.1149488304093574e-06,
+      "loss": 1.1257,
+      "step": 18479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.177389144897461,
+      "learning_rate": 3.1140350877192988e-06,
+      "loss": 1.7936,
+      "step": 18480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6833603382110596,
+      "learning_rate": 3.11312134502924e-06,
+      "loss": 1.2805,
+      "step": 18481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.405595541000366,
+      "learning_rate": 3.112207602339182e-06,
+      "loss": 1.5073,
+      "step": 18482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6748716831207275,
+      "learning_rate": 3.1112938596491233e-06,
+      "loss": 1.4563,
+      "step": 18483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.048421621322632,
+      "learning_rate": 3.1103801169590647e-06,
+      "loss": 1.4339,
+      "step": 18484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3149266242980957,
+      "learning_rate": 3.109466374269006e-06,
+      "loss": 1.3899,
+      "step": 18485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2370781898498535,
+      "learning_rate": 3.108552631578948e-06,
+      "loss": 1.4133,
+      "step": 18486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4290411472320557,
+      "learning_rate": 3.1076388888888892e-06,
+      "loss": 1.559,
+      "step": 18487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6331398487091064,
+      "learning_rate": 3.1067251461988306e-06,
+      "loss": 1.4801,
+      "step": 18488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.279930591583252,
+      "learning_rate": 3.1058114035087724e-06,
+      "loss": 1.4757,
+      "step": 18489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2664520740509033,
+      "learning_rate": 3.1048976608187138e-06,
+      "loss": 1.2684,
+      "step": 18490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.555149793624878,
+      "learning_rate": 3.103983918128655e-06,
+      "loss": 1.4116,
+      "step": 18491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.026404619216919,
+      "learning_rate": 3.1030701754385965e-06,
+      "loss": 1.604,
+      "step": 18492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2323806285858154,
+      "learning_rate": 3.1021564327485383e-06,
+      "loss": 1.3035,
+      "step": 18493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.400519847869873,
+      "learning_rate": 3.1012426900584797e-06,
+      "loss": 1.5364,
+      "step": 18494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.998896598815918,
+      "learning_rate": 3.100328947368421e-06,
+      "loss": 2.0568,
+      "step": 18495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5842645168304443,
+      "learning_rate": 3.0994152046783624e-06,
+      "loss": 1.4798,
+      "step": 18496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.179710626602173,
+      "learning_rate": 3.0985014619883042e-06,
+      "loss": 1.529,
+      "step": 18497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1979129314422607,
+      "learning_rate": 3.0975877192982456e-06,
+      "loss": 1.7844,
+      "step": 18498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.54905104637146,
+      "learning_rate": 3.096673976608187e-06,
+      "loss": 1.4059,
+      "step": 18499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3512237071990967,
+      "learning_rate": 3.095760233918129e-06,
+      "loss": 1.2991,
+      "step": 18500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.333975315093994,
+      "learning_rate": 3.0948464912280706e-06,
+      "loss": 1.3141,
+      "step": 18501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.510385274887085,
+      "learning_rate": 3.0939327485380124e-06,
+      "loss": 1.3365,
+      "step": 18502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5699477195739746,
+      "learning_rate": 3.0930190058479538e-06,
+      "loss": 0.9892,
+      "step": 18503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.406447649002075,
+      "learning_rate": 3.092105263157895e-06,
+      "loss": 1.4561,
+      "step": 18504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.389648199081421,
+      "learning_rate": 3.0911915204678365e-06,
+      "loss": 1.3483,
+      "step": 18505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3372552394866943,
+      "learning_rate": 3.0902777777777783e-06,
+      "loss": 1.3844,
+      "step": 18506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.673835515975952,
+      "learning_rate": 3.0893640350877197e-06,
+      "loss": 1.8925,
+      "step": 18507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.511143684387207,
+      "learning_rate": 3.088450292397661e-06,
+      "loss": 1.4355,
+      "step": 18508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.503661632537842,
+      "learning_rate": 3.0875365497076024e-06,
+      "loss": 1.2131,
+      "step": 18509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.639345169067383,
+      "learning_rate": 3.0866228070175442e-06,
+      "loss": 1.6503,
+      "step": 18510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.476787805557251,
+      "learning_rate": 3.0857090643274856e-06,
+      "loss": 1.3725,
+      "step": 18511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.452709674835205,
+      "learning_rate": 3.084795321637427e-06,
+      "loss": 1.481,
+      "step": 18512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.806690216064453,
+      "learning_rate": 3.0838815789473688e-06,
+      "loss": 1.4006,
+      "step": 18513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5326294898986816,
+      "learning_rate": 3.08296783625731e-06,
+      "loss": 1.1118,
+      "step": 18514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7734909057617188,
+      "learning_rate": 3.0820540935672515e-06,
+      "loss": 1.4981,
+      "step": 18515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.556816816329956,
+      "learning_rate": 3.081140350877193e-06,
+      "loss": 1.677,
+      "step": 18516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3909432888031006,
+      "learning_rate": 3.0802266081871347e-06,
+      "loss": 1.3157,
+      "step": 18517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.814056634902954,
+      "learning_rate": 3.079312865497076e-06,
+      "loss": 1.7083,
+      "step": 18518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0762879848480225,
+      "learning_rate": 3.0783991228070174e-06,
+      "loss": 1.4293,
+      "step": 18519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9937005043029785,
+      "learning_rate": 3.0774853801169592e-06,
+      "loss": 1.5842,
+      "step": 18520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3150458335876465,
+      "learning_rate": 3.0765716374269006e-06,
+      "loss": 1.3615,
+      "step": 18521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5844976902008057,
+      "learning_rate": 3.0756578947368424e-06,
+      "loss": 1.8592,
+      "step": 18522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7432456016540527,
+      "learning_rate": 3.0747441520467842e-06,
+      "loss": 1.5527,
+      "step": 18523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4727280139923096,
+      "learning_rate": 3.0738304093567256e-06,
+      "loss": 1.3692,
+      "step": 18524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.573467493057251,
+      "learning_rate": 3.072916666666667e-06,
+      "loss": 1.1859,
+      "step": 18525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2068138122558594,
+      "learning_rate": 3.0720029239766088e-06,
+      "loss": 1.691,
+      "step": 18526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9870481491088867,
+      "learning_rate": 3.07108918128655e-06,
+      "loss": 1.2041,
+      "step": 18527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.077288866043091,
+      "learning_rate": 3.0701754385964915e-06,
+      "loss": 1.9487,
+      "step": 18528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5403006076812744,
+      "learning_rate": 3.069261695906433e-06,
+      "loss": 1.4435,
+      "step": 18529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.653506278991699,
+      "learning_rate": 3.0683479532163747e-06,
+      "loss": 1.7282,
+      "step": 18530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9506211280822754,
+      "learning_rate": 3.067434210526316e-06,
+      "loss": 1.1568,
+      "step": 18531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.452564001083374,
+      "learning_rate": 3.0665204678362574e-06,
+      "loss": 1.4635,
+      "step": 18532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.43363618850708,
+      "learning_rate": 3.0656067251461992e-06,
+      "loss": 1.2356,
+      "step": 18533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.006383180618286,
+      "learning_rate": 3.0646929824561406e-06,
+      "loss": 1.593,
+      "step": 18534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2467589378356934,
+      "learning_rate": 3.063779239766082e-06,
+      "loss": 1.1548,
+      "step": 18535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8078408241271973,
+      "learning_rate": 3.0628654970760234e-06,
+      "loss": 1.6827,
+      "step": 18536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.458611011505127,
+      "learning_rate": 3.061951754385965e-06,
+      "loss": 1.3403,
+      "step": 18537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2697315216064453,
+      "learning_rate": 3.0610380116959065e-06,
+      "loss": 1.1475,
+      "step": 18538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.291398525238037,
+      "learning_rate": 3.060124269005848e-06,
+      "loss": 1.2776,
+      "step": 18539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0286824703216553,
+      "learning_rate": 3.0592105263157897e-06,
+      "loss": 1.3398,
+      "step": 18540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4508349895477295,
+      "learning_rate": 3.058296783625731e-06,
+      "loss": 1.6729,
+      "step": 18541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3675732612609863,
+      "learning_rate": 3.0573830409356724e-06,
+      "loss": 1.3673,
+      "step": 18542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.15993070602417,
+      "learning_rate": 3.0564692982456147e-06,
+      "loss": 1.1866,
+      "step": 18543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4187192916870117,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 1.2099,
+      "step": 18544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9973812103271484,
+      "learning_rate": 3.0546418128654974e-06,
+      "loss": 1.9995,
+      "step": 18545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2001214027404785,
+      "learning_rate": 3.0537280701754392e-06,
+      "loss": 1.2143,
+      "step": 18546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.598532199859619,
+      "learning_rate": 3.0528143274853806e-06,
+      "loss": 1.6987,
+      "step": 18547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9845902919769287,
+      "learning_rate": 3.051900584795322e-06,
+      "loss": 0.9111,
+      "step": 18548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6170694828033447,
+      "learning_rate": 3.0509868421052633e-06,
+      "loss": 1.4781,
+      "step": 18549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.566835403442383,
+      "learning_rate": 3.050073099415205e-06,
+      "loss": 1.5238,
+      "step": 18550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4571781158447266,
+      "learning_rate": 3.0491593567251465e-06,
+      "loss": 1.3795,
+      "step": 18551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2232534885406494,
+      "learning_rate": 3.048245614035088e-06,
+      "loss": 1.2327,
+      "step": 18552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.334662675857544,
+      "learning_rate": 3.0473318713450297e-06,
+      "loss": 1.4323,
+      "step": 18553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6815953254699707,
+      "learning_rate": 3.046418128654971e-06,
+      "loss": 1.5343,
+      "step": 18554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.604395627975464,
+      "learning_rate": 3.0455043859649124e-06,
+      "loss": 1.1713,
+      "step": 18555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7062456607818604,
+      "learning_rate": 3.044590643274854e-06,
+      "loss": 1.4579,
+      "step": 18556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5338845252990723,
+      "learning_rate": 3.0436769005847956e-06,
+      "loss": 1.1944,
+      "step": 18557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2439322471618652,
+      "learning_rate": 3.042763157894737e-06,
+      "loss": 1.3895,
+      "step": 18558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.689716100692749,
+      "learning_rate": 3.0418494152046784e-06,
+      "loss": 1.4824,
+      "step": 18559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3073508739471436,
+      "learning_rate": 3.04093567251462e-06,
+      "loss": 1.347,
+      "step": 18560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4052162170410156,
+      "learning_rate": 3.0400219298245615e-06,
+      "loss": 1.3105,
+      "step": 18561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4211509227752686,
+      "learning_rate": 3.039108187134503e-06,
+      "loss": 1.2506,
+      "step": 18562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.641498327255249,
+      "learning_rate": 3.0381944444444443e-06,
+      "loss": 1.5605,
+      "step": 18563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4850528240203857,
+      "learning_rate": 3.037280701754386e-06,
+      "loss": 1.4661,
+      "step": 18564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8481647968292236,
+      "learning_rate": 3.036366959064328e-06,
+      "loss": 1.7778,
+      "step": 18565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.441007375717163,
+      "learning_rate": 3.0354532163742697e-06,
+      "loss": 1.4561,
+      "step": 18566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7344186305999756,
+      "learning_rate": 3.034539473684211e-06,
+      "loss": 1.8657,
+      "step": 18567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.40144681930542,
+      "learning_rate": 3.0336257309941524e-06,
+      "loss": 1.3352,
+      "step": 18568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4839072227478027,
+      "learning_rate": 3.032711988304094e-06,
+      "loss": 1.7224,
+      "step": 18569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4596941471099854,
+      "learning_rate": 3.0317982456140356e-06,
+      "loss": 1.4814,
+      "step": 18570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3702609539031982,
+      "learning_rate": 3.030884502923977e-06,
+      "loss": 1.2426,
+      "step": 18571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4049572944641113,
+      "learning_rate": 3.0299707602339183e-06,
+      "loss": 1.5897,
+      "step": 18572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0471291542053223,
+      "learning_rate": 3.02905701754386e-06,
+      "loss": 1.2121,
+      "step": 18573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.608175754547119,
+      "learning_rate": 3.0281432748538015e-06,
+      "loss": 1.6035,
+      "step": 18574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.666760206222534,
+      "learning_rate": 3.027229532163743e-06,
+      "loss": 1.2989,
+      "step": 18575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0904433727264404,
+      "learning_rate": 3.0263157894736843e-06,
+      "loss": 1.3925,
+      "step": 18576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.948400616645813,
+      "learning_rate": 3.025402046783626e-06,
+      "loss": 1.3763,
+      "step": 18577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5507681369781494,
+      "learning_rate": 3.0244883040935674e-06,
+      "loss": 1.3218,
+      "step": 18578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.513828992843628,
+      "learning_rate": 3.023574561403509e-06,
+      "loss": 1.4427,
+      "step": 18579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2552342414855957,
+      "learning_rate": 3.0226608187134506e-06,
+      "loss": 1.267,
+      "step": 18580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1853926181793213,
+      "learning_rate": 3.021747076023392e-06,
+      "loss": 1.3925,
+      "step": 18581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0893328189849854,
+      "learning_rate": 3.0208333333333334e-06,
+      "loss": 1.3616,
+      "step": 18582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5621798038482666,
+      "learning_rate": 3.0199195906432747e-06,
+      "loss": 1.5423,
+      "step": 18583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.392209529876709,
+      "learning_rate": 3.0190058479532165e-06,
+      "loss": 1.4775,
+      "step": 18584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.469233751296997,
+      "learning_rate": 3.018092105263158e-06,
+      "loss": 1.6852,
+      "step": 18585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.459696054458618,
+      "learning_rate": 3.0171783625731e-06,
+      "loss": 1.5861,
+      "step": 18586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3251047134399414,
+      "learning_rate": 3.0162646198830415e-06,
+      "loss": 1.2091,
+      "step": 18587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3792331218719482,
+      "learning_rate": 3.015350877192983e-06,
+      "loss": 1.5409,
+      "step": 18588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0456838607788086,
+      "learning_rate": 3.0144371345029242e-06,
+      "loss": 1.394,
+      "step": 18589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6643049716949463,
+      "learning_rate": 3.013523391812866e-06,
+      "loss": 1.2468,
+      "step": 18590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.055844783782959,
+      "learning_rate": 3.0126096491228074e-06,
+      "loss": 1.4596,
+      "step": 18591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1854283809661865,
+      "learning_rate": 3.011695906432749e-06,
+      "loss": 1.0963,
+      "step": 18592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.151278018951416,
+      "learning_rate": 3.0107821637426906e-06,
+      "loss": 1.3688,
+      "step": 18593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7410213947296143,
+      "learning_rate": 3.009868421052632e-06,
+      "loss": 1.8364,
+      "step": 18594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.540095329284668,
+      "learning_rate": 3.0089546783625733e-06,
+      "loss": 1.4531,
+      "step": 18595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.272120952606201,
+      "learning_rate": 3.0080409356725147e-06,
+      "loss": 1.3961,
+      "step": 18596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3792102336883545,
+      "learning_rate": 3.0071271929824565e-06,
+      "loss": 1.3838,
+      "step": 18597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5801265239715576,
+      "learning_rate": 3.006213450292398e-06,
+      "loss": 1.3625,
+      "step": 18598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2912466526031494,
+      "learning_rate": 3.0052997076023393e-06,
+      "loss": 1.476,
+      "step": 18599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4685542583465576,
+      "learning_rate": 3.004385964912281e-06,
+      "loss": 1.4688,
+      "step": 18600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.233280897140503,
+      "learning_rate": 3.0034722222222224e-06,
+      "loss": 1.4898,
+      "step": 18601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0290277004241943,
+      "learning_rate": 3.002558479532164e-06,
+      "loss": 1.1984,
+      "step": 18602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4701592922210693,
+      "learning_rate": 3.001644736842105e-06,
+      "loss": 1.2828,
+      "step": 18603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.108152151107788,
+      "learning_rate": 3.000730994152047e-06,
+      "loss": 1.3058,
+      "step": 18604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6203763484954834,
+      "learning_rate": 2.9998172514619884e-06,
+      "loss": 1.407,
+      "step": 18605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2332680225372314,
+      "learning_rate": 2.9989035087719297e-06,
+      "loss": 1.187,
+      "step": 18606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4245824813842773,
+      "learning_rate": 2.997989766081872e-06,
+      "loss": 1.348,
+      "step": 18607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.449866771697998,
+      "learning_rate": 2.9970760233918133e-06,
+      "loss": 1.6592,
+      "step": 18608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.946009635925293,
+      "learning_rate": 2.9961622807017547e-06,
+      "loss": 1.3036,
+      "step": 18609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7890818119049072,
+      "learning_rate": 2.9952485380116965e-06,
+      "loss": 1.4725,
+      "step": 18610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.050095558166504,
+      "learning_rate": 2.994334795321638e-06,
+      "loss": 1.5095,
+      "step": 18611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.495776891708374,
+      "learning_rate": 2.9934210526315792e-06,
+      "loss": 1.4104,
+      "step": 18612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1475958824157715,
+      "learning_rate": 2.992507309941521e-06,
+      "loss": 1.1939,
+      "step": 18613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3410816192626953,
+      "learning_rate": 2.9915935672514624e-06,
+      "loss": 1.3464,
+      "step": 18614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2304139137268066,
+      "learning_rate": 2.990679824561404e-06,
+      "loss": 1.3587,
+      "step": 18615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1379776000976562,
+      "learning_rate": 2.989766081871345e-06,
+      "loss": 2.1954,
+      "step": 18616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5387964248657227,
+      "learning_rate": 2.988852339181287e-06,
+      "loss": 1.6593,
+      "step": 18617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8104164600372314,
+      "learning_rate": 2.9879385964912283e-06,
+      "loss": 1.5576,
+      "step": 18618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8662524223327637,
+      "learning_rate": 2.9870248538011697e-06,
+      "loss": 1.5199,
+      "step": 18619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4608352184295654,
+      "learning_rate": 2.986111111111111e-06,
+      "loss": 1.6124,
+      "step": 18620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.180763006210327,
+      "learning_rate": 2.985197368421053e-06,
+      "loss": 1.2737,
+      "step": 18621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7573788166046143,
+      "learning_rate": 2.9842836257309943e-06,
+      "loss": 1.5129,
+      "step": 18622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.671222686767578,
+      "learning_rate": 2.9833698830409356e-06,
+      "loss": 1.5263,
+      "step": 18623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.106147289276123,
+      "learning_rate": 2.9824561403508774e-06,
+      "loss": 1.2851,
+      "step": 18624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.504365921020508,
+      "learning_rate": 2.981542397660819e-06,
+      "loss": 1.4439,
+      "step": 18625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.446300745010376,
+      "learning_rate": 2.98062865497076e-06,
+      "loss": 1.425,
+      "step": 18626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1948583126068115,
+      "learning_rate": 2.9797149122807016e-06,
+      "loss": 1.1374,
+      "step": 18627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.38800048828125,
+      "learning_rate": 2.9788011695906434e-06,
+      "loss": 1.5891,
+      "step": 18628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.63685417175293,
+      "learning_rate": 2.977887426900585e-06,
+      "loss": 1.3799,
+      "step": 18629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4551315307617188,
+      "learning_rate": 2.976973684210527e-06,
+      "loss": 1.4565,
+      "step": 18630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8209195137023926,
+      "learning_rate": 2.9760599415204683e-06,
+      "loss": 1.6114,
+      "step": 18631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.169246196746826,
+      "learning_rate": 2.9751461988304097e-06,
+      "loss": 1.3593,
+      "step": 18632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8626956939697266,
+      "learning_rate": 2.974232456140351e-06,
+      "loss": 1.8561,
+      "step": 18633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6927223205566406,
+      "learning_rate": 2.973318713450293e-06,
+      "loss": 1.467,
+      "step": 18634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7042946815490723,
+      "learning_rate": 2.9724049707602342e-06,
+      "loss": 1.3726,
+      "step": 18635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2807400226593018,
+      "learning_rate": 2.9714912280701756e-06,
+      "loss": 1.5705,
+      "step": 18636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.15328049659729,
+      "learning_rate": 2.9705774853801174e-06,
+      "loss": 1.2418,
+      "step": 18637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.188666582107544,
+      "learning_rate": 2.969663742690059e-06,
+      "loss": 1.2235,
+      "step": 18638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.798128366470337,
+      "learning_rate": 2.96875e-06,
+      "loss": 1.6901,
+      "step": 18639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.80051326751709,
+      "learning_rate": 2.9678362573099415e-06,
+      "loss": 1.7559,
+      "step": 18640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5795586109161377,
+      "learning_rate": 2.9669225146198833e-06,
+      "loss": 1.3471,
+      "step": 18641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6181252002716064,
+      "learning_rate": 2.9660087719298247e-06,
+      "loss": 1.4574,
+      "step": 18642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8910129070281982,
+      "learning_rate": 2.965095029239766e-06,
+      "loss": 1.5376,
+      "step": 18643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6674399375915527,
+      "learning_rate": 2.964181286549708e-06,
+      "loss": 1.4555,
+      "step": 18644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3380520343780518,
+      "learning_rate": 2.9632675438596493e-06,
+      "loss": 1.3472,
+      "step": 18645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1321468353271484,
+      "learning_rate": 2.9623538011695906e-06,
+      "loss": 1.2664,
+      "step": 18646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.145503044128418,
+      "learning_rate": 2.961440058479532e-06,
+      "loss": 1.2601,
+      "step": 18647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.266075849533081,
+      "learning_rate": 2.960526315789474e-06,
+      "loss": 1.3508,
+      "step": 18648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.648301839828491,
+      "learning_rate": 2.959612573099415e-06,
+      "loss": 1.7208,
+      "step": 18649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.605602979660034,
+      "learning_rate": 2.9586988304093574e-06,
+      "loss": 1.4376,
+      "step": 18650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.653444528579712,
+      "learning_rate": 2.9577850877192988e-06,
+      "loss": 1.5635,
+      "step": 18651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.985604166984558,
+      "learning_rate": 2.95687134502924e-06,
+      "loss": 1.3805,
+      "step": 18652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8425631523132324,
+      "learning_rate": 2.9559576023391815e-06,
+      "loss": 1.6301,
+      "step": 18653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8021347522735596,
+      "learning_rate": 2.9550438596491233e-06,
+      "loss": 1.1815,
+      "step": 18654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.332568645477295,
+      "learning_rate": 2.9541301169590647e-06,
+      "loss": 1.16,
+      "step": 18655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8592214584350586,
+      "learning_rate": 2.953216374269006e-06,
+      "loss": 1.917,
+      "step": 18656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.387876033782959,
+      "learning_rate": 2.952302631578948e-06,
+      "loss": 1.4517,
+      "step": 18657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6512696743011475,
+      "learning_rate": 2.9513888888888892e-06,
+      "loss": 1.54,
+      "step": 18658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.565476894378662,
+      "learning_rate": 2.9504751461988306e-06,
+      "loss": 1.5092,
+      "step": 18659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.809753656387329,
+      "learning_rate": 2.949561403508772e-06,
+      "loss": 1.262,
+      "step": 18660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1779801845550537,
+      "learning_rate": 2.948647660818714e-06,
+      "loss": 1.3081,
+      "step": 18661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9768288135528564,
+      "learning_rate": 2.947733918128655e-06,
+      "loss": 1.8639,
+      "step": 18662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.571169137954712,
+      "learning_rate": 2.9468201754385965e-06,
+      "loss": 1.3105,
+      "step": 18663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1262712478637695,
+      "learning_rate": 2.9459064327485383e-06,
+      "loss": 1.1454,
+      "step": 18664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.119537353515625,
+      "learning_rate": 2.9449926900584797e-06,
+      "loss": 1.4463,
+      "step": 18665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6210968494415283,
+      "learning_rate": 2.944078947368421e-06,
+      "loss": 1.5717,
+      "step": 18666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3739407062530518,
+      "learning_rate": 2.9431652046783625e-06,
+      "loss": 1.3972,
+      "step": 18667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8221490383148193,
+      "learning_rate": 2.9422514619883043e-06,
+      "loss": 1.6531,
+      "step": 18668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2408745288848877,
+      "learning_rate": 2.9413377192982456e-06,
+      "loss": 1.1793,
+      "step": 18669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.171447515487671,
+      "learning_rate": 2.940423976608187e-06,
+      "loss": 1.1698,
+      "step": 18670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2742297649383545,
+      "learning_rate": 2.939510233918129e-06,
+      "loss": 1.2026,
+      "step": 18671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4330568313598633,
+      "learning_rate": 2.9385964912280706e-06,
+      "loss": 1.4785,
+      "step": 18672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.039947509765625,
+      "learning_rate": 2.937682748538012e-06,
+      "loss": 1.3012,
+      "step": 18673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.415924549102783,
+      "learning_rate": 2.9367690058479538e-06,
+      "loss": 1.5163,
+      "step": 18674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2938053607940674,
+      "learning_rate": 2.935855263157895e-06,
+      "loss": 1.4761,
+      "step": 18675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4169561862945557,
+      "learning_rate": 2.9349415204678365e-06,
+      "loss": 1.5567,
+      "step": 18676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.296079635620117,
+      "learning_rate": 2.9340277777777783e-06,
+      "loss": 1.1686,
+      "step": 18677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3966147899627686,
+      "learning_rate": 2.9331140350877197e-06,
+      "loss": 1.4702,
+      "step": 18678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.358051061630249,
+      "learning_rate": 2.932200292397661e-06,
+      "loss": 1.3626,
+      "step": 18679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.61394739151001,
+      "learning_rate": 2.9312865497076024e-06,
+      "loss": 1.0997,
+      "step": 18680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5674710273742676,
+      "learning_rate": 2.9303728070175442e-06,
+      "loss": 1.4154,
+      "step": 18681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4828362464904785,
+      "learning_rate": 2.9294590643274856e-06,
+      "loss": 1.331,
+      "step": 18682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5393099784851074,
+      "learning_rate": 2.928545321637427e-06,
+      "loss": 1.5351,
+      "step": 18683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8891713619232178,
+      "learning_rate": 2.927631578947369e-06,
+      "loss": 1.1628,
+      "step": 18684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0496463775634766,
+      "learning_rate": 2.92671783625731e-06,
+      "loss": 1.1977,
+      "step": 18685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.588820695877075,
+      "learning_rate": 2.9258040935672515e-06,
+      "loss": 1.616,
+      "step": 18686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1779775619506836,
+      "learning_rate": 2.924890350877193e-06,
+      "loss": 1.3048,
+      "step": 18687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.079782485961914,
+      "learning_rate": 2.9239766081871347e-06,
+      "loss": 1.173,
+      "step": 18688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2944679260253906,
+      "learning_rate": 2.923062865497076e-06,
+      "loss": 1.3322,
+      "step": 18689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.333143472671509,
+      "learning_rate": 2.9221491228070175e-06,
+      "loss": 1.4361,
+      "step": 18690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6398098468780518,
+      "learning_rate": 2.9212353801169593e-06,
+      "loss": 1.7868,
+      "step": 18691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6835556030273438,
+      "learning_rate": 2.9203216374269006e-06,
+      "loss": 1.6198,
+      "step": 18692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9896366596221924,
+      "learning_rate": 2.9194078947368424e-06,
+      "loss": 1.6517,
+      "step": 18693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.468235731124878,
+      "learning_rate": 2.9184941520467842e-06,
+      "loss": 1.5051,
+      "step": 18694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6046133041381836,
+      "learning_rate": 2.9175804093567256e-06,
+      "loss": 1.4602,
+      "step": 18695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2218947410583496,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 1.1082,
+      "step": 18696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1922221183776855,
+      "learning_rate": 2.9157529239766088e-06,
+      "loss": 1.2441,
+      "step": 18697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.699632167816162,
+      "learning_rate": 2.91483918128655e-06,
+      "loss": 1.609,
+      "step": 18698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.156501531600952,
+      "learning_rate": 2.9139254385964915e-06,
+      "loss": 1.3913,
+      "step": 18699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5652592182159424,
+      "learning_rate": 2.913011695906433e-06,
+      "loss": 1.5862,
+      "step": 18700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2772154808044434,
+      "learning_rate": 2.9120979532163747e-06,
+      "loss": 1.7145,
+      "step": 18701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.367009401321411,
+      "learning_rate": 2.911184210526316e-06,
+      "loss": 1.3087,
+      "step": 18702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4601287841796875,
+      "learning_rate": 2.9102704678362574e-06,
+      "loss": 1.5681,
+      "step": 18703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3971238136291504,
+      "learning_rate": 2.9093567251461992e-06,
+      "loss": 1.3636,
+      "step": 18704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.689239263534546,
+      "learning_rate": 2.9084429824561406e-06,
+      "loss": 1.2811,
+      "step": 18705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7641406059265137,
+      "learning_rate": 2.907529239766082e-06,
+      "loss": 1.4748,
+      "step": 18706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.825119972229004,
+      "learning_rate": 2.9066154970760234e-06,
+      "loss": 1.7528,
+      "step": 18707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6958999633789062,
+      "learning_rate": 2.905701754385965e-06,
+      "loss": 1.504,
+      "step": 18708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.074350595474243,
+      "learning_rate": 2.9047880116959065e-06,
+      "loss": 1.2561,
+      "step": 18709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.983701229095459,
+      "learning_rate": 2.903874269005848e-06,
+      "loss": 1.7526,
+      "step": 18710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.833005666732788,
+      "learning_rate": 2.9029605263157897e-06,
+      "loss": 1.2337,
+      "step": 18711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2194323539733887,
+      "learning_rate": 2.902046783625731e-06,
+      "loss": 1.2847,
+      "step": 18712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6319966316223145,
+      "learning_rate": 2.9011330409356725e-06,
+      "loss": 1.4819,
+      "step": 18713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8856284618377686,
+      "learning_rate": 2.9002192982456147e-06,
+      "loss": 1.5918,
+      "step": 18714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3792724609375,
+      "learning_rate": 2.899305555555556e-06,
+      "loss": 1.64,
+      "step": 18715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5873727798461914,
+      "learning_rate": 2.8983918128654974e-06,
+      "loss": 1.6956,
+      "step": 18716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4835641384124756,
+      "learning_rate": 2.8974780701754392e-06,
+      "loss": 1.4638,
+      "step": 18717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5239439010620117,
+      "learning_rate": 2.8965643274853806e-06,
+      "loss": 1.5953,
+      "step": 18718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5129199028015137,
+      "learning_rate": 2.895650584795322e-06,
+      "loss": 1.3086,
+      "step": 18719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2789456844329834,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 1.692,
+      "step": 18720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1002824306488037,
+      "learning_rate": 2.893823099415205e-06,
+      "loss": 1.3286,
+      "step": 18721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8920834064483643,
+      "learning_rate": 2.8929093567251465e-06,
+      "loss": 1.5498,
+      "step": 18722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.744572401046753,
+      "learning_rate": 2.891995614035088e-06,
+      "loss": 1.5876,
+      "step": 18723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4751479625701904,
+      "learning_rate": 2.8910818713450297e-06,
+      "loss": 1.33,
+      "step": 18724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9262977838516235,
+      "learning_rate": 2.890168128654971e-06,
+      "loss": 1.2495,
+      "step": 18725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7930827140808105,
+      "learning_rate": 2.8892543859649124e-06,
+      "loss": 1.4216,
+      "step": 18726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.303133487701416,
+      "learning_rate": 2.888340643274854e-06,
+      "loss": 1.251,
+      "step": 18727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2975094318389893,
+      "learning_rate": 2.8874269005847956e-06,
+      "loss": 1.3915,
+      "step": 18728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6020612716674805,
+      "learning_rate": 2.886513157894737e-06,
+      "loss": 1.354,
+      "step": 18729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3467631340026855,
+      "learning_rate": 2.8855994152046784e-06,
+      "loss": 1.594,
+      "step": 18730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.564607620239258,
+      "learning_rate": 2.8846856725146197e-06,
+      "loss": 1.1975,
+      "step": 18731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8589367866516113,
+      "learning_rate": 2.8837719298245615e-06,
+      "loss": 1.5305,
+      "step": 18732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4513113498687744,
+      "learning_rate": 2.882858187134503e-06,
+      "loss": 1.5405,
+      "step": 18733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3291943073272705,
+      "learning_rate": 2.8819444444444443e-06,
+      "loss": 1.2338,
+      "step": 18734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3575820922851562,
+      "learning_rate": 2.881030701754386e-06,
+      "loss": 1.5204,
+      "step": 18735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4288816452026367,
+      "learning_rate": 2.880116959064328e-06,
+      "loss": 1.3637,
+      "step": 18736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.493515729904175,
+      "learning_rate": 2.8792032163742697e-06,
+      "loss": 1.4151,
+      "step": 18737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.280872106552124,
+      "learning_rate": 2.878289473684211e-06,
+      "loss": 1.4806,
+      "step": 18738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6785354614257812,
+      "learning_rate": 2.8773757309941524e-06,
+      "loss": 1.515,
+      "step": 18739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4942514896392822,
+      "learning_rate": 2.876461988304094e-06,
+      "loss": 1.5619,
+      "step": 18740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4002487659454346,
+      "learning_rate": 2.8755482456140356e-06,
+      "loss": 1.39,
+      "step": 18741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9333596229553223,
+      "learning_rate": 2.874634502923977e-06,
+      "loss": 1.2876,
+      "step": 18742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4469540119171143,
+      "learning_rate": 2.8737207602339183e-06,
+      "loss": 1.601,
+      "step": 18743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7736331224441528,
+      "learning_rate": 2.8728070175438597e-06,
+      "loss": 0.9327,
+      "step": 18744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6756653785705566,
+      "learning_rate": 2.8718932748538015e-06,
+      "loss": 1.5932,
+      "step": 18745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6627590656280518,
+      "learning_rate": 2.870979532163743e-06,
+      "loss": 1.4328,
+      "step": 18746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4231603145599365,
+      "learning_rate": 2.8700657894736843e-06,
+      "loss": 1.3961,
+      "step": 18747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.569244623184204,
+      "learning_rate": 2.869152046783626e-06,
+      "loss": 1.3573,
+      "step": 18748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.552119255065918,
+      "learning_rate": 2.8682383040935674e-06,
+      "loss": 1.5112,
+      "step": 18749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8188047409057617,
+      "learning_rate": 2.867324561403509e-06,
+      "loss": 1.2794,
+      "step": 18750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.873178243637085,
+      "learning_rate": 2.86641081871345e-06,
+      "loss": 1.6343,
+      "step": 18751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.574579954147339,
+      "learning_rate": 2.865497076023392e-06,
+      "loss": 1.1849,
+      "step": 18752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5385677814483643,
+      "learning_rate": 2.8645833333333334e-06,
+      "loss": 1.6398,
+      "step": 18753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.864750385284424,
+      "learning_rate": 2.8636695906432747e-06,
+      "loss": 1.6021,
+      "step": 18754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.168682336807251,
+      "learning_rate": 2.8627558479532165e-06,
+      "loss": 1.4821,
+      "step": 18755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4365234375,
+      "learning_rate": 2.861842105263158e-06,
+      "loss": 1.6555,
+      "step": 18756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.387519121170044,
+      "learning_rate": 2.8609283625730997e-06,
+      "loss": 1.948,
+      "step": 18757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.871349573135376,
+      "learning_rate": 2.8600146198830415e-06,
+      "loss": 1.7103,
+      "step": 18758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6969492435455322,
+      "learning_rate": 2.859100877192983e-06,
+      "loss": 1.6182,
+      "step": 18759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8558435440063477,
+      "learning_rate": 2.8581871345029243e-06,
+      "loss": 1.6802,
+      "step": 18760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4494974613189697,
+      "learning_rate": 2.857273391812866e-06,
+      "loss": 1.3696,
+      "step": 18761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.011521339416504,
+      "learning_rate": 2.8563596491228074e-06,
+      "loss": 1.3341,
+      "step": 18762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.389362335205078,
+      "learning_rate": 2.855445906432749e-06,
+      "loss": 1.0824,
+      "step": 18763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4520866870880127,
+      "learning_rate": 2.85453216374269e-06,
+      "loss": 1.5558,
+      "step": 18764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.74568510055542,
+      "learning_rate": 2.853618421052632e-06,
+      "loss": 1.6036,
+      "step": 18765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4296510219573975,
+      "learning_rate": 2.8527046783625733e-06,
+      "loss": 1.4902,
+      "step": 18766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.750431776046753,
+      "learning_rate": 2.8517909356725147e-06,
+      "loss": 1.7913,
+      "step": 18767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3624980449676514,
+      "learning_rate": 2.8508771929824565e-06,
+      "loss": 1.0765,
+      "step": 18768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1921703815460205,
+      "learning_rate": 2.849963450292398e-06,
+      "loss": 1.2452,
+      "step": 18769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6753599643707275,
+      "learning_rate": 2.8490497076023393e-06,
+      "loss": 1.5161,
+      "step": 18770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6061363220214844,
+      "learning_rate": 2.8481359649122806e-06,
+      "loss": 1.6728,
+      "step": 18771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.432643175125122,
+      "learning_rate": 2.8472222222222224e-06,
+      "loss": 1.6242,
+      "step": 18772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.617661714553833,
+      "learning_rate": 2.846308479532164e-06,
+      "loss": 1.5046,
+      "step": 18773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5245201587677,
+      "learning_rate": 2.845394736842105e-06,
+      "loss": 1.2842,
+      "step": 18774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6069018840789795,
+      "learning_rate": 2.844480994152047e-06,
+      "loss": 1.6143,
+      "step": 18775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3168141841888428,
+      "learning_rate": 2.8435672514619884e-06,
+      "loss": 1.4204,
+      "step": 18776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.600670337677002,
+      "learning_rate": 2.8426535087719297e-06,
+      "loss": 1.365,
+      "step": 18777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.292466640472412,
+      "learning_rate": 2.841739766081872e-06,
+      "loss": 1.2806,
+      "step": 18778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8838951587677,
+      "learning_rate": 2.8408260233918133e-06,
+      "loss": 1.3894,
+      "step": 18779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7907618284225464,
+      "learning_rate": 2.8399122807017547e-06,
+      "loss": 1.0391,
+      "step": 18780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5426626205444336,
+      "learning_rate": 2.8389985380116965e-06,
+      "loss": 1.3491,
+      "step": 18781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5502243041992188,
+      "learning_rate": 2.838084795321638e-06,
+      "loss": 1.5087,
+      "step": 18782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2682852745056152,
+      "learning_rate": 2.8371710526315793e-06,
+      "loss": 1.2631,
+      "step": 18783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.887075901031494,
+      "learning_rate": 2.8362573099415206e-06,
+      "loss": 1.7173,
+      "step": 18784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2272989749908447,
+      "learning_rate": 2.8353435672514624e-06,
+      "loss": 1.617,
+      "step": 18785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5837931632995605,
+      "learning_rate": 2.834429824561404e-06,
+      "loss": 1.5076,
+      "step": 18786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.804384708404541,
+      "learning_rate": 2.833516081871345e-06,
+      "loss": 1.8087,
+      "step": 18787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3556573390960693,
+      "learning_rate": 2.832602339181287e-06,
+      "loss": 1.4419,
+      "step": 18788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.673192024230957,
+      "learning_rate": 2.8316885964912283e-06,
+      "loss": 1.5601,
+      "step": 18789
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.510866165161133,
+      "learning_rate": 2.8307748538011697e-06,
+      "loss": 1.2657,
+      "step": 18790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4401187896728516,
+      "learning_rate": 2.829861111111111e-06,
+      "loss": 1.4388,
+      "step": 18791
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.479532480239868,
+      "learning_rate": 2.828947368421053e-06,
+      "loss": 1.655,
+      "step": 18792
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.245340585708618,
+      "learning_rate": 2.8280336257309943e-06,
+      "loss": 1.3057,
+      "step": 18793
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4347519874572754,
+      "learning_rate": 2.8271198830409356e-06,
+      "loss": 1.5112,
+      "step": 18794
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7858376502990723,
+      "learning_rate": 2.8262061403508774e-06,
+      "loss": 1.631,
+      "step": 18795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.254178762435913,
+      "learning_rate": 2.825292397660819e-06,
+      "loss": 1.175,
+      "step": 18796
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5616559982299805,
+      "learning_rate": 2.82437865497076e-06,
+      "loss": 1.3895,
+      "step": 18797
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4603898525238037,
+      "learning_rate": 2.8234649122807016e-06,
+      "loss": 1.2741,
+      "step": 18798
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.245011568069458,
+      "learning_rate": 2.8225511695906434e-06,
+      "loss": 1.4506,
+      "step": 18799
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.244189500808716,
+      "learning_rate": 2.821637426900585e-06,
+      "loss": 1.2557,
+      "step": 18800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.339839458465576,
+      "learning_rate": 2.820723684210527e-06,
+      "loss": 1.6249,
+      "step": 18801
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3636066913604736,
+      "learning_rate": 2.8198099415204683e-06,
+      "loss": 1.3886,
+      "step": 18802
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.757497549057007,
+      "learning_rate": 2.8188961988304097e-06,
+      "loss": 1.9676,
+      "step": 18803
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6870222091674805,
+      "learning_rate": 2.817982456140351e-06,
+      "loss": 1.8266,
+      "step": 18804
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.123713731765747,
+      "learning_rate": 2.817068713450293e-06,
+      "loss": 1.1783,
+      "step": 18805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.074449300765991,
+      "learning_rate": 2.8161549707602343e-06,
+      "loss": 1.1549,
+      "step": 18806
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.482187032699585,
+      "learning_rate": 2.8152412280701756e-06,
+      "loss": 1.1434,
+      "step": 18807
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6774277687072754,
+      "learning_rate": 2.8143274853801174e-06,
+      "loss": 1.5302,
+      "step": 18808
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6544222831726074,
+      "learning_rate": 2.813413742690059e-06,
+      "loss": 1.5654,
+      "step": 18809
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9441872835159302,
+      "learning_rate": 2.8125e-06,
+      "loss": 1.1731,
+      "step": 18810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.47857666015625,
+      "learning_rate": 2.8115862573099416e-06,
+      "loss": 1.4377,
+      "step": 18811
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.458517074584961,
+      "learning_rate": 2.8106725146198833e-06,
+      "loss": 1.2664,
+      "step": 18812
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1201324462890625,
+      "learning_rate": 2.8097587719298247e-06,
+      "loss": 1.3034,
+      "step": 18813
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1939072608947754,
+      "learning_rate": 2.808845029239766e-06,
+      "loss": 1.871,
+      "step": 18814
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.068171262741089,
+      "learning_rate": 2.807931286549708e-06,
+      "loss": 1.0116,
+      "step": 18815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3904876708984375,
+      "learning_rate": 2.8070175438596493e-06,
+      "loss": 1.4634,
+      "step": 18816
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.83504581451416,
+      "learning_rate": 2.8061038011695906e-06,
+      "loss": 1.7253,
+      "step": 18817
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3374762535095215,
+      "learning_rate": 2.805190058479532e-06,
+      "loss": 1.3271,
+      "step": 18818
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3069450855255127,
+      "learning_rate": 2.804276315789474e-06,
+      "loss": 1.2962,
+      "step": 18819
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0128538608551025,
+      "learning_rate": 2.803362573099415e-06,
+      "loss": 1.1461,
+      "step": 18820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0982284545898438,
+      "learning_rate": 2.8024488304093574e-06,
+      "loss": 1.6825,
+      "step": 18821
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3526933193206787,
+      "learning_rate": 2.8015350877192988e-06,
+      "loss": 1.4681,
+      "step": 18822
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.464189291000366,
+      "learning_rate": 2.80062134502924e-06,
+      "loss": 1.4512,
+      "step": 18823
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.728806257247925,
+      "learning_rate": 2.7997076023391815e-06,
+      "loss": 1.869,
+      "step": 18824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.508945941925049,
+      "learning_rate": 2.7987938596491233e-06,
+      "loss": 1.5015,
+      "step": 18825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1079459190368652,
+      "learning_rate": 2.7978801169590647e-06,
+      "loss": 1.4184,
+      "step": 18826
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.317856550216675,
+      "learning_rate": 2.796966374269006e-06,
+      "loss": 1.2337,
+      "step": 18827
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.166764974594116,
+      "learning_rate": 2.796052631578948e-06,
+      "loss": 1.647,
+      "step": 18828
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.484591245651245,
+      "learning_rate": 2.7951388888888893e-06,
+      "loss": 1.3425,
+      "step": 18829
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.06396484375,
+      "learning_rate": 2.7942251461988306e-06,
+      "loss": 1.3057,
+      "step": 18830
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5784265995025635,
+      "learning_rate": 2.793311403508772e-06,
+      "loss": 1.4406,
+      "step": 18831
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9459447860717773,
+      "learning_rate": 2.792397660818714e-06,
+      "loss": 1.2997,
+      "step": 18832
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9159789085388184,
+      "learning_rate": 2.791483918128655e-06,
+      "loss": 1.2297,
+      "step": 18833
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.220752477645874,
+      "learning_rate": 2.7905701754385966e-06,
+      "loss": 1.1436,
+      "step": 18834
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7000887393951416,
+      "learning_rate": 2.7896564327485383e-06,
+      "loss": 1.4675,
+      "step": 18835
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.973252773284912,
+      "learning_rate": 2.7887426900584797e-06,
+      "loss": 1.4356,
+      "step": 18836
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3669180870056152,
+      "learning_rate": 2.787828947368421e-06,
+      "loss": 1.2582,
+      "step": 18837
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9581353664398193,
+      "learning_rate": 2.7869152046783625e-06,
+      "loss": 1.2294,
+      "step": 18838
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.952251434326172,
+      "learning_rate": 2.7860014619883043e-06,
+      "loss": 1.3283,
+      "step": 18839
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7596707344055176,
+      "learning_rate": 2.7850877192982456e-06,
+      "loss": 1.5926,
+      "step": 18840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1649084091186523,
+      "learning_rate": 2.784173976608187e-06,
+      "loss": 1.4405,
+      "step": 18841
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7448084354400635,
+      "learning_rate": 2.783260233918129e-06,
+      "loss": 1.7878,
+      "step": 18842
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7086222171783447,
+      "learning_rate": 2.7823464912280706e-06,
+      "loss": 1.384,
+      "step": 18843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.569575071334839,
+      "learning_rate": 2.781432748538012e-06,
+      "loss": 1.4919,
+      "step": 18844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.249485492706299,
+      "learning_rate": 2.7805190058479538e-06,
+      "loss": 1.3108,
+      "step": 18845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1885430812835693,
+      "learning_rate": 2.779605263157895e-06,
+      "loss": 1.2236,
+      "step": 18846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7074296474456787,
+      "learning_rate": 2.7786915204678365e-06,
+      "loss": 1.3694,
+      "step": 18847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3625636100769043,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 1.4713,
+      "step": 18848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5391557216644287,
+      "learning_rate": 2.7768640350877197e-06,
+      "loss": 1.6824,
+      "step": 18849
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3593764305114746,
+      "learning_rate": 2.775950292397661e-06,
+      "loss": 1.3567,
+      "step": 18850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.121405601501465,
+      "learning_rate": 2.7750365497076025e-06,
+      "loss": 1.4398,
+      "step": 18851
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.159611701965332,
+      "learning_rate": 2.7741228070175443e-06,
+      "loss": 1.1152,
+      "step": 18852
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.107281446456909,
+      "learning_rate": 2.7732090643274856e-06,
+      "loss": 1.2336,
+      "step": 18853
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.152191638946533,
+      "learning_rate": 2.772295321637427e-06,
+      "loss": 1.6158,
+      "step": 18854
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2002387046813965,
+      "learning_rate": 2.7713815789473684e-06,
+      "loss": 1.4037,
+      "step": 18855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7979917526245117,
+      "learning_rate": 2.77046783625731e-06,
+      "loss": 1.5218,
+      "step": 18856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2792465686798096,
+      "learning_rate": 2.7695540935672516e-06,
+      "loss": 1.4522,
+      "step": 18857
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.987181544303894,
+      "learning_rate": 2.768640350877193e-06,
+      "loss": 1.4971,
+      "step": 18858
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.103729486465454,
+      "learning_rate": 2.7677266081871347e-06,
+      "loss": 1.3285,
+      "step": 18859
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4724087715148926,
+      "learning_rate": 2.766812865497076e-06,
+      "loss": 1.2115,
+      "step": 18860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.071082592010498,
+      "learning_rate": 2.7658991228070175e-06,
+      "loss": 1.1638,
+      "step": 18861
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.367774486541748,
+      "learning_rate": 2.764985380116959e-06,
+      "loss": 1.3089,
+      "step": 18862
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5394585132598877,
+      "learning_rate": 2.7640716374269006e-06,
+      "loss": 1.4322,
+      "step": 18863
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2246487140655518,
+      "learning_rate": 2.7631578947368424e-06,
+      "loss": 1.3166,
+      "step": 18864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4114654064178467,
+      "learning_rate": 2.7622441520467842e-06,
+      "loss": 1.5859,
+      "step": 18865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.323559045791626,
+      "learning_rate": 2.7613304093567256e-06,
+      "loss": 1.5609,
+      "step": 18866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6115171909332275,
+      "learning_rate": 2.760416666666667e-06,
+      "loss": 1.5631,
+      "step": 18867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5909969806671143,
+      "learning_rate": 2.7595029239766084e-06,
+      "loss": 1.1921,
+      "step": 18868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.689760208129883,
+      "learning_rate": 2.75858918128655e-06,
+      "loss": 1.4775,
+      "step": 18869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.323408842086792,
+      "learning_rate": 2.7576754385964915e-06,
+      "loss": 1.4358,
+      "step": 18870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.855720281600952,
+      "learning_rate": 2.756761695906433e-06,
+      "loss": 1.4497,
+      "step": 18871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2998626232147217,
+      "learning_rate": 2.7558479532163747e-06,
+      "loss": 1.4586,
+      "step": 18872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.314464807510376,
+      "learning_rate": 2.754934210526316e-06,
+      "loss": 1.5088,
+      "step": 18873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.62585186958313,
+      "learning_rate": 2.7540204678362575e-06,
+      "loss": 1.7071,
+      "step": 18874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2431209087371826,
+      "learning_rate": 2.753106725146199e-06,
+      "loss": 1.1955,
+      "step": 18875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1798088550567627,
+      "learning_rate": 2.7521929824561406e-06,
+      "loss": 1.3288,
+      "step": 18876
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9401612281799316,
+      "learning_rate": 2.751279239766082e-06,
+      "loss": 1.6412,
+      "step": 18877
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.259010076522827,
+      "learning_rate": 2.7503654970760234e-06,
+      "loss": 1.2482,
+      "step": 18878
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.416616916656494,
+      "learning_rate": 2.749451754385965e-06,
+      "loss": 1.4777,
+      "step": 18879
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.102022409439087,
+      "learning_rate": 2.7485380116959066e-06,
+      "loss": 1.5021,
+      "step": 18880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2144222259521484,
+      "learning_rate": 2.747624269005848e-06,
+      "loss": 1.1766,
+      "step": 18881
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5953333377838135,
+      "learning_rate": 2.7467105263157893e-06,
+      "loss": 1.5271,
+      "step": 18882
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1559085845947266,
+      "learning_rate": 2.745796783625731e-06,
+      "loss": 1.9948,
+      "step": 18883
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1958858966827393,
+      "learning_rate": 2.7448830409356725e-06,
+      "loss": 1.0479,
+      "step": 18884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.738438367843628,
+      "learning_rate": 2.7439692982456147e-06,
+      "loss": 1.6065,
+      "step": 18885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.26769757270813,
+      "learning_rate": 2.743055555555556e-06,
+      "loss": 1.4266,
+      "step": 18886
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2015187740325928,
+      "learning_rate": 2.7421418128654974e-06,
+      "loss": 1.2808,
+      "step": 18887
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.996577262878418,
+      "learning_rate": 2.741228070175439e-06,
+      "loss": 1.2574,
+      "step": 18888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1416189670562744,
+      "learning_rate": 2.7403143274853806e-06,
+      "loss": 1.3772,
+      "step": 18889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3094325065612793,
+      "learning_rate": 2.739400584795322e-06,
+      "loss": 1.0761,
+      "step": 18890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.460589647293091,
+      "learning_rate": 2.7384868421052634e-06,
+      "loss": 1.4629,
+      "step": 18891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1617560386657715,
+      "learning_rate": 2.737573099415205e-06,
+      "loss": 2.1198,
+      "step": 18892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.023406505584717,
+      "learning_rate": 2.7366593567251465e-06,
+      "loss": 1.1762,
+      "step": 18893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.335759162902832,
+      "learning_rate": 2.735745614035088e-06,
+      "loss": 1.0831,
+      "step": 18894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.307682991027832,
+      "learning_rate": 2.7348318713450293e-06,
+      "loss": 1.1946,
+      "step": 18895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1831719875335693,
+      "learning_rate": 2.733918128654971e-06,
+      "loss": 1.288,
+      "step": 18896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2811779975891113,
+      "learning_rate": 2.7330043859649125e-06,
+      "loss": 1.2148,
+      "step": 18897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8606443405151367,
+      "learning_rate": 2.732090643274854e-06,
+      "loss": 1.8862,
+      "step": 18898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4100027084350586,
+      "learning_rate": 2.7311769005847956e-06,
+      "loss": 1.3628,
+      "step": 18899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0141119956970215,
+      "learning_rate": 2.730263157894737e-06,
+      "loss": 1.7391,
+      "step": 18900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.080747604370117,
+      "learning_rate": 2.7293494152046784e-06,
+      "loss": 1.3535,
+      "step": 18901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.738025665283203,
+      "learning_rate": 2.7284356725146198e-06,
+      "loss": 1.2959,
+      "step": 18902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3399288654327393,
+      "learning_rate": 2.7275219298245616e-06,
+      "loss": 1.1449,
+      "step": 18903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.127124786376953,
+      "learning_rate": 2.726608187134503e-06,
+      "loss": 1.5573,
+      "step": 18904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.717801570892334,
+      "learning_rate": 2.7256944444444443e-06,
+      "loss": 1.8243,
+      "step": 18905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.95365571975708,
+      "learning_rate": 2.724780701754386e-06,
+      "loss": 1.6554,
+      "step": 18906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5441367626190186,
+      "learning_rate": 2.723866959064328e-06,
+      "loss": 1.2906,
+      "step": 18907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0371110439300537,
+      "learning_rate": 2.7229532163742693e-06,
+      "loss": 1.9378,
+      "step": 18908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.853980302810669,
+      "learning_rate": 2.722039473684211e-06,
+      "loss": 1.5398,
+      "step": 18909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0546510219573975,
+      "learning_rate": 2.7211257309941524e-06,
+      "loss": 1.649,
+      "step": 18910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1471800804138184,
+      "learning_rate": 2.720211988304094e-06,
+      "loss": 1.5932,
+      "step": 18911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.771139144897461,
+      "learning_rate": 2.7192982456140356e-06,
+      "loss": 1.9089,
+      "step": 18912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5172393321990967,
+      "learning_rate": 2.718384502923977e-06,
+      "loss": 1.4937,
+      "step": 18913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.630869150161743,
+      "learning_rate": 2.7174707602339184e-06,
+      "loss": 1.5518,
+      "step": 18914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7495555877685547,
+      "learning_rate": 2.7165570175438597e-06,
+      "loss": 2.014,
+      "step": 18915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3392367362976074,
+      "learning_rate": 2.7156432748538015e-06,
+      "loss": 1.1333,
+      "step": 18916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6496291160583496,
+      "learning_rate": 2.714729532163743e-06,
+      "loss": 1.2772,
+      "step": 18917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.237215280532837,
+      "learning_rate": 2.7138157894736843e-06,
+      "loss": 1.2243,
+      "step": 18918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5529086589813232,
+      "learning_rate": 2.712902046783626e-06,
+      "loss": 1.3197,
+      "step": 18919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.924499750137329,
+      "learning_rate": 2.7119883040935675e-06,
+      "loss": 1.6608,
+      "step": 18920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.395874261856079,
+      "learning_rate": 2.711074561403509e-06,
+      "loss": 1.6235,
+      "step": 18921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9112184047698975,
+      "learning_rate": 2.71016081871345e-06,
+      "loss": 1.7069,
+      "step": 18922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6018714904785156,
+      "learning_rate": 2.709247076023392e-06,
+      "loss": 1.4352,
+      "step": 18923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.245234727859497,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 1.9377,
+      "step": 18924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.491431713104248,
+      "learning_rate": 2.7074195906432748e-06,
+      "loss": 1.5001,
+      "step": 18925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3860726356506348,
+      "learning_rate": 2.7065058479532166e-06,
+      "loss": 1.3642,
+      "step": 18926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7063519954681396,
+      "learning_rate": 2.705592105263158e-06,
+      "loss": 1.228,
+      "step": 18927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3024861812591553,
+      "learning_rate": 2.7046783625730997e-06,
+      "loss": 1.0884,
+      "step": 18928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.438541889190674,
+      "learning_rate": 2.7037646198830415e-06,
+      "loss": 1.3472,
+      "step": 18929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4167730808258057,
+      "learning_rate": 2.702850877192983e-06,
+      "loss": 1.343,
+      "step": 18930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2170963287353516,
+      "learning_rate": 2.7019371345029243e-06,
+      "loss": 1.2178,
+      "step": 18931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.309075355529785,
+      "learning_rate": 2.701023391812866e-06,
+      "loss": 1.3594,
+      "step": 18932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1810057163238525,
+      "learning_rate": 2.7001096491228074e-06,
+      "loss": 1.2457,
+      "step": 18933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8190462589263916,
+      "learning_rate": 2.699195906432749e-06,
+      "loss": 1.6999,
+      "step": 18934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.637557029724121,
+      "learning_rate": 2.69828216374269e-06,
+      "loss": 1.424,
+      "step": 18935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.687281608581543,
+      "learning_rate": 2.697368421052632e-06,
+      "loss": 1.6845,
+      "step": 18936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3660573959350586,
+      "learning_rate": 2.6964546783625734e-06,
+      "loss": 1.4838,
+      "step": 18937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8940398693084717,
+      "learning_rate": 2.6955409356725147e-06,
+      "loss": 1.3808,
+      "step": 18938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.552943468093872,
+      "learning_rate": 2.6946271929824565e-06,
+      "loss": 1.6187,
+      "step": 18939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9237003326416016,
+      "learning_rate": 2.693713450292398e-06,
+      "loss": 1.5741,
+      "step": 18940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6316170692443848,
+      "learning_rate": 2.6927997076023393e-06,
+      "loss": 1.5309,
+      "step": 18941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.448115348815918,
+      "learning_rate": 2.6918859649122807e-06,
+      "loss": 1.0382,
+      "step": 18942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5427000522613525,
+      "learning_rate": 2.6909722222222225e-06,
+      "loss": 1.7502,
+      "step": 18943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3667566776275635,
+      "learning_rate": 2.690058479532164e-06,
+      "loss": 1.4906,
+      "step": 18944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3610188961029053,
+      "learning_rate": 2.689144736842105e-06,
+      "loss": 1.2683,
+      "step": 18945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.119171142578125,
+      "learning_rate": 2.688230994152047e-06,
+      "loss": 1.3877,
+      "step": 18946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8495922088623047,
+      "learning_rate": 2.6873172514619884e-06,
+      "loss": 2.0117,
+      "step": 18947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3916890621185303,
+      "learning_rate": 2.6864035087719298e-06,
+      "loss": 1.5491,
+      "step": 18948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7974424362182617,
+      "learning_rate": 2.685489766081872e-06,
+      "loss": 1.5378,
+      "step": 18949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.594632387161255,
+      "learning_rate": 2.6845760233918133e-06,
+      "loss": 1.4421,
+      "step": 18950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.670111656188965,
+      "learning_rate": 2.6836622807017547e-06,
+      "loss": 1.7691,
+      "step": 18951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8892931938171387,
+      "learning_rate": 2.6827485380116965e-06,
+      "loss": 1.4444,
+      "step": 18952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1256823539733887,
+      "learning_rate": 2.681834795321638e-06,
+      "loss": 1.1886,
+      "step": 18953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2934141159057617,
+      "learning_rate": 2.6809210526315793e-06,
+      "loss": 1.2882,
+      "step": 18954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.486055850982666,
+      "learning_rate": 2.6800073099415206e-06,
+      "loss": 1.2328,
+      "step": 18955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.597015380859375,
+      "learning_rate": 2.6790935672514624e-06,
+      "loss": 1.1776,
+      "step": 18956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6915018558502197,
+      "learning_rate": 2.678179824561404e-06,
+      "loss": 1.4137,
+      "step": 18957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3511247634887695,
+      "learning_rate": 2.677266081871345e-06,
+      "loss": 1.1983,
+      "step": 18958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.050133466720581,
+      "learning_rate": 2.676352339181287e-06,
+      "loss": 1.5216,
+      "step": 18959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0144569873809814,
+      "learning_rate": 2.6754385964912284e-06,
+      "loss": 1.0281,
+      "step": 18960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4606857299804688,
+      "learning_rate": 2.6745248538011697e-06,
+      "loss": 1.652,
+      "step": 18961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.594095468521118,
+      "learning_rate": 2.673611111111111e-06,
+      "loss": 1.5,
+      "step": 18962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.739145040512085,
+      "learning_rate": 2.672697368421053e-06,
+      "loss": 1.8568,
+      "step": 18963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3602492809295654,
+      "learning_rate": 2.6717836257309943e-06,
+      "loss": 1.6426,
+      "step": 18964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.321934461593628,
+      "learning_rate": 2.6708698830409357e-06,
+      "loss": 1.2688,
+      "step": 18965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7619826793670654,
+      "learning_rate": 2.6699561403508775e-06,
+      "loss": 1.3311,
+      "step": 18966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.684921979904175,
+      "learning_rate": 2.669042397660819e-06,
+      "loss": 1.3951,
+      "step": 18967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2107207775115967,
+      "learning_rate": 2.66812865497076e-06,
+      "loss": 1.3925,
+      "step": 18968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2220842838287354,
+      "learning_rate": 2.6672149122807016e-06,
+      "loss": 1.1947,
+      "step": 18969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.245945692062378,
+      "learning_rate": 2.6663011695906434e-06,
+      "loss": 1.5141,
+      "step": 18970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2195918560028076,
+      "learning_rate": 2.665387426900585e-06,
+      "loss": 1.4228,
+      "step": 18971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.525703191757202,
+      "learning_rate": 2.664473684210527e-06,
+      "loss": 1.2923,
+      "step": 18972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.340885877609253,
+      "learning_rate": 2.6635599415204683e-06,
+      "loss": 1.6274,
+      "step": 18973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.059922456741333,
+      "learning_rate": 2.6626461988304097e-06,
+      "loss": 1.0853,
+      "step": 18974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.554739236831665,
+      "learning_rate": 2.661732456140351e-06,
+      "loss": 1.7966,
+      "step": 18975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8944811820983887,
+      "learning_rate": 2.660818713450293e-06,
+      "loss": 1.8375,
+      "step": 18976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.989964485168457,
+      "learning_rate": 2.6599049707602343e-06,
+      "loss": 1.6762,
+      "step": 18977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9338090419769287,
+      "learning_rate": 2.6589912280701756e-06,
+      "loss": 1.1813,
+      "step": 18978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.484787702560425,
+      "learning_rate": 2.658077485380117e-06,
+      "loss": 1.427,
+      "step": 18979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.672987699508667,
+      "learning_rate": 2.657163742690059e-06,
+      "loss": 1.4778,
+      "step": 18980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2411892414093018,
+      "learning_rate": 2.65625e-06,
+      "loss": 1.2031,
+      "step": 18981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2020328044891357,
+      "learning_rate": 2.6553362573099416e-06,
+      "loss": 1.5928,
+      "step": 18982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2850594520568848,
+      "learning_rate": 2.6544225146198834e-06,
+      "loss": 1.2936,
+      "step": 18983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3997573852539062,
+      "learning_rate": 2.6535087719298247e-06,
+      "loss": 1.2702,
+      "step": 18984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3689730167388916,
+      "learning_rate": 2.652595029239766e-06,
+      "loss": 1.32,
+      "step": 18985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.991206169128418,
+      "learning_rate": 2.6516812865497075e-06,
+      "loss": 1.5413,
+      "step": 18986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0256736278533936,
+      "learning_rate": 2.6507675438596493e-06,
+      "loss": 1.5416,
+      "step": 18987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2528655529022217,
+      "learning_rate": 2.6498538011695907e-06,
+      "loss": 1.836,
+      "step": 18988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.806121349334717,
+      "learning_rate": 2.648940058479532e-06,
+      "loss": 1.7356,
+      "step": 18989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5490496158599854,
+      "learning_rate": 2.648026315789474e-06,
+      "loss": 1.3189,
+      "step": 18990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.10782790184021,
+      "learning_rate": 2.647112573099415e-06,
+      "loss": 1.1272,
+      "step": 18991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.228178024291992,
+      "learning_rate": 2.646198830409357e-06,
+      "loss": 1.2908,
+      "step": 18992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6273300647735596,
+      "learning_rate": 2.645285087719299e-06,
+      "loss": 1.2988,
+      "step": 18993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5260655879974365,
+      "learning_rate": 2.64437134502924e-06,
+      "loss": 1.4807,
+      "step": 18994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5555992126464844,
+      "learning_rate": 2.6434576023391816e-06,
+      "loss": 1.4928,
+      "step": 18995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.539933919906616,
+      "learning_rate": 2.6425438596491233e-06,
+      "loss": 1.4956,
+      "step": 18996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.696336269378662,
+      "learning_rate": 2.6416301169590647e-06,
+      "loss": 1.717,
+      "step": 18997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.433943271636963,
+      "learning_rate": 2.640716374269006e-06,
+      "loss": 1.3528,
+      "step": 18998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6130318641662598,
+      "learning_rate": 2.6398026315789475e-06,
+      "loss": 1.4694,
+      "step": 18999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.850424289703369,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 1.4861,
+      "step": 19000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0346648693084717,
+      "learning_rate": 2.6379751461988306e-06,
+      "loss": 2.012,
+      "step": 19001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.387765645980835,
+      "learning_rate": 2.637061403508772e-06,
+      "loss": 1.3982,
+      "step": 19002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5828697681427,
+      "learning_rate": 2.636147660818714e-06,
+      "loss": 1.5812,
+      "step": 19003
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7790799140930176,
+      "learning_rate": 2.635233918128655e-06,
+      "loss": 1.9736,
+      "step": 19004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.417198657989502,
+      "learning_rate": 2.6343201754385966e-06,
+      "loss": 1.4429,
+      "step": 19005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.913037657737732,
+      "learning_rate": 2.633406432748538e-06,
+      "loss": 0.9643,
+      "step": 19006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.170984983444214,
+      "learning_rate": 2.6324926900584797e-06,
+      "loss": 1.3575,
+      "step": 19007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.847907781600952,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 1.6548,
+      "step": 19008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.418928623199463,
+      "learning_rate": 2.6306652046783625e-06,
+      "loss": 1.3541,
+      "step": 19009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.187086343765259,
+      "learning_rate": 2.6297514619883043e-06,
+      "loss": 1.247,
+      "step": 19010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0351855754852295,
+      "learning_rate": 2.6288377192982457e-06,
+      "loss": 1.782,
+      "step": 19011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.106956720352173,
+      "learning_rate": 2.627923976608187e-06,
+      "loss": 1.3109,
+      "step": 19012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3772318363189697,
+      "learning_rate": 2.6270102339181284e-06,
+      "loss": 1.2914,
+      "step": 19013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.353915214538574,
+      "learning_rate": 2.6260964912280706e-06,
+      "loss": 1.4482,
+      "step": 19014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.116323232650757,
+      "learning_rate": 2.625182748538012e-06,
+      "loss": 1.1619,
+      "step": 19015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9341875314712524,
+      "learning_rate": 2.624269005847954e-06,
+      "loss": 1.1133,
+      "step": 19016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7978503704071045,
+      "learning_rate": 2.623355263157895e-06,
+      "loss": 1.1698,
+      "step": 19017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4281864166259766,
+      "learning_rate": 2.6224415204678366e-06,
+      "loss": 1.4408,
+      "step": 19018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6768453121185303,
+      "learning_rate": 2.621527777777778e-06,
+      "loss": 1.2386,
+      "step": 19019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.720475435256958,
+      "learning_rate": 2.6206140350877197e-06,
+      "loss": 1.4685,
+      "step": 19020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5393991470336914,
+      "learning_rate": 2.619700292397661e-06,
+      "loss": 1.3306,
+      "step": 19021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3690085411071777,
+      "learning_rate": 2.6187865497076025e-06,
+      "loss": 1.6827,
+      "step": 19022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7720229625701904,
+      "learning_rate": 2.6178728070175443e-06,
+      "loss": 1.2587,
+      "step": 19023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2921886444091797,
+      "learning_rate": 2.6169590643274856e-06,
+      "loss": 1.217,
+      "step": 19024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.656467914581299,
+      "learning_rate": 2.616045321637427e-06,
+      "loss": 1.3533,
+      "step": 19025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6704885959625244,
+      "learning_rate": 2.6151315789473684e-06,
+      "loss": 1.4153,
+      "step": 19026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.29921555519104,
+      "learning_rate": 2.61421783625731e-06,
+      "loss": 1.5151,
+      "step": 19027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4049441814422607,
+      "learning_rate": 2.6133040935672516e-06,
+      "loss": 1.4493,
+      "step": 19028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0777814388275146,
+      "learning_rate": 2.612390350877193e-06,
+      "loss": 1.4483,
+      "step": 19029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.295077323913574,
+      "learning_rate": 2.6114766081871347e-06,
+      "loss": 1.2943,
+      "step": 19030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0350966453552246,
+      "learning_rate": 2.610562865497076e-06,
+      "loss": 2.0858,
+      "step": 19031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4831011295318604,
+      "learning_rate": 2.6096491228070175e-06,
+      "loss": 1.156,
+      "step": 19032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0833702087402344,
+      "learning_rate": 2.608735380116959e-06,
+      "loss": 1.2314,
+      "step": 19033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.54362416267395,
+      "learning_rate": 2.6078216374269007e-06,
+      "loss": 1.5589,
+      "step": 19034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2821969985961914,
+      "learning_rate": 2.6069078947368425e-06,
+      "loss": 1.4354,
+      "step": 19035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3787691593170166,
+      "learning_rate": 2.6059941520467843e-06,
+      "loss": 1.7011,
+      "step": 19036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5310332775115967,
+      "learning_rate": 2.6050804093567256e-06,
+      "loss": 1.4868,
+      "step": 19037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.313150405883789,
+      "learning_rate": 2.604166666666667e-06,
+      "loss": 1.5714,
+      "step": 19038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6959683895111084,
+      "learning_rate": 2.6032529239766084e-06,
+      "loss": 1.5274,
+      "step": 19039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9097633361816406,
+      "learning_rate": 2.60233918128655e-06,
+      "loss": 1.6588,
+      "step": 19040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1935770511627197,
+      "learning_rate": 2.6014254385964916e-06,
+      "loss": 1.5183,
+      "step": 19041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.581740617752075,
+      "learning_rate": 2.600511695906433e-06,
+      "loss": 1.248,
+      "step": 19042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4767587184906006,
+      "learning_rate": 2.5995979532163747e-06,
+      "loss": 1.5245,
+      "step": 19043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.424928665161133,
+      "learning_rate": 2.598684210526316e-06,
+      "loss": 1.5941,
+      "step": 19044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7355051040649414,
+      "learning_rate": 2.5977704678362575e-06,
+      "loss": 1.5201,
+      "step": 19045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.50689959526062,
+      "learning_rate": 2.596856725146199e-06,
+      "loss": 1.6938,
+      "step": 19046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1561636924743652,
+      "learning_rate": 2.5959429824561406e-06,
+      "loss": 1.3123,
+      "step": 19047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0437772274017334,
+      "learning_rate": 2.595029239766082e-06,
+      "loss": 1.6218,
+      "step": 19048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2918331623077393,
+      "learning_rate": 2.5941154970760234e-06,
+      "loss": 1.4933,
+      "step": 19049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.81878662109375,
+      "learning_rate": 2.593201754385965e-06,
+      "loss": 1.6294,
+      "step": 19050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.987734794616699,
+      "learning_rate": 2.5922880116959066e-06,
+      "loss": 1.7707,
+      "step": 19051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3228189945220947,
+      "learning_rate": 2.591374269005848e-06,
+      "loss": 1.1614,
+      "step": 19052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.567335605621338,
+      "learning_rate": 2.5904605263157893e-06,
+      "loss": 1.6802,
+      "step": 19053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0412700176239014,
+      "learning_rate": 2.589546783625731e-06,
+      "loss": 1.2526,
+      "step": 19054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.389737367630005,
+      "learning_rate": 2.5886330409356725e-06,
+      "loss": 1.6212,
+      "step": 19055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.960137128829956,
+      "learning_rate": 2.5877192982456147e-06,
+      "loss": 1.6053,
+      "step": 19056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2745144367218018,
+      "learning_rate": 2.586805555555556e-06,
+      "loss": 1.3732,
+      "step": 19057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.384089708328247,
+      "learning_rate": 2.5858918128654975e-06,
+      "loss": 1.4487,
+      "step": 19058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2923548221588135,
+      "learning_rate": 2.584978070175439e-06,
+      "loss": 1.2318,
+      "step": 19059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.43092679977417,
+      "learning_rate": 2.5840643274853806e-06,
+      "loss": 1.7309,
+      "step": 19060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3150947093963623,
+      "learning_rate": 2.583150584795322e-06,
+      "loss": 1.4587,
+      "step": 19061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8763949871063232,
+      "learning_rate": 2.5822368421052634e-06,
+      "loss": 1.2065,
+      "step": 19062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7912354469299316,
+      "learning_rate": 2.581323099415205e-06,
+      "loss": 1.7245,
+      "step": 19063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.682063341140747,
+      "learning_rate": 2.5804093567251465e-06,
+      "loss": 1.4203,
+      "step": 19064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.334908962249756,
+      "learning_rate": 2.579495614035088e-06,
+      "loss": 1.2273,
+      "step": 19065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.381840944290161,
+      "learning_rate": 2.5785818713450293e-06,
+      "loss": 1.3408,
+      "step": 19066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1290383338928223,
+      "learning_rate": 2.577668128654971e-06,
+      "loss": 1.4527,
+      "step": 19067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.029308319091797,
+      "learning_rate": 2.5767543859649125e-06,
+      "loss": 1.5361,
+      "step": 19068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.299194574356079,
+      "learning_rate": 2.575840643274854e-06,
+      "loss": 1.5598,
+      "step": 19069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.629582166671753,
+      "learning_rate": 2.5749269005847956e-06,
+      "loss": 1.3907,
+      "step": 19070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.519151210784912,
+      "learning_rate": 2.574013157894737e-06,
+      "loss": 1.3682,
+      "step": 19071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1262335777282715,
+      "learning_rate": 2.5730994152046784e-06,
+      "loss": 1.8672,
+      "step": 19072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.528756618499756,
+      "learning_rate": 2.5721856725146198e-06,
+      "loss": 1.55,
+      "step": 19073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4068031311035156,
+      "learning_rate": 2.5712719298245616e-06,
+      "loss": 1.5316,
+      "step": 19074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.371678590774536,
+      "learning_rate": 2.570358187134503e-06,
+      "loss": 1.2788,
+      "step": 19075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.44718599319458,
+      "learning_rate": 2.5694444444444443e-06,
+      "loss": 1.4888,
+      "step": 19076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2780494689941406,
+      "learning_rate": 2.568530701754386e-06,
+      "loss": 1.456,
+      "step": 19077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2061455249786377,
+      "learning_rate": 2.567616959064328e-06,
+      "loss": 1.3838,
+      "step": 19078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9521241188049316,
+      "learning_rate": 2.5667032163742693e-06,
+      "loss": 1.6956,
+      "step": 19079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.808476686477661,
+      "learning_rate": 2.565789473684211e-06,
+      "loss": 1.6164,
+      "step": 19080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5280463695526123,
+      "learning_rate": 2.5648757309941525e-06,
+      "loss": 1.6319,
+      "step": 19081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.503427028656006,
+      "learning_rate": 2.563961988304094e-06,
+      "loss": 1.5753,
+      "step": 19082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4287736415863037,
+      "learning_rate": 2.5630482456140356e-06,
+      "loss": 1.6618,
+      "step": 19083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5090761184692383,
+      "learning_rate": 2.562134502923977e-06,
+      "loss": 1.3812,
+      "step": 19084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8672988414764404,
+      "learning_rate": 2.5612207602339184e-06,
+      "loss": 1.5331,
+      "step": 19085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.873199462890625,
+      "learning_rate": 2.5603070175438598e-06,
+      "loss": 1.6572,
+      "step": 19086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3321897983551025,
+      "learning_rate": 2.5593932748538015e-06,
+      "loss": 1.2631,
+      "step": 19087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2476015090942383,
+      "learning_rate": 2.558479532163743e-06,
+      "loss": 1.2902,
+      "step": 19088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6578376293182373,
+      "learning_rate": 2.5575657894736843e-06,
+      "loss": 1.8515,
+      "step": 19089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6135623455047607,
+      "learning_rate": 2.556652046783626e-06,
+      "loss": 1.5476,
+      "step": 19090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.847768545150757,
+      "learning_rate": 2.5557383040935675e-06,
+      "loss": 1.7564,
+      "step": 19091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.039789915084839,
+      "learning_rate": 2.554824561403509e-06,
+      "loss": 1.8145,
+      "step": 19092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.39109468460083,
+      "learning_rate": 2.5539108187134502e-06,
+      "loss": 1.3364,
+      "step": 19093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.566032886505127,
+      "learning_rate": 2.552997076023392e-06,
+      "loss": 1.2741,
+      "step": 19094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.209200859069824,
+      "learning_rate": 2.5520833333333334e-06,
+      "loss": 1.2548,
+      "step": 19095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.447986364364624,
+      "learning_rate": 2.5511695906432748e-06,
+      "loss": 1.2764,
+      "step": 19096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.11099910736084,
+      "learning_rate": 2.550255847953216e-06,
+      "loss": 1.1261,
+      "step": 19097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.187833309173584,
+      "learning_rate": 2.549342105263158e-06,
+      "loss": 1.2792,
+      "step": 19098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6222479343414307,
+      "learning_rate": 2.5484283625730997e-06,
+      "loss": 1.614,
+      "step": 19099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4336400032043457,
+      "learning_rate": 2.5475146198830415e-06,
+      "loss": 1.4245,
+      "step": 19100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9511008262634277,
+      "learning_rate": 2.546600877192983e-06,
+      "loss": 1.5638,
+      "step": 19101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.977470636367798,
+      "learning_rate": 2.5456871345029243e-06,
+      "loss": 1.8128,
+      "step": 19102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.493155002593994,
+      "learning_rate": 2.5447733918128657e-06,
+      "loss": 1.4489,
+      "step": 19103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1130921840667725,
+      "learning_rate": 2.5438596491228075e-06,
+      "loss": 1.2057,
+      "step": 19104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.565908908843994,
+      "learning_rate": 2.542945906432749e-06,
+      "loss": 1.3102,
+      "step": 19105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.06443452835083,
+      "learning_rate": 2.54203216374269e-06,
+      "loss": 1.2055,
+      "step": 19106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3674588203430176,
+      "learning_rate": 2.541118421052632e-06,
+      "loss": 1.4944,
+      "step": 19107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.020695686340332,
+      "learning_rate": 2.5402046783625734e-06,
+      "loss": 1.2211,
+      "step": 19108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6829285621643066,
+      "learning_rate": 2.5392909356725148e-06,
+      "loss": 1.7102,
+      "step": 19109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5984532833099365,
+      "learning_rate": 2.538377192982456e-06,
+      "loss": 1.5064,
+      "step": 19110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.553163528442383,
+      "learning_rate": 2.537463450292398e-06,
+      "loss": 1.4735,
+      "step": 19111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.017685651779175,
+      "learning_rate": 2.5365497076023393e-06,
+      "loss": 1.8296,
+      "step": 19112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.48549222946167,
+      "learning_rate": 2.5356359649122807e-06,
+      "loss": 1.4396,
+      "step": 19113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.627173662185669,
+      "learning_rate": 2.5347222222222225e-06,
+      "loss": 1.7812,
+      "step": 19114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6097559928894043,
+      "learning_rate": 2.533808479532164e-06,
+      "loss": 1.2634,
+      "step": 19115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.493049144744873,
+      "learning_rate": 2.5328947368421052e-06,
+      "loss": 1.4584,
+      "step": 19116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1714799404144287,
+      "learning_rate": 2.5319809941520466e-06,
+      "loss": 1.2018,
+      "step": 19117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5706191062927246,
+      "learning_rate": 2.5310672514619884e-06,
+      "loss": 1.5271,
+      "step": 19118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3528196811676025,
+      "learning_rate": 2.5301535087719298e-06,
+      "loss": 1.4611,
+      "step": 19119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7457082271575928,
+      "learning_rate": 2.529239766081872e-06,
+      "loss": 1.4854,
+      "step": 19120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6626052856445312,
+      "learning_rate": 2.5283260233918134e-06,
+      "loss": 1.2466,
+      "step": 19121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.883791208267212,
+      "learning_rate": 2.5274122807017547e-06,
+      "loss": 1.5493,
+      "step": 19122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3722023963928223,
+      "learning_rate": 2.526498538011696e-06,
+      "loss": 1.4744,
+      "step": 19123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4964451789855957,
+      "learning_rate": 2.525584795321638e-06,
+      "loss": 1.5737,
+      "step": 19124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.663745164871216,
+      "learning_rate": 2.5246710526315793e-06,
+      "loss": 1.4976,
+      "step": 19125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.188046932220459,
+      "learning_rate": 2.5237573099415207e-06,
+      "loss": 1.6956,
+      "step": 19126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0732784271240234,
+      "learning_rate": 2.5228435672514625e-06,
+      "loss": 1.43,
+      "step": 19127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.385739326477051,
+      "learning_rate": 2.521929824561404e-06,
+      "loss": 1.2409,
+      "step": 19128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.921226978302002,
+      "learning_rate": 2.521016081871345e-06,
+      "loss": 1.4541,
+      "step": 19129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5746750831604004,
+      "learning_rate": 2.5201023391812866e-06,
+      "loss": 1.3808,
+      "step": 19130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8708293437957764,
+      "learning_rate": 2.5191885964912284e-06,
+      "loss": 1.8221,
+      "step": 19131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4070262908935547,
+      "learning_rate": 2.5182748538011698e-06,
+      "loss": 1.5071,
+      "step": 19132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0382657051086426,
+      "learning_rate": 2.517361111111111e-06,
+      "loss": 1.0971,
+      "step": 19133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.816236734390259,
+      "learning_rate": 2.516447368421053e-06,
+      "loss": 1.704,
+      "step": 19134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8948733806610107,
+      "learning_rate": 2.5155336257309943e-06,
+      "loss": 1.6613,
+      "step": 19135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4222638607025146,
+      "learning_rate": 2.5146198830409357e-06,
+      "loss": 1.2803,
+      "step": 19136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1746153831481934,
+      "learning_rate": 2.513706140350877e-06,
+      "loss": 1.0731,
+      "step": 19137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.210873603820801,
+      "learning_rate": 2.512792397660819e-06,
+      "loss": 1.6391,
+      "step": 19138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9924410581588745,
+      "learning_rate": 2.5118786549707602e-06,
+      "loss": 1.2857,
+      "step": 19139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6537816524505615,
+      "learning_rate": 2.5109649122807016e-06,
+      "loss": 1.5266,
+      "step": 19140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7467970848083496,
+      "learning_rate": 2.5100511695906434e-06,
+      "loss": 1.3948,
+      "step": 19141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4490129947662354,
+      "learning_rate": 2.509137426900585e-06,
+      "loss": 1.4186,
+      "step": 19142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.54736065864563,
+      "learning_rate": 2.5082236842105266e-06,
+      "loss": 1.4499,
+      "step": 19143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.731069326400757,
+      "learning_rate": 2.5073099415204684e-06,
+      "loss": 1.5628,
+      "step": 19144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.529289722442627,
+      "learning_rate": 2.5063961988304097e-06,
+      "loss": 1.4607,
+      "step": 19145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2606167793273926,
+      "learning_rate": 2.505482456140351e-06,
+      "loss": 1.3368,
+      "step": 19146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.403266429901123,
+      "learning_rate": 2.504568713450293e-06,
+      "loss": 1.4313,
+      "step": 19147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0562171936035156,
+      "learning_rate": 2.5036549707602343e-06,
+      "loss": 1.461,
+      "step": 19148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6460659503936768,
+      "learning_rate": 2.5027412280701757e-06,
+      "loss": 1.2001,
+      "step": 19149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.467433452606201,
+      "learning_rate": 2.501827485380117e-06,
+      "loss": 1.3006,
+      "step": 19150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.414341688156128,
+      "learning_rate": 2.500913742690059e-06,
+      "loss": 1.4256,
+      "step": 19151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1995866298675537,
+      "learning_rate": 2.5e-06,
+      "loss": 1.7414,
+      "step": 19152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.48848295211792,
+      "learning_rate": 2.4990862573099416e-06,
+      "loss": 1.4151,
+      "step": 19153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.409557342529297,
+      "learning_rate": 2.4981725146198834e-06,
+      "loss": 1.0254,
+      "step": 19154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5916528701782227,
+      "learning_rate": 2.4972587719298248e-06,
+      "loss": 1.5159,
+      "step": 19155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.465538263320923,
+      "learning_rate": 2.496345029239766e-06,
+      "loss": 1.5264,
+      "step": 19156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.731191396713257,
+      "learning_rate": 2.4954312865497075e-06,
+      "loss": 1.4852,
+      "step": 19157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.501661777496338,
+      "learning_rate": 2.4945175438596493e-06,
+      "loss": 1.3457,
+      "step": 19158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.636512517929077,
+      "learning_rate": 2.493603801169591e-06,
+      "loss": 1.7243,
+      "step": 19159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2759954929351807,
+      "learning_rate": 2.4926900584795325e-06,
+      "loss": 1.5851,
+      "step": 19160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6737804412841797,
+      "learning_rate": 2.491776315789474e-06,
+      "loss": 1.4442,
+      "step": 19161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.390256404876709,
+      "learning_rate": 2.4908625730994152e-06,
+      "loss": 1.3492,
+      "step": 19162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4929635524749756,
+      "learning_rate": 2.489948830409357e-06,
+      "loss": 1.2835,
+      "step": 19163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.873605251312256,
+      "learning_rate": 2.4890350877192984e-06,
+      "loss": 1.5578,
+      "step": 19164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7439188957214355,
+      "learning_rate": 2.4881213450292398e-06,
+      "loss": 1.4713,
+      "step": 19165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1348185539245605,
+      "learning_rate": 2.4872076023391816e-06,
+      "loss": 1.1128,
+      "step": 19166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.425762891769409,
+      "learning_rate": 2.486293859649123e-06,
+      "loss": 1.2561,
+      "step": 19167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5501396656036377,
+      "learning_rate": 2.4853801169590643e-06,
+      "loss": 1.2496,
+      "step": 19168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.265824317932129,
+      "learning_rate": 2.484466374269006e-06,
+      "loss": 1.2338,
+      "step": 19169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1415209770202637,
+      "learning_rate": 2.4835526315789475e-06,
+      "loss": 1.7141,
+      "step": 19170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8013319969177246,
+      "learning_rate": 2.4826388888888893e-06,
+      "loss": 1.5733,
+      "step": 19171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1621270179748535,
+      "learning_rate": 2.4817251461988307e-06,
+      "loss": 1.3741,
+      "step": 19172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.750997304916382,
+      "learning_rate": 2.480811403508772e-06,
+      "loss": 1.4329,
+      "step": 19173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.277996063232422,
+      "learning_rate": 2.479897660818714e-06,
+      "loss": 1.9251,
+      "step": 19174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4787416458129883,
+      "learning_rate": 2.478983918128655e-06,
+      "loss": 1.4487,
+      "step": 19175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.44244122505188,
+      "learning_rate": 2.4780701754385966e-06,
+      "loss": 1.4802,
+      "step": 19176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.713200092315674,
+      "learning_rate": 2.477156432748538e-06,
+      "loss": 1.5451,
+      "step": 19177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3898558616638184,
+      "learning_rate": 2.4762426900584798e-06,
+      "loss": 1.3522,
+      "step": 19178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5022451877593994,
+      "learning_rate": 2.4753289473684215e-06,
+      "loss": 1.4201,
+      "step": 19179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5531764030456543,
+      "learning_rate": 2.474415204678363e-06,
+      "loss": 1.5821,
+      "step": 19180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7299389839172363,
+      "learning_rate": 2.4735014619883043e-06,
+      "loss": 1.5703,
+      "step": 19181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.557311773300171,
+      "learning_rate": 2.4725877192982457e-06,
+      "loss": 1.3857,
+      "step": 19182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5246565341949463,
+      "learning_rate": 2.4716739766081875e-06,
+      "loss": 1.4212,
+      "step": 19183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.085810661315918,
+      "learning_rate": 2.470760233918129e-06,
+      "loss": 1.1105,
+      "step": 19184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7855305671691895,
+      "learning_rate": 2.4698464912280702e-06,
+      "loss": 1.4062,
+      "step": 19185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.620288372039795,
+      "learning_rate": 2.468932748538012e-06,
+      "loss": 1.4685,
+      "step": 19186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4515271186828613,
+      "learning_rate": 2.4680190058479534e-06,
+      "loss": 1.3925,
+      "step": 19187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.457268714904785,
+      "learning_rate": 2.4671052631578948e-06,
+      "loss": 2.0538,
+      "step": 19188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.787885904312134,
+      "learning_rate": 2.466191520467836e-06,
+      "loss": 1.4372,
+      "step": 19189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1702346801757812,
+      "learning_rate": 2.465277777777778e-06,
+      "loss": 2.0593,
+      "step": 19190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.351991891860962,
+      "learning_rate": 2.4643640350877197e-06,
+      "loss": 1.5372,
+      "step": 19191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5218353271484375,
+      "learning_rate": 2.463450292397661e-06,
+      "loss": 1.6822,
+      "step": 19192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.144307851791382,
+      "learning_rate": 2.4625365497076025e-06,
+      "loss": 1.2452,
+      "step": 19193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8316192626953125,
+      "learning_rate": 2.4616228070175443e-06,
+      "loss": 1.4315,
+      "step": 19194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.713252305984497,
+      "learning_rate": 2.4607090643274857e-06,
+      "loss": 1.7028,
+      "step": 19195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6790835857391357,
+      "learning_rate": 2.459795321637427e-06,
+      "loss": 1.6431,
+      "step": 19196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.783907890319824,
+      "learning_rate": 2.4588815789473684e-06,
+      "loss": 1.5259,
+      "step": 19197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4987151622772217,
+      "learning_rate": 2.45796783625731e-06,
+      "loss": 1.6904,
+      "step": 19198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.686305046081543,
+      "learning_rate": 2.4570540935672516e-06,
+      "loss": 1.2809,
+      "step": 19199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.503335952758789,
+      "learning_rate": 2.456140350877193e-06,
+      "loss": 1.3904,
+      "step": 19200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.406087875366211,
+      "learning_rate": 2.4552266081871348e-06,
+      "loss": 1.2679,
+      "step": 19201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3099162578582764,
+      "learning_rate": 2.454312865497076e-06,
+      "loss": 1.3432,
+      "step": 19202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6241273880004883,
+      "learning_rate": 2.453399122807018e-06,
+      "loss": 1.6443,
+      "step": 19203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7663516998291016,
+      "learning_rate": 2.4524853801169593e-06,
+      "loss": 1.6031,
+      "step": 19204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.684419631958008,
+      "learning_rate": 2.4515716374269007e-06,
+      "loss": 1.6988,
+      "step": 19205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7473220825195312,
+      "learning_rate": 2.4506578947368425e-06,
+      "loss": 1.4254,
+      "step": 19206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3781089782714844,
+      "learning_rate": 2.449744152046784e-06,
+      "loss": 1.2257,
+      "step": 19207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.389436721801758,
+      "learning_rate": 2.4488304093567252e-06,
+      "loss": 1.3201,
+      "step": 19208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7195472717285156,
+      "learning_rate": 2.4479166666666666e-06,
+      "loss": 2.0527,
+      "step": 19209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.06156063079834,
+      "learning_rate": 2.4470029239766084e-06,
+      "loss": 1.2356,
+      "step": 19210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.902625322341919,
+      "learning_rate": 2.44608918128655e-06,
+      "loss": 1.2831,
+      "step": 19211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.096243381500244,
+      "learning_rate": 2.4451754385964916e-06,
+      "loss": 1.1243,
+      "step": 19212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1951026916503906,
+      "learning_rate": 2.444261695906433e-06,
+      "loss": 1.2113,
+      "step": 19213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.68144154548645,
+      "learning_rate": 2.4433479532163743e-06,
+      "loss": 1.3922,
+      "step": 19214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.562079429626465,
+      "learning_rate": 2.442434210526316e-06,
+      "loss": 1.6363,
+      "step": 19215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.385972023010254,
+      "learning_rate": 2.4415204678362575e-06,
+      "loss": 1.3608,
+      "step": 19216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.532621383666992,
+      "learning_rate": 2.440606725146199e-06,
+      "loss": 1.3297,
+      "step": 19217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7722935676574707,
+      "learning_rate": 2.4396929824561407e-06,
+      "loss": 1.4653,
+      "step": 19218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.629601001739502,
+      "learning_rate": 2.438779239766082e-06,
+      "loss": 1.3286,
+      "step": 19219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8307344913482666,
+      "learning_rate": 2.4378654970760234e-06,
+      "loss": 1.2705,
+      "step": 19220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.308204174041748,
+      "learning_rate": 2.4369517543859648e-06,
+      "loss": 1.4312,
+      "step": 19221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.01621413230896,
+      "learning_rate": 2.4360380116959066e-06,
+      "loss": 1.6494,
+      "step": 19222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.091294765472412,
+      "learning_rate": 2.4351242690058484e-06,
+      "loss": 1.1874,
+      "step": 19223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3877413272857666,
+      "learning_rate": 2.4342105263157898e-06,
+      "loss": 1.1556,
+      "step": 19224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8770458698272705,
+      "learning_rate": 2.433296783625731e-06,
+      "loss": 1.5308,
+      "step": 19225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6708061695098877,
+      "learning_rate": 2.432383040935673e-06,
+      "loss": 1.6171,
+      "step": 19226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9798229932785034,
+      "learning_rate": 2.4314692982456143e-06,
+      "loss": 1.1366,
+      "step": 19227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.337251663208008,
+      "learning_rate": 2.4305555555555557e-06,
+      "loss": 1.3201,
+      "step": 19228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.390808343887329,
+      "learning_rate": 2.429641812865497e-06,
+      "loss": 1.2803,
+      "step": 19229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3144266605377197,
+      "learning_rate": 2.428728070175439e-06,
+      "loss": 1.2649,
+      "step": 19230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5298726558685303,
+      "learning_rate": 2.4278143274853802e-06,
+      "loss": 1.4502,
+      "step": 19231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.567246675491333,
+      "learning_rate": 2.4269005847953216e-06,
+      "loss": 1.5434,
+      "step": 19232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.748382806777954,
+      "learning_rate": 2.4259868421052634e-06,
+      "loss": 1.9004,
+      "step": 19233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.507030725479126,
+      "learning_rate": 2.4250730994152048e-06,
+      "loss": 1.4858,
+      "step": 19234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.252106189727783,
+      "learning_rate": 2.4241593567251466e-06,
+      "loss": 1.2246,
+      "step": 19235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.512208938598633,
+      "learning_rate": 2.423245614035088e-06,
+      "loss": 1.3169,
+      "step": 19236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.190635919570923,
+      "learning_rate": 2.4223318713450293e-06,
+      "loss": 1.0634,
+      "step": 19237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.105255365371704,
+      "learning_rate": 2.421418128654971e-06,
+      "loss": 1.7093,
+      "step": 19238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.55643630027771,
+      "learning_rate": 2.4205043859649125e-06,
+      "loss": 1.5438,
+      "step": 19239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2746810913085938,
+      "learning_rate": 2.419590643274854e-06,
+      "loss": 1.2406,
+      "step": 19240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2454676628112793,
+      "learning_rate": 2.4186769005847952e-06,
+      "loss": 1.3766,
+      "step": 19241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8635361194610596,
+      "learning_rate": 2.417763157894737e-06,
+      "loss": 1.6209,
+      "step": 19242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1257455348968506,
+      "learning_rate": 2.416849415204679e-06,
+      "loss": 1.2982,
+      "step": 19243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.268211841583252,
+      "learning_rate": 2.41593567251462e-06,
+      "loss": 1.3241,
+      "step": 19244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.451343059539795,
+      "learning_rate": 2.4150219298245616e-06,
+      "loss": 1.5453,
+      "step": 19245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8639073371887207,
+      "learning_rate": 2.4141081871345034e-06,
+      "loss": 1.6609,
+      "step": 19246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.593895435333252,
+      "learning_rate": 2.4131944444444448e-06,
+      "loss": 1.3304,
+      "step": 19247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4344441890716553,
+      "learning_rate": 2.412280701754386e-06,
+      "loss": 1.2429,
+      "step": 19248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7551209926605225,
+      "learning_rate": 2.4113669590643275e-06,
+      "loss": 1.3557,
+      "step": 19249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.709993362426758,
+      "learning_rate": 2.4104532163742693e-06,
+      "loss": 1.4078,
+      "step": 19250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.287543296813965,
+      "learning_rate": 2.4095394736842107e-06,
+      "loss": 1.3066,
+      "step": 19251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.660398006439209,
+      "learning_rate": 2.408625730994152e-06,
+      "loss": 1.7591,
+      "step": 19252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.795783042907715,
+      "learning_rate": 2.407711988304094e-06,
+      "loss": 1.7451,
+      "step": 19253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1000492572784424,
+      "learning_rate": 2.4067982456140352e-06,
+      "loss": 1.2772,
+      "step": 19254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5216822624206543,
+      "learning_rate": 2.405884502923977e-06,
+      "loss": 1.2065,
+      "step": 19255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.355705738067627,
+      "learning_rate": 2.4049707602339184e-06,
+      "loss": 1.6991,
+      "step": 19256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4459376335144043,
+      "learning_rate": 2.4040570175438598e-06,
+      "loss": 1.4903,
+      "step": 19257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5300889015197754,
+      "learning_rate": 2.4031432748538016e-06,
+      "loss": 1.5083,
+      "step": 19258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.445958137512207,
+      "learning_rate": 2.402229532163743e-06,
+      "loss": 1.3971,
+      "step": 19259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.295827865600586,
+      "learning_rate": 2.4013157894736843e-06,
+      "loss": 1.306,
+      "step": 19260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.422706127166748,
+      "learning_rate": 2.4004020467836257e-06,
+      "loss": 1.6788,
+      "step": 19261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.54416823387146,
+      "learning_rate": 2.3994883040935675e-06,
+      "loss": 1.0237,
+      "step": 19262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6232950687408447,
+      "learning_rate": 2.398574561403509e-06,
+      "loss": 1.1261,
+      "step": 19263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2839486598968506,
+      "learning_rate": 2.3976608187134502e-06,
+      "loss": 1.2293,
+      "step": 19264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3517563343048096,
+      "learning_rate": 2.396747076023392e-06,
+      "loss": 1.5755,
+      "step": 19265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8663249015808105,
+      "learning_rate": 2.395833333333334e-06,
+      "loss": 1.7585,
+      "step": 19266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.35371732711792,
+      "learning_rate": 2.394919590643275e-06,
+      "loss": 1.5187,
+      "step": 19267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3582088947296143,
+      "learning_rate": 2.3940058479532166e-06,
+      "loss": 1.3514,
+      "step": 19268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.63517165184021,
+      "learning_rate": 2.393092105263158e-06,
+      "loss": 1.6582,
+      "step": 19269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.265549421310425,
+      "learning_rate": 2.3921783625730998e-06,
+      "loss": 1.4026,
+      "step": 19270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1076650619506836,
+      "learning_rate": 2.391264619883041e-06,
+      "loss": 1.3429,
+      "step": 19271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4040884971618652,
+      "learning_rate": 2.3903508771929825e-06,
+      "loss": 1.3745,
+      "step": 19272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.600053310394287,
+      "learning_rate": 2.389437134502924e-06,
+      "loss": 1.3771,
+      "step": 19273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7098567485809326,
+      "learning_rate": 2.3885233918128657e-06,
+      "loss": 1.5287,
+      "step": 19274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3199713230133057,
+      "learning_rate": 2.3876096491228075e-06,
+      "loss": 1.2673,
+      "step": 19275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.269515037536621,
+      "learning_rate": 2.386695906432749e-06,
+      "loss": 1.2857,
+      "step": 19276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7289724349975586,
+      "learning_rate": 2.3857821637426902e-06,
+      "loss": 1.3727,
+      "step": 19277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3698291778564453,
+      "learning_rate": 2.384868421052632e-06,
+      "loss": 1.3424,
+      "step": 19278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9766788482666016,
+      "learning_rate": 2.3839546783625734e-06,
+      "loss": 1.7676,
+      "step": 19279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.257134437561035,
+      "learning_rate": 2.3830409356725148e-06,
+      "loss": 1.3207,
+      "step": 19280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2187180519104004,
+      "learning_rate": 2.382127192982456e-06,
+      "loss": 1.8951,
+      "step": 19281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8314526081085205,
+      "learning_rate": 2.381213450292398e-06,
+      "loss": 1.2872,
+      "step": 19282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.226747989654541,
+      "learning_rate": 2.3802997076023393e-06,
+      "loss": 1.3518,
+      "step": 19283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.546877145767212,
+      "learning_rate": 2.3793859649122807e-06,
+      "loss": 1.5096,
+      "step": 19284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1166791915893555,
+      "learning_rate": 2.3784722222222225e-06,
+      "loss": 1.8246,
+      "step": 19285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.909409523010254,
+      "learning_rate": 2.377558479532164e-06,
+      "loss": 1.736,
+      "step": 19286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6517210006713867,
+      "learning_rate": 2.3766447368421057e-06,
+      "loss": 1.5714,
+      "step": 19287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9086092710494995,
+      "learning_rate": 2.375730994152047e-06,
+      "loss": 1.1273,
+      "step": 19288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3995916843414307,
+      "learning_rate": 2.3748172514619884e-06,
+      "loss": 1.6303,
+      "step": 19289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.312551259994507,
+      "learning_rate": 2.37390350877193e-06,
+      "loss": 1.2755,
+      "step": 19290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1575381755828857,
+      "learning_rate": 2.3729897660818716e-06,
+      "loss": 1.0164,
+      "step": 19291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3276991844177246,
+      "learning_rate": 2.372076023391813e-06,
+      "loss": 1.4205,
+      "step": 19292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 33.63002014160156,
+      "learning_rate": 2.3711622807017543e-06,
+      "loss": 1.352,
+      "step": 19293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6087803840637207,
+      "learning_rate": 2.370248538011696e-06,
+      "loss": 1.4675,
+      "step": 19294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.268458366394043,
+      "learning_rate": 2.3693347953216375e-06,
+      "loss": 1.3724,
+      "step": 19295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5158169269561768,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 1.4436,
+      "step": 19296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.481834888458252,
+      "learning_rate": 2.3675073099415207e-06,
+      "loss": 1.5889,
+      "step": 19297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4557392597198486,
+      "learning_rate": 2.3665935672514625e-06,
+      "loss": 1.1758,
+      "step": 19298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4994845390319824,
+      "learning_rate": 2.365679824561404e-06,
+      "loss": 1.2472,
+      "step": 19299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.656064033508301,
+      "learning_rate": 2.3647660818713452e-06,
+      "loss": 1.7102,
+      "step": 19300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7221720218658447,
+      "learning_rate": 2.3638523391812866e-06,
+      "loss": 1.4181,
+      "step": 19301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4890520572662354,
+      "learning_rate": 2.3629385964912284e-06,
+      "loss": 1.5939,
+      "step": 19302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.951124429702759,
+      "learning_rate": 2.3620248538011698e-06,
+      "loss": 1.2266,
+      "step": 19303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6894898414611816,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 1.454,
+      "step": 19304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.804333209991455,
+      "learning_rate": 2.360197368421053e-06,
+      "loss": 1.4743,
+      "step": 19305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.903821349143982,
+      "learning_rate": 2.3592836257309943e-06,
+      "loss": 1.2226,
+      "step": 19306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.024556875228882,
+      "learning_rate": 2.358369883040936e-06,
+      "loss": 1.5984,
+      "step": 19307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1659929752349854,
+      "learning_rate": 2.3574561403508775e-06,
+      "loss": 1.6868,
+      "step": 19308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.222414016723633,
+      "learning_rate": 2.356542397660819e-06,
+      "loss": 0.9572,
+      "step": 19309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.399240732192993,
+      "learning_rate": 2.3556286549707607e-06,
+      "loss": 1.3192,
+      "step": 19310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.630828380584717,
+      "learning_rate": 2.354714912280702e-06,
+      "loss": 1.975,
+      "step": 19311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4117166996002197,
+      "learning_rate": 2.3538011695906434e-06,
+      "loss": 1.3504,
+      "step": 19312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4906249046325684,
+      "learning_rate": 2.3528874269005848e-06,
+      "loss": 1.4653,
+      "step": 19313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8183798789978027,
+      "learning_rate": 2.3519736842105266e-06,
+      "loss": 1.4537,
+      "step": 19314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5688223838806152,
+      "learning_rate": 2.351059941520468e-06,
+      "loss": 1.5063,
+      "step": 19315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.312321901321411,
+      "learning_rate": 2.3501461988304093e-06,
+      "loss": 1.4856,
+      "step": 19316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.948988914489746,
+      "learning_rate": 2.349232456140351e-06,
+      "loss": 1.4887,
+      "step": 19317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7850704193115234,
+      "learning_rate": 2.348318713450293e-06,
+      "loss": 1.7573,
+      "step": 19318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2870078086853027,
+      "learning_rate": 2.3474049707602343e-06,
+      "loss": 1.4205,
+      "step": 19319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.317793607711792,
+      "learning_rate": 2.3464912280701757e-06,
+      "loss": 1.2596,
+      "step": 19320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.058363437652588,
+      "learning_rate": 2.345577485380117e-06,
+      "loss": 1.28,
+      "step": 19321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.101938247680664,
+      "learning_rate": 2.344663742690059e-06,
+      "loss": 1.1893,
+      "step": 19322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7268588542938232,
+      "learning_rate": 2.3437500000000002e-06,
+      "loss": 1.6935,
+      "step": 19323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3281915187835693,
+      "learning_rate": 2.3428362573099416e-06,
+      "loss": 1.2581,
+      "step": 19324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4758830070495605,
+      "learning_rate": 2.3419225146198834e-06,
+      "loss": 1.1931,
+      "step": 19325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.349747657775879,
+      "learning_rate": 2.3410087719298248e-06,
+      "loss": 1.3205,
+      "step": 19326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1510496139526367,
+      "learning_rate": 2.340095029239766e-06,
+      "loss": 1.1725,
+      "step": 19327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2667136192321777,
+      "learning_rate": 2.3391812865497075e-06,
+      "loss": 1.5556,
+      "step": 19328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3025150299072266,
+      "learning_rate": 2.3382675438596493e-06,
+      "loss": 1.8294,
+      "step": 19329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1282222270965576,
+      "learning_rate": 2.337353801169591e-06,
+      "loss": 1.0224,
+      "step": 19330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.726856231689453,
+      "learning_rate": 2.3364400584795325e-06,
+      "loss": 1.5485,
+      "step": 19331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.087175369262695,
+      "learning_rate": 2.335526315789474e-06,
+      "loss": 1.6635,
+      "step": 19332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6302313804626465,
+      "learning_rate": 2.3346125730994152e-06,
+      "loss": 1.2486,
+      "step": 19333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5715487003326416,
+      "learning_rate": 2.333698830409357e-06,
+      "loss": 1.364,
+      "step": 19334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4719278812408447,
+      "learning_rate": 2.3327850877192984e-06,
+      "loss": 1.5123,
+      "step": 19335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.567288398742676,
+      "learning_rate": 2.3318713450292398e-06,
+      "loss": 1.5088,
+      "step": 19336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.617363929748535,
+      "learning_rate": 2.3309576023391816e-06,
+      "loss": 1.5737,
+      "step": 19337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7072486877441406,
+      "learning_rate": 2.330043859649123e-06,
+      "loss": 1.3765,
+      "step": 19338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3546559810638428,
+      "learning_rate": 2.3291301169590643e-06,
+      "loss": 1.3548,
+      "step": 19339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.235738515853882,
+      "learning_rate": 2.328216374269006e-06,
+      "loss": 1.1847,
+      "step": 19340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0222392082214355,
+      "learning_rate": 2.3273026315789475e-06,
+      "loss": 1.2768,
+      "step": 19341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2233896255493164,
+      "learning_rate": 2.3263888888888893e-06,
+      "loss": 1.0974,
+      "step": 19342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.500777006149292,
+      "learning_rate": 2.3254751461988307e-06,
+      "loss": 1.4088,
+      "step": 19343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3315987586975098,
+      "learning_rate": 2.324561403508772e-06,
+      "loss": 1.1991,
+      "step": 19344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5683913230895996,
+      "learning_rate": 2.3236476608187134e-06,
+      "loss": 1.3177,
+      "step": 19345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9606235027313232,
+      "learning_rate": 2.3227339181286552e-06,
+      "loss": 1.5681,
+      "step": 19346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6085546016693115,
+      "learning_rate": 2.3218201754385966e-06,
+      "loss": 1.4953,
+      "step": 19347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1618688106536865,
+      "learning_rate": 2.320906432748538e-06,
+      "loss": 1.4299,
+      "step": 19348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2115728855133057,
+      "learning_rate": 2.3199926900584798e-06,
+      "loss": 1.4205,
+      "step": 19349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3051090240478516,
+      "learning_rate": 2.3190789473684216e-06,
+      "loss": 1.3309,
+      "step": 19350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.386366128921509,
+      "learning_rate": 2.318165204678363e-06,
+      "loss": 1.6136,
+      "step": 19351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.701132297515869,
+      "learning_rate": 2.3172514619883043e-06,
+      "loss": 1.4631,
+      "step": 19352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.290709972381592,
+      "learning_rate": 2.3163377192982457e-06,
+      "loss": 1.6457,
+      "step": 19353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3576555252075195,
+      "learning_rate": 2.3154239766081875e-06,
+      "loss": 1.1971,
+      "step": 19354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6263790130615234,
+      "learning_rate": 2.314510233918129e-06,
+      "loss": 1.5122,
+      "step": 19355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8505430221557617,
+      "learning_rate": 2.3135964912280702e-06,
+      "loss": 1.619,
+      "step": 19356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.301420211791992,
+      "learning_rate": 2.312682748538012e-06,
+      "loss": 1.3634,
+      "step": 19357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0401930809020996,
+      "learning_rate": 2.3117690058479534e-06,
+      "loss": 1.4704,
+      "step": 19358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.816425323486328,
+      "learning_rate": 2.3108552631578948e-06,
+      "loss": 1.3264,
+      "step": 19359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.530322551727295,
+      "learning_rate": 2.309941520467836e-06,
+      "loss": 1.2937,
+      "step": 19360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3068065643310547,
+      "learning_rate": 2.309027777777778e-06,
+      "loss": 1.2878,
+      "step": 19361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7154593467712402,
+      "learning_rate": 2.3081140350877197e-06,
+      "loss": 1.8472,
+      "step": 19362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.250978469848633,
+      "learning_rate": 2.307200292397661e-06,
+      "loss": 1.7731,
+      "step": 19363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0617854595184326,
+      "learning_rate": 2.3062865497076025e-06,
+      "loss": 1.1138,
+      "step": 19364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4008264541625977,
+      "learning_rate": 2.305372807017544e-06,
+      "loss": 1.3036,
+      "step": 19365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1260106563568115,
+      "learning_rate": 2.3044590643274857e-06,
+      "loss": 1.325,
+      "step": 19366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.374717950820923,
+      "learning_rate": 2.303545321637427e-06,
+      "loss": 1.2955,
+      "step": 19367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.076859474182129,
+      "learning_rate": 2.3026315789473684e-06,
+      "loss": 1.1652,
+      "step": 19368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.74454665184021,
+      "learning_rate": 2.3017178362573102e-06,
+      "loss": 1.6411,
+      "step": 19369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6991801261901855,
+      "learning_rate": 2.3008040935672516e-06,
+      "loss": 1.4004,
+      "step": 19370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.371760845184326,
+      "learning_rate": 2.299890350877193e-06,
+      "loss": 1.2623,
+      "step": 19371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7982654571533203,
+      "learning_rate": 2.2989766081871348e-06,
+      "loss": 1.6157,
+      "step": 19372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.239502429962158,
+      "learning_rate": 2.298062865497076e-06,
+      "loss": 1.2253,
+      "step": 19373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.035069465637207,
+      "learning_rate": 2.297149122807018e-06,
+      "loss": 1.3611,
+      "step": 19374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.409374713897705,
+      "learning_rate": 2.2962353801169593e-06,
+      "loss": 2.0106,
+      "step": 19375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.812340021133423,
+      "learning_rate": 2.2953216374269007e-06,
+      "loss": 1.8124,
+      "step": 19376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4825081825256348,
+      "learning_rate": 2.2944078947368425e-06,
+      "loss": 1.5628,
+      "step": 19377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1062960624694824,
+      "learning_rate": 2.293494152046784e-06,
+      "loss": 1.6479,
+      "step": 19378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4289872646331787,
+      "learning_rate": 2.2925804093567252e-06,
+      "loss": 1.5078,
+      "step": 19379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7519619464874268,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 1.3472,
+      "step": 19380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.876455545425415,
+      "learning_rate": 2.2907529239766084e-06,
+      "loss": 1.5606,
+      "step": 19381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3422276973724365,
+      "learning_rate": 2.28983918128655e-06,
+      "loss": 1.1815,
+      "step": 19382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0989272594451904,
+      "learning_rate": 2.2889254385964916e-06,
+      "loss": 1.202,
+      "step": 19383
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.117734909057617,
+      "learning_rate": 2.288011695906433e-06,
+      "loss": 1.4735,
+      "step": 19384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2392079830169678,
+      "learning_rate": 2.2870979532163743e-06,
+      "loss": 1.8391,
+      "step": 19385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3975777626037598,
+      "learning_rate": 2.286184210526316e-06,
+      "loss": 1.2091,
+      "step": 19386
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1261720657348633,
+      "learning_rate": 2.2852704678362575e-06,
+      "loss": 1.1882,
+      "step": 19387
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.354222297668457,
+      "learning_rate": 2.284356725146199e-06,
+      "loss": 1.3784,
+      "step": 19388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.540503978729248,
+      "learning_rate": 2.2834429824561407e-06,
+      "loss": 1.3066,
+      "step": 19389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0555992126464844,
+      "learning_rate": 2.282529239766082e-06,
+      "loss": 1.8869,
+      "step": 19390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.300926685333252,
+      "learning_rate": 2.2816154970760234e-06,
+      "loss": 1.3986,
+      "step": 19391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.002401113510132,
+      "learning_rate": 2.280701754385965e-06,
+      "loss": 1.1512,
+      "step": 19392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.83392071723938,
+      "learning_rate": 2.2797880116959066e-06,
+      "loss": 1.922,
+      "step": 19393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1219642162323,
+      "learning_rate": 2.2788742690058484e-06,
+      "loss": 1.287,
+      "step": 19394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3533928394317627,
+      "learning_rate": 2.2779605263157898e-06,
+      "loss": 1.5608,
+      "step": 19395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0959298610687256,
+      "learning_rate": 2.277046783625731e-06,
+      "loss": 1.3857,
+      "step": 19396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5189664363861084,
+      "learning_rate": 2.2761330409356725e-06,
+      "loss": 1.4604,
+      "step": 19397
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.337963104248047,
+      "learning_rate": 2.2752192982456143e-06,
+      "loss": 1.1617,
+      "step": 19398
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4063024520874023,
+      "learning_rate": 2.2743055555555557e-06,
+      "loss": 1.3958,
+      "step": 19399
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.751747131347656,
+      "learning_rate": 2.273391812865497e-06,
+      "loss": 1.4733,
+      "step": 19400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.922412157058716,
+      "learning_rate": 2.272478070175439e-06,
+      "loss": 1.5972,
+      "step": 19401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4577713012695312,
+      "learning_rate": 2.2715643274853802e-06,
+      "loss": 1.3911,
+      "step": 19402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6131091117858887,
+      "learning_rate": 2.2706505847953216e-06,
+      "loss": 1.4018,
+      "step": 19403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.850013494491577,
+      "learning_rate": 2.2697368421052634e-06,
+      "loss": 1.6401,
+      "step": 19404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9240365028381348,
+      "learning_rate": 2.2688230994152048e-06,
+      "loss": 1.8271,
+      "step": 19405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.396360397338867,
+      "learning_rate": 2.2679093567251466e-06,
+      "loss": 1.2299,
+      "step": 19406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.551060438156128,
+      "learning_rate": 2.266995614035088e-06,
+      "loss": 1.433,
+      "step": 19407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6563103199005127,
+      "learning_rate": 2.2660818713450293e-06,
+      "loss": 1.3334,
+      "step": 19408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.628781318664551,
+      "learning_rate": 2.265168128654971e-06,
+      "loss": 1.3203,
+      "step": 19409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.962963342666626,
+      "learning_rate": 2.2642543859649125e-06,
+      "loss": 1.3811,
+      "step": 19410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9459269046783447,
+      "learning_rate": 2.263340643274854e-06,
+      "loss": 1.6357,
+      "step": 19411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3641958236694336,
+      "learning_rate": 2.2624269005847952e-06,
+      "loss": 1.669,
+      "step": 19412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7361152172088623,
+      "learning_rate": 2.261513157894737e-06,
+      "loss": 1.2522,
+      "step": 19413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.514200210571289,
+      "learning_rate": 2.260599415204679e-06,
+      "loss": 1.3979,
+      "step": 19414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9306563138961792,
+      "learning_rate": 2.2596856725146202e-06,
+      "loss": 1.1709,
+      "step": 19415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6202118396759033,
+      "learning_rate": 2.2587719298245616e-06,
+      "loss": 1.652,
+      "step": 19416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.966184139251709,
+      "learning_rate": 2.257858187134503e-06,
+      "loss": 0.989,
+      "step": 19417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.409045696258545,
+      "learning_rate": 2.2569444444444448e-06,
+      "loss": 1.2366,
+      "step": 19418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1501874923706055,
+      "learning_rate": 2.256030701754386e-06,
+      "loss": 1.2381,
+      "step": 19419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2121527194976807,
+      "learning_rate": 2.2551169590643275e-06,
+      "loss": 1.2122,
+      "step": 19420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8292484283447266,
+      "learning_rate": 2.2542032163742693e-06,
+      "loss": 1.363,
+      "step": 19421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9387121200561523,
+      "learning_rate": 2.2532894736842107e-06,
+      "loss": 1.7127,
+      "step": 19422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7308473587036133,
+      "learning_rate": 2.252375730994152e-06,
+      "loss": 1.3576,
+      "step": 19423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5244529247283936,
+      "learning_rate": 2.2514619883040934e-06,
+      "loss": 1.5221,
+      "step": 19424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.454754114151001,
+      "learning_rate": 2.2505482456140352e-06,
+      "loss": 1.3465,
+      "step": 19425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.470980167388916,
+      "learning_rate": 2.249634502923977e-06,
+      "loss": 1.3013,
+      "step": 19426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1571505069732666,
+      "learning_rate": 2.2487207602339184e-06,
+      "loss": 1.3561,
+      "step": 19427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2626705169677734,
+      "learning_rate": 2.2478070175438598e-06,
+      "loss": 1.1638,
+      "step": 19428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1917057037353516,
+      "learning_rate": 2.2468932748538016e-06,
+      "loss": 1.5218,
+      "step": 19429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0199811458587646,
+      "learning_rate": 2.245979532163743e-06,
+      "loss": 1.6934,
+      "step": 19430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.660076379776001,
+      "learning_rate": 2.2450657894736843e-06,
+      "loss": 1.8252,
+      "step": 19431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6909451484680176,
+      "learning_rate": 2.2441520467836257e-06,
+      "loss": 1.582,
+      "step": 19432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.728494167327881,
+      "learning_rate": 2.2432383040935675e-06,
+      "loss": 1.5408,
+      "step": 19433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4177775382995605,
+      "learning_rate": 2.242324561403509e-06,
+      "loss": 1.4953,
+      "step": 19434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2555599212646484,
+      "learning_rate": 2.2414108187134502e-06,
+      "loss": 1.4412,
+      "step": 19435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7267227172851562,
+      "learning_rate": 2.240497076023392e-06,
+      "loss": 1.4388,
+      "step": 19436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5649683475494385,
+      "learning_rate": 2.2395833333333334e-06,
+      "loss": 1.8469,
+      "step": 19437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.63334321975708,
+      "learning_rate": 2.2386695906432752e-06,
+      "loss": 1.4993,
+      "step": 19438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2147061824798584,
+      "learning_rate": 2.2377558479532166e-06,
+      "loss": 0.9911,
+      "step": 19439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6913063526153564,
+      "learning_rate": 2.236842105263158e-06,
+      "loss": 1.4627,
+      "step": 19440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.704569101333618,
+      "learning_rate": 2.2359283625730998e-06,
+      "loss": 1.6397,
+      "step": 19441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1434898376464844,
+      "learning_rate": 2.235014619883041e-06,
+      "loss": 1.2626,
+      "step": 19442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4891252517700195,
+      "learning_rate": 2.2341008771929825e-06,
+      "loss": 1.3631,
+      "step": 19443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3421175479888916,
+      "learning_rate": 2.233187134502924e-06,
+      "loss": 1.1287,
+      "step": 19444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.064763307571411,
+      "learning_rate": 2.2322733918128657e-06,
+      "loss": 1.1786,
+      "step": 19445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7570269107818604,
+      "learning_rate": 2.2313596491228075e-06,
+      "loss": 1.6291,
+      "step": 19446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.058131694793701,
+      "learning_rate": 2.230445906432749e-06,
+      "loss": 1.1218,
+      "step": 19447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3460724353790283,
+      "learning_rate": 2.2295321637426902e-06,
+      "loss": 1.3046,
+      "step": 19448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3430428504943848,
+      "learning_rate": 2.228618421052632e-06,
+      "loss": 1.9358,
+      "step": 19449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5636935234069824,
+      "learning_rate": 2.2277046783625734e-06,
+      "loss": 1.4287,
+      "step": 19450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8444647789001465,
+      "learning_rate": 2.2267909356725148e-06,
+      "loss": 1.5773,
+      "step": 19451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1791446208953857,
+      "learning_rate": 2.225877192982456e-06,
+      "loss": 1.2429,
+      "step": 19452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6283018589019775,
+      "learning_rate": 2.224963450292398e-06,
+      "loss": 1.3997,
+      "step": 19453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6407949924468994,
+      "learning_rate": 2.2240497076023393e-06,
+      "loss": 1.6657,
+      "step": 19454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7701289653778076,
+      "learning_rate": 2.2231359649122807e-06,
+      "loss": 1.7898,
+      "step": 19455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.16166615486145,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.0604,
+      "step": 19456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.364588975906372,
+      "learning_rate": 2.221308479532164e-06,
+      "loss": 1.587,
+      "step": 19457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8278489112854004,
+      "learning_rate": 2.2203947368421057e-06,
+      "loss": 1.6784,
+      "step": 19458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4290733337402344,
+      "learning_rate": 2.219480994152047e-06,
+      "loss": 1.3806,
+      "step": 19459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4490630626678467,
+      "learning_rate": 2.2185672514619884e-06,
+      "loss": 1.6259,
+      "step": 19460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3495798110961914,
+      "learning_rate": 2.2176535087719302e-06,
+      "loss": 1.2825,
+      "step": 19461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8894472122192383,
+      "learning_rate": 2.2167397660818716e-06,
+      "loss": 1.4466,
+      "step": 19462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.713120698928833,
+      "learning_rate": 2.215826023391813e-06,
+      "loss": 1.5659,
+      "step": 19463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.025752067565918,
+      "learning_rate": 2.2149122807017543e-06,
+      "loss": 1.5515,
+      "step": 19464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.685638427734375,
+      "learning_rate": 2.213998538011696e-06,
+      "loss": 1.3283,
+      "step": 19465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.018656015396118,
+      "learning_rate": 2.2130847953216375e-06,
+      "loss": 1.6592,
+      "step": 19466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.726012706756592,
+      "learning_rate": 2.212171052631579e-06,
+      "loss": 1.5513,
+      "step": 19467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.492670774459839,
+      "learning_rate": 2.2112573099415207e-06,
+      "loss": 1.3699,
+      "step": 19468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2154722213745117,
+      "learning_rate": 2.210343567251462e-06,
+      "loss": 1.294,
+      "step": 19469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.35554838180542,
+      "learning_rate": 2.209429824561404e-06,
+      "loss": 1.4232,
+      "step": 19470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7608017921447754,
+      "learning_rate": 2.2085160818713452e-06,
+      "loss": 1.4182,
+      "step": 19471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9175949096679688,
+      "learning_rate": 2.2076023391812866e-06,
+      "loss": 1.6864,
+      "step": 19472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.379006862640381,
+      "learning_rate": 2.2066885964912284e-06,
+      "loss": 1.3878,
+      "step": 19473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5067977905273438,
+      "learning_rate": 2.2057748538011698e-06,
+      "loss": 1.4564,
+      "step": 19474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.011747121810913,
+      "learning_rate": 2.204861111111111e-06,
+      "loss": 1.5176,
+      "step": 19475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9554524421691895,
+      "learning_rate": 2.2039473684210525e-06,
+      "loss": 1.4755,
+      "step": 19476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1895768642425537,
+      "learning_rate": 2.2030336257309943e-06,
+      "loss": 1.2618,
+      "step": 19477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5731871128082275,
+      "learning_rate": 2.202119883040936e-06,
+      "loss": 1.7653,
+      "step": 19478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.409661054611206,
+      "learning_rate": 2.2012061403508775e-06,
+      "loss": 1.4478,
+      "step": 19479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8529210090637207,
+      "learning_rate": 2.200292397660819e-06,
+      "loss": 1.5963,
+      "step": 19480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.843860387802124,
+      "learning_rate": 2.1993786549707607e-06,
+      "loss": 1.6095,
+      "step": 19481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1558878421783447,
+      "learning_rate": 2.198464912280702e-06,
+      "loss": 1.5195,
+      "step": 19482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8286774158477783,
+      "learning_rate": 2.1975511695906434e-06,
+      "loss": 1.8441,
+      "step": 19483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4093894958496094,
+      "learning_rate": 2.196637426900585e-06,
+      "loss": 1.2292,
+      "step": 19484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6267921924591064,
+      "learning_rate": 2.1957236842105266e-06,
+      "loss": 1.5744,
+      "step": 19485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8510725498199463,
+      "learning_rate": 2.194809941520468e-06,
+      "loss": 0.9992,
+      "step": 19486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2888026237487793,
+      "learning_rate": 2.1938961988304093e-06,
+      "loss": 1.3715,
+      "step": 19487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3189585208892822,
+      "learning_rate": 2.192982456140351e-06,
+      "loss": 1.5533,
+      "step": 19488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.42541241645813,
+      "learning_rate": 2.1920687134502925e-06,
+      "loss": 1.181,
+      "step": 19489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.896911144256592,
+      "learning_rate": 2.1911549707602343e-06,
+      "loss": 1.7709,
+      "step": 19490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.508843183517456,
+      "learning_rate": 2.1902412280701757e-06,
+      "loss": 1.5108,
+      "step": 19491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.58380126953125,
+      "learning_rate": 2.189327485380117e-06,
+      "loss": 1.2033,
+      "step": 19492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.368142604827881,
+      "learning_rate": 2.188413742690059e-06,
+      "loss": 1.33,
+      "step": 19493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5630006790161133,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 1.4463,
+      "step": 19494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6313791275024414,
+      "learning_rate": 2.1865862573099416e-06,
+      "loss": 1.7268,
+      "step": 19495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.027994155883789,
+      "learning_rate": 2.185672514619883e-06,
+      "loss": 2.1805,
+      "step": 19496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6255669593811035,
+      "learning_rate": 2.1847587719298248e-06,
+      "loss": 1.4945,
+      "step": 19497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1881301403045654,
+      "learning_rate": 2.183845029239766e-06,
+      "loss": 1.2918,
+      "step": 19498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1430795192718506,
+      "learning_rate": 2.1829312865497075e-06,
+      "loss": 1.4183,
+      "step": 19499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3312816619873047,
+      "learning_rate": 2.1820175438596493e-06,
+      "loss": 1.4082,
+      "step": 19500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9333527088165283,
+      "learning_rate": 2.181103801169591e-06,
+      "loss": 1.1251,
+      "step": 19501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.305218458175659,
+      "learning_rate": 2.1801900584795325e-06,
+      "loss": 1.3216,
+      "step": 19502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5246641635894775,
+      "learning_rate": 2.179276315789474e-06,
+      "loss": 0.9702,
+      "step": 19503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5971438884735107,
+      "learning_rate": 2.1783625730994152e-06,
+      "loss": 1.4497,
+      "step": 19504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.623896837234497,
+      "learning_rate": 2.177448830409357e-06,
+      "loss": 1.7652,
+      "step": 19505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5007033348083496,
+      "learning_rate": 2.1765350877192984e-06,
+      "loss": 1.5766,
+      "step": 19506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3048222064971924,
+      "learning_rate": 2.17562134502924e-06,
+      "loss": 1.455,
+      "step": 19507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8620400428771973,
+      "learning_rate": 2.174707602339181e-06,
+      "loss": 1.5684,
+      "step": 19508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5048511028289795,
+      "learning_rate": 2.173793859649123e-06,
+      "loss": 1.504,
+      "step": 19509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6784934997558594,
+      "learning_rate": 2.1728801169590643e-06,
+      "loss": 1.6236,
+      "step": 19510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.236193895339966,
+      "learning_rate": 2.171966374269006e-06,
+      "loss": 1.2701,
+      "step": 19511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2549283504486084,
+      "learning_rate": 2.1710526315789475e-06,
+      "loss": 1.9291,
+      "step": 19512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3556418418884277,
+      "learning_rate": 2.1701388888888893e-06,
+      "loss": 1.7407,
+      "step": 19513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.289849281311035,
+      "learning_rate": 2.1692251461988307e-06,
+      "loss": 1.452,
+      "step": 19514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8859310150146484,
+      "learning_rate": 2.168311403508772e-06,
+      "loss": 1.627,
+      "step": 19515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6538751125335693,
+      "learning_rate": 2.1673976608187134e-06,
+      "loss": 1.7977,
+      "step": 19516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5906002521514893,
+      "learning_rate": 2.1664839181286552e-06,
+      "loss": 1.4845,
+      "step": 19517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3482391834259033,
+      "learning_rate": 2.1655701754385966e-06,
+      "loss": 1.3073,
+      "step": 19518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.853193521499634,
+      "learning_rate": 2.164656432748538e-06,
+      "loss": 1.2723,
+      "step": 19519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.826411724090576,
+      "learning_rate": 2.1637426900584798e-06,
+      "loss": 1.5797,
+      "step": 19520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.632647752761841,
+      "learning_rate": 2.162828947368421e-06,
+      "loss": 1.4584,
+      "step": 19521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.27634334564209,
+      "learning_rate": 2.161915204678363e-06,
+      "loss": 1.8711,
+      "step": 19522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3587026596069336,
+      "learning_rate": 2.1610014619883043e-06,
+      "loss": 1.1444,
+      "step": 19523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2440052032470703,
+      "learning_rate": 2.1600877192982457e-06,
+      "loss": 1.2991,
+      "step": 19524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8541929721832275,
+      "learning_rate": 2.1591739766081875e-06,
+      "loss": 1.3521,
+      "step": 19525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3733768463134766,
+      "learning_rate": 2.158260233918129e-06,
+      "loss": 1.3416,
+      "step": 19526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3776447772979736,
+      "learning_rate": 2.1573464912280702e-06,
+      "loss": 1.5597,
+      "step": 19527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4090614318847656,
+      "learning_rate": 2.1564327485380116e-06,
+      "loss": 1.1857,
+      "step": 19528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.81168270111084,
+      "learning_rate": 2.1555190058479534e-06,
+      "loss": 1.5945,
+      "step": 19529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4888598918914795,
+      "learning_rate": 2.154605263157895e-06,
+      "loss": 1.2999,
+      "step": 19530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.782383918762207,
+      "learning_rate": 2.153691520467836e-06,
+      "loss": 2.0818,
+      "step": 19531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.495943307876587,
+      "learning_rate": 2.152777777777778e-06,
+      "loss": 1.5993,
+      "step": 19532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2949788570404053,
+      "learning_rate": 2.1518640350877198e-06,
+      "loss": 1.526,
+      "step": 19533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.350158214569092,
+      "learning_rate": 2.150950292397661e-06,
+      "loss": 1.2235,
+      "step": 19534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6508994102478027,
+      "learning_rate": 2.1500365497076025e-06,
+      "loss": 1.6928,
+      "step": 19535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.314020872116089,
+      "learning_rate": 2.149122807017544e-06,
+      "loss": 1.2206,
+      "step": 19536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.58035945892334,
+      "learning_rate": 2.1482090643274857e-06,
+      "loss": 1.5704,
+      "step": 19537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.627951145172119,
+      "learning_rate": 2.147295321637427e-06,
+      "loss": 1.396,
+      "step": 19538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9051156044006348,
+      "learning_rate": 2.1463815789473684e-06,
+      "loss": 1.752,
+      "step": 19539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6012697219848633,
+      "learning_rate": 2.1454678362573102e-06,
+      "loss": 1.5636,
+      "step": 19540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2636003494262695,
+      "learning_rate": 2.1445540935672516e-06,
+      "loss": 1.0755,
+      "step": 19541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5156733989715576,
+      "learning_rate": 2.143640350877193e-06,
+      "loss": 1.2466,
+      "step": 19542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.972160816192627,
+      "learning_rate": 2.1427266081871348e-06,
+      "loss": 1.0325,
+      "step": 19543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6941473484039307,
+      "learning_rate": 2.141812865497076e-06,
+      "loss": 1.5461,
+      "step": 19544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4938864707946777,
+      "learning_rate": 2.140899122807018e-06,
+      "loss": 1.4989,
+      "step": 19545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8546836376190186,
+      "learning_rate": 2.1399853801169593e-06,
+      "loss": 1.3939,
+      "step": 19546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.189338207244873,
+      "learning_rate": 2.1390716374269007e-06,
+      "loss": 1.275,
+      "step": 19547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5948386192321777,
+      "learning_rate": 2.138157894736842e-06,
+      "loss": 1.697,
+      "step": 19548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2284765243530273,
+      "learning_rate": 2.137244152046784e-06,
+      "loss": 1.2696,
+      "step": 19549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.856868028640747,
+      "learning_rate": 2.1363304093567252e-06,
+      "loss": 1.1538,
+      "step": 19550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.13132381439209,
+      "learning_rate": 2.1354166666666666e-06,
+      "loss": 1.3136,
+      "step": 19551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9822800159454346,
+      "learning_rate": 2.1345029239766084e-06,
+      "loss": 1.5427,
+      "step": 19552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.535557985305786,
+      "learning_rate": 2.1335891812865502e-06,
+      "loss": 1.4821,
+      "step": 19553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3718221187591553,
+      "learning_rate": 2.1326754385964916e-06,
+      "loss": 1.6253,
+      "step": 19554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.016085386276245,
+      "learning_rate": 2.131761695906433e-06,
+      "loss": 1.1363,
+      "step": 19555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.205836534500122,
+      "learning_rate": 2.1308479532163743e-06,
+      "loss": 1.8189,
+      "step": 19556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.068142890930176,
+      "learning_rate": 2.129934210526316e-06,
+      "loss": 1.3447,
+      "step": 19557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.459930896759033,
+      "learning_rate": 2.1290204678362575e-06,
+      "loss": 1.5934,
+      "step": 19558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.251843214035034,
+      "learning_rate": 2.128106725146199e-06,
+      "loss": 1.2754,
+      "step": 19559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.586148738861084,
+      "learning_rate": 2.1271929824561407e-06,
+      "loss": 1.5544,
+      "step": 19560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.770094156265259,
+      "learning_rate": 2.126279239766082e-06,
+      "loss": 1.5751,
+      "step": 19561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.610062837600708,
+      "learning_rate": 2.1253654970760234e-06,
+      "loss": 1.701,
+      "step": 19562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.063059091567993,
+      "learning_rate": 2.124451754385965e-06,
+      "loss": 1.2483,
+      "step": 19563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.349576950073242,
+      "learning_rate": 2.1235380116959066e-06,
+      "loss": 1.5999,
+      "step": 19564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4767191410064697,
+      "learning_rate": 2.1226242690058484e-06,
+      "loss": 1.6286,
+      "step": 19565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.772602081298828,
+      "learning_rate": 2.1217105263157898e-06,
+      "loss": 1.7691,
+      "step": 19566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.41801381111145,
+      "learning_rate": 2.120796783625731e-06,
+      "loss": 1.638,
+      "step": 19567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9828195571899414,
+      "learning_rate": 2.1198830409356725e-06,
+      "loss": 1.5627,
+      "step": 19568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.039886236190796,
+      "learning_rate": 2.1189692982456143e-06,
+      "loss": 1.2169,
+      "step": 19569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.545365571975708,
+      "learning_rate": 2.1180555555555557e-06,
+      "loss": 1.4718,
+      "step": 19570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7461093664169312,
+      "learning_rate": 2.117141812865497e-06,
+      "loss": 0.9406,
+      "step": 19571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.733916759490967,
+      "learning_rate": 2.116228070175439e-06,
+      "loss": 1.2282,
+      "step": 19572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8595023155212402,
+      "learning_rate": 2.1153143274853802e-06,
+      "loss": 1.523,
+      "step": 19573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5533220767974854,
+      "learning_rate": 2.1144005847953216e-06,
+      "loss": 1.27,
+      "step": 19574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2949626445770264,
+      "learning_rate": 2.1134868421052634e-06,
+      "loss": 1.3881,
+      "step": 19575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7897634506225586,
+      "learning_rate": 2.112573099415205e-06,
+      "loss": 1.3824,
+      "step": 19576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5167195796966553,
+      "learning_rate": 2.1116593567251466e-06,
+      "loss": 1.4876,
+      "step": 19577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1965177059173584,
+      "learning_rate": 2.110745614035088e-06,
+      "loss": 1.3431,
+      "step": 19578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1691513061523438,
+      "learning_rate": 2.1098318713450293e-06,
+      "loss": 1.2637,
+      "step": 19579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4482316970825195,
+      "learning_rate": 2.1089181286549707e-06,
+      "loss": 1.5218,
+      "step": 19580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.558950185775757,
+      "learning_rate": 2.1080043859649125e-06,
+      "loss": 1.4276,
+      "step": 19581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.510550022125244,
+      "learning_rate": 2.107090643274854e-06,
+      "loss": 1.4879,
+      "step": 19582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.415081262588501,
+      "learning_rate": 2.1061769005847953e-06,
+      "loss": 1.6134,
+      "step": 19583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3105828762054443,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.4207,
+      "step": 19584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.640294075012207,
+      "learning_rate": 2.104349415204679e-06,
+      "loss": 1.5701,
+      "step": 19585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.774797797203064,
+      "learning_rate": 2.1034356725146202e-06,
+      "loss": 0.9406,
+      "step": 19586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.841899871826172,
+      "learning_rate": 2.1025219298245616e-06,
+      "loss": 1.4919,
+      "step": 19587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9555751085281372,
+      "learning_rate": 2.101608187134503e-06,
+      "loss": 1.2908,
+      "step": 19588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.655534505844116,
+      "learning_rate": 2.1006944444444448e-06,
+      "loss": 1.6652,
+      "step": 19589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.889503002166748,
+      "learning_rate": 2.099780701754386e-06,
+      "loss": 1.6144,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9003376960754395,
+      "learning_rate": 2.0988669590643275e-06,
+      "loss": 1.6632,
+      "step": 19591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4973530769348145,
+      "learning_rate": 2.0979532163742693e-06,
+      "loss": 1.5239,
+      "step": 19592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0948116779327393,
+      "learning_rate": 2.0970394736842107e-06,
+      "loss": 2.0219,
+      "step": 19593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5920281410217285,
+      "learning_rate": 2.096125730994152e-06,
+      "loss": 1.5318,
+      "step": 19594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2832231521606445,
+      "learning_rate": 2.0952119883040934e-06,
+      "loss": 1.444,
+      "step": 19595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3446943759918213,
+      "learning_rate": 2.0942982456140352e-06,
+      "loss": 1.5268,
+      "step": 19596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.687619924545288,
+      "learning_rate": 2.093384502923977e-06,
+      "loss": 1.3626,
+      "step": 19597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9109716415405273,
+      "learning_rate": 2.0924707602339184e-06,
+      "loss": 1.0513,
+      "step": 19598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1037514209747314,
+      "learning_rate": 2.09155701754386e-06,
+      "loss": 1.3376,
+      "step": 19599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.144991636276245,
+      "learning_rate": 2.090643274853801e-06,
+      "loss": 1.2645,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.184751033782959,
+      "learning_rate": 2.089729532163743e-06,
+      "loss": 1.3052,
+      "step": 19601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6565399169921875,
+      "learning_rate": 2.0888157894736843e-06,
+      "loss": 1.4754,
+      "step": 19602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1491878032684326,
+      "learning_rate": 2.0879020467836257e-06,
+      "loss": 1.4463,
+      "step": 19603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9470813274383545,
+      "learning_rate": 2.0869883040935675e-06,
+      "loss": 1.8583,
+      "step": 19604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.376384735107422,
+      "learning_rate": 2.086074561403509e-06,
+      "loss": 1.4949,
+      "step": 19605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1944620609283447,
+      "learning_rate": 2.0851608187134503e-06,
+      "loss": 1.2659,
+      "step": 19606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.900512933731079,
+      "learning_rate": 2.084247076023392e-06,
+      "loss": 1.4114,
+      "step": 19607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4280827045440674,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.643,
+      "step": 19608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.541196346282959,
+      "learning_rate": 2.0824195906432752e-06,
+      "loss": 1.5138,
+      "step": 19609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.566774368286133,
+      "learning_rate": 2.0815058479532166e-06,
+      "loss": 1.4414,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.360555648803711,
+      "learning_rate": 2.080592105263158e-06,
+      "loss": 1.0828,
+      "step": 19611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.145322322845459,
+      "learning_rate": 2.0796783625730998e-06,
+      "loss": 1.2323,
+      "step": 19612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6406893730163574,
+      "learning_rate": 2.078764619883041e-06,
+      "loss": 1.5072,
+      "step": 19613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4241907596588135,
+      "learning_rate": 2.0778508771929825e-06,
+      "loss": 1.5307,
+      "step": 19614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6829657554626465,
+      "learning_rate": 2.076937134502924e-06,
+      "loss": 1.5181,
+      "step": 19615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.013201951980591,
+      "learning_rate": 2.0760233918128657e-06,
+      "loss": 1.7031,
+      "step": 19616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.064183235168457,
+      "learning_rate": 2.0751096491228075e-06,
+      "loss": 1.0318,
+      "step": 19617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.285600423812866,
+      "learning_rate": 2.074195906432749e-06,
+      "loss": 1.3887,
+      "step": 19618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.159410238265991,
+      "learning_rate": 2.0732821637426902e-06,
+      "loss": 1.3958,
+      "step": 19619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6272788047790527,
+      "learning_rate": 2.0723684210526316e-06,
+      "loss": 1.4008,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.251132011413574,
+      "learning_rate": 2.0714546783625734e-06,
+      "loss": 1.2826,
+      "step": 19621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.824453353881836,
+      "learning_rate": 2.070540935672515e-06,
+      "loss": 1.791,
+      "step": 19622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2183008193969727,
+      "learning_rate": 2.069627192982456e-06,
+      "loss": 1.3795,
+      "step": 19623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8943629264831543,
+      "learning_rate": 2.068713450292398e-06,
+      "loss": 1.4986,
+      "step": 19624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9181787967681885,
+      "learning_rate": 2.0677997076023393e-06,
+      "loss": 1.6312,
+      "step": 19625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.929994583129883,
+      "learning_rate": 2.0668859649122807e-06,
+      "loss": 1.671,
+      "step": 19626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.690164089202881,
+      "learning_rate": 2.065972222222222e-06,
+      "loss": 1.6163,
+      "step": 19627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.021937370300293,
+      "learning_rate": 2.065058479532164e-06,
+      "loss": 1.6814,
+      "step": 19628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4796218872070312,
+      "learning_rate": 2.0641447368421057e-06,
+      "loss": 1.3714,
+      "step": 19629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2805378437042236,
+      "learning_rate": 2.063230994152047e-06,
+      "loss": 1.2487,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.537484645843506,
+      "learning_rate": 2.0623172514619884e-06,
+      "loss": 1.5794,
+      "step": 19631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.412093162536621,
+      "learning_rate": 2.06140350877193e-06,
+      "loss": 1.7208,
+      "step": 19632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9612948894500732,
+      "learning_rate": 2.0604897660818716e-06,
+      "loss": 1.6343,
+      "step": 19633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5176711082458496,
+      "learning_rate": 2.059576023391813e-06,
+      "loss": 1.3598,
+      "step": 19634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2783377170562744,
+      "learning_rate": 2.0586622807017544e-06,
+      "loss": 1.2023,
+      "step": 19635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.595102071762085,
+      "learning_rate": 2.057748538011696e-06,
+      "loss": 1.6641,
+      "step": 19636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.355794668197632,
+      "learning_rate": 2.0568347953216375e-06,
+      "loss": 1.3879,
+      "step": 19637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8938305377960205,
+      "learning_rate": 2.055921052631579e-06,
+      "loss": 1.6625,
+      "step": 19638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.570189952850342,
+      "learning_rate": 2.0550073099415207e-06,
+      "loss": 1.5409,
+      "step": 19639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3173129558563232,
+      "learning_rate": 2.054093567251462e-06,
+      "loss": 1.3319,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6546051502227783,
+      "learning_rate": 2.053179824561404e-06,
+      "loss": 1.1847,
+      "step": 19641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0354437828063965,
+      "learning_rate": 2.0522660818713452e-06,
+      "loss": 1.7024,
+      "step": 19642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2293050289154053,
+      "learning_rate": 2.0513523391812866e-06,
+      "loss": 1.2115,
+      "step": 19643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1348912715911865,
+      "learning_rate": 2.0504385964912284e-06,
+      "loss": 1.3656,
+      "step": 19644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.802572011947632,
+      "learning_rate": 2.04952485380117e-06,
+      "loss": 1.4834,
+      "step": 19645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.476412296295166,
+      "learning_rate": 2.048611111111111e-06,
+      "loss": 1.2532,
+      "step": 19646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.689347982406616,
+      "learning_rate": 2.0476973684210525e-06,
+      "loss": 1.633,
+      "step": 19647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.131641387939453,
+      "learning_rate": 2.0467836257309943e-06,
+      "loss": 1.325,
+      "step": 19648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0170180797576904,
+      "learning_rate": 2.045869883040936e-06,
+      "loss": 1.8094,
+      "step": 19649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.766749858856201,
+      "learning_rate": 2.0449561403508775e-06,
+      "loss": 1.1544,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6569604873657227,
+      "learning_rate": 2.044042397660819e-06,
+      "loss": 1.4363,
+      "step": 19651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7109570503234863,
+      "learning_rate": 2.0431286549707603e-06,
+      "loss": 1.6651,
+      "step": 19652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5450527667999268,
+      "learning_rate": 2.042214912280702e-06,
+      "loss": 1.5484,
+      "step": 19653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0588958263397217,
+      "learning_rate": 2.0413011695906434e-06,
+      "loss": 1.5982,
+      "step": 19654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.937894344329834,
+      "learning_rate": 2.040387426900585e-06,
+      "loss": 1.6532,
+      "step": 19655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3341963291168213,
+      "learning_rate": 2.0394736842105266e-06,
+      "loss": 1.3786,
+      "step": 19656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.943315267562866,
+      "learning_rate": 2.038559941520468e-06,
+      "loss": 1.5485,
+      "step": 19657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8386881351470947,
+      "learning_rate": 2.0376461988304094e-06,
+      "loss": 1.3566,
+      "step": 19658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.666724443435669,
+      "learning_rate": 2.0367324561403507e-06,
+      "loss": 1.1467,
+      "step": 19659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4725708961486816,
+      "learning_rate": 2.0358187134502925e-06,
+      "loss": 1.2618,
+      "step": 19660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7643752098083496,
+      "learning_rate": 2.0349049707602343e-06,
+      "loss": 1.4151,
+      "step": 19661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6273279190063477,
+      "learning_rate": 2.0339912280701757e-06,
+      "loss": 1.3518,
+      "step": 19662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.589263439178467,
+      "learning_rate": 2.033077485380117e-06,
+      "loss": 1.5549,
+      "step": 19663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7091593742370605,
+      "learning_rate": 2.032163742690059e-06,
+      "loss": 1.5824,
+      "step": 19664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.143185615539551,
+      "learning_rate": 2.0312500000000002e-06,
+      "loss": 1.2956,
+      "step": 19665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2636260986328125,
+      "learning_rate": 2.0303362573099416e-06,
+      "loss": 1.4136,
+      "step": 19666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.750070571899414,
+      "learning_rate": 2.029422514619883e-06,
+      "loss": 1.3248,
+      "step": 19667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4338366985321045,
+      "learning_rate": 2.028508771929825e-06,
+      "loss": 1.4328,
+      "step": 19668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3347458839416504,
+      "learning_rate": 2.027595029239766e-06,
+      "loss": 1.4556,
+      "step": 19669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.02854061126709,
+      "learning_rate": 2.0266812865497075e-06,
+      "loss": 1.021,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6239492893218994,
+      "learning_rate": 2.0257675438596493e-06,
+      "loss": 1.4838,
+      "step": 19671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6275441646575928,
+      "learning_rate": 2.0248538011695907e-06,
+      "loss": 1.72,
+      "step": 19672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9616447687149048,
+      "learning_rate": 2.0239400584795325e-06,
+      "loss": 1.2563,
+      "step": 19673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.551194429397583,
+      "learning_rate": 2.023026315789474e-06,
+      "loss": 1.4436,
+      "step": 19674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.873513698577881,
+      "learning_rate": 2.0221125730994153e-06,
+      "loss": 1.8154,
+      "step": 19675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.760085344314575,
+      "learning_rate": 2.021198830409357e-06,
+      "loss": 1.7185,
+      "step": 19676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.487222909927368,
+      "learning_rate": 2.0202850877192984e-06,
+      "loss": 1.1175,
+      "step": 19677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4885263442993164,
+      "learning_rate": 2.01937134502924e-06,
+      "loss": 1.4178,
+      "step": 19678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4808735847473145,
+      "learning_rate": 2.018457602339181e-06,
+      "loss": 1.6425,
+      "step": 19679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.151946783065796,
+      "learning_rate": 2.017543859649123e-06,
+      "loss": 1.2457,
+      "step": 19680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.31418514251709,
+      "learning_rate": 2.0166301169590644e-06,
+      "loss": 1.4004,
+      "step": 19681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.854163646697998,
+      "learning_rate": 2.015716374269006e-06,
+      "loss": 1.4455,
+      "step": 19682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6914737224578857,
+      "learning_rate": 2.0148026315789475e-06,
+      "loss": 1.3012,
+      "step": 19683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.292489767074585,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 1.373,
+      "step": 19684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.148024559020996,
+      "learning_rate": 2.0129751461988307e-06,
+      "loss": 1.8489,
+      "step": 19685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7055559158325195,
+      "learning_rate": 2.012061403508772e-06,
+      "loss": 1.612,
+      "step": 19686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5368905067443848,
+      "learning_rate": 2.0111476608187134e-06,
+      "loss": 1.1338,
+      "step": 19687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5958027839660645,
+      "learning_rate": 2.0102339181286552e-06,
+      "loss": 1.6803,
+      "step": 19688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.366745948791504,
+      "learning_rate": 2.0093201754385966e-06,
+      "loss": 1.6545,
+      "step": 19689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9147119522094727,
+      "learning_rate": 2.008406432748538e-06,
+      "loss": 1.0773,
+      "step": 19690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4438862800598145,
+      "learning_rate": 2.0074926900584794e-06,
+      "loss": 1.2119,
+      "step": 19691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4698164463043213,
+      "learning_rate": 2.006578947368421e-06,
+      "loss": 1.2377,
+      "step": 19692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2678890228271484,
+      "learning_rate": 2.005665204678363e-06,
+      "loss": 1.3927,
+      "step": 19693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2961618900299072,
+      "learning_rate": 2.0047514619883043e-06,
+      "loss": 1.3851,
+      "step": 19694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4521055221557617,
+      "learning_rate": 2.0038377192982457e-06,
+      "loss": 1.4108,
+      "step": 19695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.49127459526062,
+      "learning_rate": 2.0029239766081875e-06,
+      "loss": 1.3624,
+      "step": 19696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6882503032684326,
+      "learning_rate": 2.002010233918129e-06,
+      "loss": 1.7459,
+      "step": 19697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.279414653778076,
+      "learning_rate": 2.0010964912280703e-06,
+      "loss": 1.3128,
+      "step": 19698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2558956146240234,
+      "learning_rate": 2.0001827485380116e-06,
+      "loss": 1.6307,
+      "step": 19699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0332624912261963,
+      "learning_rate": 1.9992690058479534e-06,
+      "loss": 1.5648,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.020590305328369,
+      "learning_rate": 1.998355263157895e-06,
+      "loss": 1.8898,
+      "step": 19701
+    },
+    {
+      "epoch": 0.9,
+      "eval_train_bleu": 0.2295048586870517,
+      "eval_train_loss": 2.1054584980010986,
+      "eval_train_runtime": 24.3116,
+      "eval_train_samples_per_second": 41.174,
+      "eval_train_steps_per_second": 1.316,
+      "step": 19701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7684948444366455,
+      "learning_rate": 1.997441520467836e-06,
+      "loss": 2.0133,
+      "step": 19702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3817532062530518,
+      "learning_rate": 1.996527777777778e-06,
+      "loss": 1.2596,
+      "step": 19703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.695242166519165,
+      "learning_rate": 1.9956140350877194e-06,
+      "loss": 1.5147,
+      "step": 19704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3552322387695312,
+      "learning_rate": 1.994700292397661e-06,
+      "loss": 1.2825,
+      "step": 19705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0533201694488525,
+      "learning_rate": 1.9937865497076025e-06,
+      "loss": 1.1372,
+      "step": 19706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1312544345855713,
+      "learning_rate": 1.992872807017544e-06,
+      "loss": 1.2343,
+      "step": 19707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5614616870880127,
+      "learning_rate": 1.9919590643274857e-06,
+      "loss": 1.3387,
+      "step": 19708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.625239133834839,
+      "learning_rate": 1.991045321637427e-06,
+      "loss": 1.4746,
+      "step": 19709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4722790718078613,
+      "learning_rate": 1.9901315789473684e-06,
+      "loss": 1.4081,
+      "step": 19710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7046444416046143,
+      "learning_rate": 1.98921783625731e-06,
+      "loss": 1.4785,
+      "step": 19711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.15205717086792,
+      "learning_rate": 1.9883040935672516e-06,
+      "loss": 1.2637,
+      "step": 19712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3726468086242676,
+      "learning_rate": 1.987390350877193e-06,
+      "loss": 1.5239,
+      "step": 19713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.470994710922241,
+      "learning_rate": 1.986476608187135e-06,
+      "loss": 1.5589,
+      "step": 19714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7364240884780884,
+      "learning_rate": 1.985562865497076e-06,
+      "loss": 1.1321,
+      "step": 19715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9499151706695557,
+      "learning_rate": 1.984649122807018e-06,
+      "loss": 1.6233,
+      "step": 19716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9124869108200073,
+      "learning_rate": 1.9837353801169593e-06,
+      "loss": 1.2717,
+      "step": 19717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6681041717529297,
+      "learning_rate": 1.9828216374269007e-06,
+      "loss": 1.4461,
+      "step": 19718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.945556402206421,
+      "learning_rate": 1.981907894736842e-06,
+      "loss": 1.6337,
+      "step": 19719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.630831718444824,
+      "learning_rate": 1.980994152046784e-06,
+      "loss": 1.6699,
+      "step": 19720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2537920475006104,
+      "learning_rate": 1.9800804093567253e-06,
+      "loss": 1.2175,
+      "step": 19721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.662036418914795,
+      "learning_rate": 1.9791666666666666e-06,
+      "loss": 1.7554,
+      "step": 19722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5374581813812256,
+      "learning_rate": 1.9782529239766084e-06,
+      "loss": 1.5464,
+      "step": 19723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5551021099090576,
+      "learning_rate": 1.97733918128655e-06,
+      "loss": 1.656,
+      "step": 19724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9894866943359375,
+      "learning_rate": 1.9764254385964916e-06,
+      "loss": 1.2567,
+      "step": 19725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5118017196655273,
+      "learning_rate": 1.975511695906433e-06,
+      "loss": 1.5497,
+      "step": 19726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.596205949783325,
+      "learning_rate": 1.9745979532163744e-06,
+      "loss": 1.6323,
+      "step": 19727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9452366828918457,
+      "learning_rate": 1.973684210526316e-06,
+      "loss": 1.0116,
+      "step": 19728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.970435380935669,
+      "learning_rate": 1.9727704678362575e-06,
+      "loss": 1.6785,
+      "step": 19729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.310892343521118,
+      "learning_rate": 1.971856725146199e-06,
+      "loss": 1.2397,
+      "step": 19730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2067346572875977,
+      "learning_rate": 1.9709429824561403e-06,
+      "loss": 1.4088,
+      "step": 19731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9113218784332275,
+      "learning_rate": 1.970029239766082e-06,
+      "loss": 1.3067,
+      "step": 19732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.401456117630005,
+      "learning_rate": 1.9691154970760234e-06,
+      "loss": 1.2166,
+      "step": 19733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3407435417175293,
+      "learning_rate": 1.968201754385965e-06,
+      "loss": 1.2352,
+      "step": 19734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.104538917541504,
+      "learning_rate": 1.9672880116959066e-06,
+      "loss": 1.7082,
+      "step": 19735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3452930450439453,
+      "learning_rate": 1.9663742690058484e-06,
+      "loss": 1.4435,
+      "step": 19736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7950406074523926,
+      "learning_rate": 1.96546052631579e-06,
+      "loss": 1.6439,
+      "step": 19737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.496870279312134,
+      "learning_rate": 1.964546783625731e-06,
+      "loss": 1.3316,
+      "step": 19738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.347921848297119,
+      "learning_rate": 1.9636330409356725e-06,
+      "loss": 1.6473,
+      "step": 19739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9760184288024902,
+      "learning_rate": 1.9627192982456143e-06,
+      "loss": 1.7196,
+      "step": 19740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.444873332977295,
+      "learning_rate": 1.9618055555555557e-06,
+      "loss": 1.5642,
+      "step": 19741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3336312770843506,
+      "learning_rate": 1.960891812865497e-06,
+      "loss": 1.4498,
+      "step": 19742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.386425256729126,
+      "learning_rate": 1.959978070175439e-06,
+      "loss": 1.3046,
+      "step": 19743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3203556537628174,
+      "learning_rate": 1.9590643274853803e-06,
+      "loss": 1.42,
+      "step": 19744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.203477144241333,
+      "learning_rate": 1.9581505847953216e-06,
+      "loss": 1.2577,
+      "step": 19745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.398682117462158,
+      "learning_rate": 1.9572368421052634e-06,
+      "loss": 1.4158,
+      "step": 19746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.842134475708008,
+      "learning_rate": 1.956323099415205e-06,
+      "loss": 1.704,
+      "step": 19747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.457582950592041,
+      "learning_rate": 1.9554093567251466e-06,
+      "loss": 1.3118,
+      "step": 19748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4521195888519287,
+      "learning_rate": 1.954495614035088e-06,
+      "loss": 1.3074,
+      "step": 19749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7509613037109375,
+      "learning_rate": 1.9535818713450294e-06,
+      "loss": 1.7076,
+      "step": 19750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4706242084503174,
+      "learning_rate": 1.9526681286549707e-06,
+      "loss": 1.4468,
+      "step": 19751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5685787200927734,
+      "learning_rate": 1.9517543859649125e-06,
+      "loss": 1.4807,
+      "step": 19752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1745383739471436,
+      "learning_rate": 1.950840643274854e-06,
+      "loss": 1.4671,
+      "step": 19753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7006192207336426,
+      "learning_rate": 1.9499269005847953e-06,
+      "loss": 1.7243,
+      "step": 19754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0000288486480713,
+      "learning_rate": 1.949013157894737e-06,
+      "loss": 1.4194,
+      "step": 19755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.037611722946167,
+      "learning_rate": 1.9480994152046784e-06,
+      "loss": 1.7874,
+      "step": 19756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5424723625183105,
+      "learning_rate": 1.9471856725146202e-06,
+      "loss": 1.2005,
+      "step": 19757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.051811456680298,
+      "learning_rate": 1.9462719298245616e-06,
+      "loss": 1.1737,
+      "step": 19758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2312538623809814,
+      "learning_rate": 1.945358187134503e-06,
+      "loss": 1.3823,
+      "step": 19759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9728124141693115,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 1.1055,
+      "step": 19760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2345967292785645,
+      "learning_rate": 1.943530701754386e-06,
+      "loss": 1.3879,
+      "step": 19761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4050002098083496,
+      "learning_rate": 1.9426169590643275e-06,
+      "loss": 1.4875,
+      "step": 19762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.958263874053955,
+      "learning_rate": 1.941703216374269e-06,
+      "loss": 1.9754,
+      "step": 19763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.238185405731201,
+      "learning_rate": 1.9407894736842107e-06,
+      "loss": 1.2929,
+      "step": 19764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8749983310699463,
+      "learning_rate": 1.939875730994152e-06,
+      "loss": 1.5795,
+      "step": 19765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.700273036956787,
+      "learning_rate": 1.9389619883040935e-06,
+      "loss": 1.4931,
+      "step": 19766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7573091983795166,
+      "learning_rate": 1.9380482456140353e-06,
+      "loss": 1.5433,
+      "step": 19767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.176098108291626,
+      "learning_rate": 1.937134502923977e-06,
+      "loss": 1.1495,
+      "step": 19768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.90468430519104,
+      "learning_rate": 1.9362207602339184e-06,
+      "loss": 1.8727,
+      "step": 19769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.714939832687378,
+      "learning_rate": 1.93530701754386e-06,
+      "loss": 1.3693,
+      "step": 19770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.147523880004883,
+      "learning_rate": 1.934393274853801e-06,
+      "loss": 1.7068,
+      "step": 19771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3474040031433105,
+      "learning_rate": 1.933479532163743e-06,
+      "loss": 1.4699,
+      "step": 19772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8419547080993652,
+      "learning_rate": 1.9325657894736844e-06,
+      "loss": 1.5912,
+      "step": 19773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2405307292938232,
+      "learning_rate": 1.9316520467836257e-06,
+      "loss": 1.3919,
+      "step": 19774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4497323036193848,
+      "learning_rate": 1.9307383040935675e-06,
+      "loss": 1.5901,
+      "step": 19775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.831979274749756,
+      "learning_rate": 1.929824561403509e-06,
+      "loss": 1.5624,
+      "step": 19776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.412052869796753,
+      "learning_rate": 1.9289108187134503e-06,
+      "loss": 1.4665,
+      "step": 19777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.556252956390381,
+      "learning_rate": 1.927997076023392e-06,
+      "loss": 1.3425,
+      "step": 19778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4684178829193115,
+      "learning_rate": 1.9270833333333334e-06,
+      "loss": 1.3994,
+      "step": 19779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9550203084945679,
+      "learning_rate": 1.9261695906432752e-06,
+      "loss": 1.177,
+      "step": 19780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6761860847473145,
+      "learning_rate": 1.9252558479532166e-06,
+      "loss": 1.4418,
+      "step": 19781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4669742584228516,
+      "learning_rate": 1.924342105263158e-06,
+      "loss": 1.5546,
+      "step": 19782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3999674320220947,
+      "learning_rate": 1.9234283625730994e-06,
+      "loss": 1.2673,
+      "step": 19783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.196977138519287,
+      "learning_rate": 1.922514619883041e-06,
+      "loss": 1.1204,
+      "step": 19784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5855114459991455,
+      "learning_rate": 1.9216008771929825e-06,
+      "loss": 1.3538,
+      "step": 19785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8898544311523438,
+      "learning_rate": 1.920687134502924e-06,
+      "loss": 1.3919,
+      "step": 19786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4250335693359375,
+      "learning_rate": 1.9197733918128657e-06,
+      "loss": 1.3974,
+      "step": 19787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.429539680480957,
+      "learning_rate": 1.9188596491228075e-06,
+      "loss": 1.5801,
+      "step": 19788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6137349605560303,
+      "learning_rate": 1.917945906432749e-06,
+      "loss": 1.4318,
+      "step": 19789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.851133108139038,
+      "learning_rate": 1.9170321637426903e-06,
+      "loss": 1.5665,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5980072021484375,
+      "learning_rate": 1.9161184210526316e-06,
+      "loss": 1.5527,
+      "step": 19791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.324298858642578,
+      "learning_rate": 1.9152046783625734e-06,
+      "loss": 1.4984,
+      "step": 19792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.656482219696045,
+      "learning_rate": 1.914290935672515e-06,
+      "loss": 1.6784,
+      "step": 19793
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.54714298248291,
+      "learning_rate": 1.913377192982456e-06,
+      "loss": 1.6114,
+      "step": 19794
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6683032512664795,
+      "learning_rate": 1.912463450292398e-06,
+      "loss": 1.6509,
+      "step": 19795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.506504774093628,
+      "learning_rate": 1.9115497076023394e-06,
+      "loss": 1.1662,
+      "step": 19796
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.460585594177246,
+      "learning_rate": 1.9106359649122807e-06,
+      "loss": 1.3546,
+      "step": 19797
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9652044773101807,
+      "learning_rate": 1.909722222222222e-06,
+      "loss": 1.8021,
+      "step": 19798
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0200400352478027,
+      "learning_rate": 1.908808479532164e-06,
+      "loss": 1.3289,
+      "step": 19799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.465318202972412,
+      "learning_rate": 1.9078947368421057e-06,
+      "loss": 1.8364,
+      "step": 19800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1800575256347656,
+      "learning_rate": 1.906980994152047e-06,
+      "loss": 1.508,
+      "step": 19801
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.630431890487671,
+      "learning_rate": 1.9060672514619884e-06,
+      "loss": 1.123,
+      "step": 19802
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.845118522644043,
+      "learning_rate": 1.90515350877193e-06,
+      "loss": 1.7184,
+      "step": 19803
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3728525638580322,
+      "learning_rate": 1.9042397660818714e-06,
+      "loss": 1.4043,
+      "step": 19804
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0645666122436523,
+      "learning_rate": 1.903326023391813e-06,
+      "loss": 1.6945,
+      "step": 19805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4295599460601807,
+      "learning_rate": 1.9024122807017546e-06,
+      "loss": 1.6134,
+      "step": 19806
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.269641399383545,
+      "learning_rate": 1.901498538011696e-06,
+      "loss": 1.2906,
+      "step": 19807
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.725247859954834,
+      "learning_rate": 1.9005847953216375e-06,
+      "loss": 1.6717,
+      "step": 19808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2929234504699707,
+      "learning_rate": 1.899671052631579e-06,
+      "loss": 1.0851,
+      "step": 19809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8413524627685547,
+      "learning_rate": 1.8987573099415207e-06,
+      "loss": 1.7138,
+      "step": 19810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.507087469100952,
+      "learning_rate": 1.8978435672514623e-06,
+      "loss": 1.4821,
+      "step": 19811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9813919067382812,
+      "learning_rate": 1.8969298245614037e-06,
+      "loss": 1.9841,
+      "step": 19812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5114667415618896,
+      "learning_rate": 1.8960160818713453e-06,
+      "loss": 1.372,
+      "step": 19813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.387087821960449,
+      "learning_rate": 1.8951023391812866e-06,
+      "loss": 1.5087,
+      "step": 19814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2645926475524902,
+      "learning_rate": 1.8941885964912282e-06,
+      "loss": 1.4004,
+      "step": 19815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.719712257385254,
+      "learning_rate": 1.8932748538011698e-06,
+      "loss": 1.3345,
+      "step": 19816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.507702112197876,
+      "learning_rate": 1.8923611111111112e-06,
+      "loss": 1.3894,
+      "step": 19817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.468045234680176,
+      "learning_rate": 1.8914473684210528e-06,
+      "loss": 1.321,
+      "step": 19818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.528186321258545,
+      "learning_rate": 1.8905336257309941e-06,
+      "loss": 1.5005,
+      "step": 19819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5722270011901855,
+      "learning_rate": 1.889619883040936e-06,
+      "loss": 1.3812,
+      "step": 19820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6635313034057617,
+      "learning_rate": 1.8887061403508775e-06,
+      "loss": 1.6767,
+      "step": 19821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2541956901550293,
+      "learning_rate": 1.887792397660819e-06,
+      "loss": 1.2365,
+      "step": 19822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9540261030197144,
+      "learning_rate": 1.8868786549707605e-06,
+      "loss": 1.1084,
+      "step": 19823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7055039405822754,
+      "learning_rate": 1.8859649122807019e-06,
+      "loss": 1.3048,
+      "step": 19824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.156034231185913,
+      "learning_rate": 1.8850511695906434e-06,
+      "loss": 1.4303,
+      "step": 19825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.508730888366699,
+      "learning_rate": 1.884137426900585e-06,
+      "loss": 1.2418,
+      "step": 19826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1240227222442627,
+      "learning_rate": 1.8832236842105264e-06,
+      "loss": 1.3225,
+      "step": 19827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5679492950439453,
+      "learning_rate": 1.882309941520468e-06,
+      "loss": 1.5725,
+      "step": 19828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6634066104888916,
+      "learning_rate": 1.8813961988304094e-06,
+      "loss": 1.7691,
+      "step": 19829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0608184337615967,
+      "learning_rate": 1.880482456140351e-06,
+      "loss": 1.8654,
+      "step": 19830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.373924732208252,
+      "learning_rate": 1.8795687134502928e-06,
+      "loss": 1.2041,
+      "step": 19831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.207869052886963,
+      "learning_rate": 1.8786549707602341e-06,
+      "loss": 1.2715,
+      "step": 19832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7258224487304688,
+      "learning_rate": 1.8777412280701757e-06,
+      "loss": 1.426,
+      "step": 19833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1434922218322754,
+      "learning_rate": 1.876827485380117e-06,
+      "loss": 1.4488,
+      "step": 19834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.533053398132324,
+      "learning_rate": 1.8759137426900587e-06,
+      "loss": 1.5956,
+      "step": 19835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3473355770111084,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 1.2006,
+      "step": 19836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.108031988143921,
+      "learning_rate": 1.8740862573099416e-06,
+      "loss": 1.3412,
+      "step": 19837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.193371534347534,
+      "learning_rate": 1.8731725146198832e-06,
+      "loss": 1.052,
+      "step": 19838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9614193439483643,
+      "learning_rate": 1.8722587719298246e-06,
+      "loss": 1.4668,
+      "step": 19839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7076497077941895,
+      "learning_rate": 1.8713450292397662e-06,
+      "loss": 1.4977,
+      "step": 19840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5009217262268066,
+      "learning_rate": 1.8704312865497076e-06,
+      "loss": 1.4785,
+      "step": 19841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7192959785461426,
+      "learning_rate": 1.8695175438596494e-06,
+      "loss": 1.8806,
+      "step": 19842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.284594774246216,
+      "learning_rate": 1.868603801169591e-06,
+      "loss": 1.3792,
+      "step": 19843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8899121284484863,
+      "learning_rate": 1.8676900584795323e-06,
+      "loss": 1.6602,
+      "step": 19844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.988239049911499,
+      "learning_rate": 1.866776315789474e-06,
+      "loss": 1.7619,
+      "step": 19845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4039220809936523,
+      "learning_rate": 1.8658625730994153e-06,
+      "loss": 1.4113,
+      "step": 19846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.707601308822632,
+      "learning_rate": 1.8649488304093569e-06,
+      "loss": 1.931,
+      "step": 19847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5062599182128906,
+      "learning_rate": 1.8640350877192984e-06,
+      "loss": 1.3882,
+      "step": 19848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.205873489379883,
+      "learning_rate": 1.8631213450292398e-06,
+      "loss": 1.2669,
+      "step": 19849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3024044036865234,
+      "learning_rate": 1.8622076023391814e-06,
+      "loss": 1.31,
+      "step": 19850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5439071655273438,
+      "learning_rate": 1.8612938596491228e-06,
+      "loss": 1.2664,
+      "step": 19851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0132384300231934,
+      "learning_rate": 1.8603801169590644e-06,
+      "loss": 1.118,
+      "step": 19852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.065373659133911,
+      "learning_rate": 1.8594663742690062e-06,
+      "loss": 1.2233,
+      "step": 19853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.427056312561035,
+      "learning_rate": 1.8585526315789475e-06,
+      "loss": 1.4175,
+      "step": 19854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3839187622070312,
+      "learning_rate": 1.8576388888888891e-06,
+      "loss": 1.2546,
+      "step": 19855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5612967014312744,
+      "learning_rate": 1.8567251461988305e-06,
+      "loss": 1.7684,
+      "step": 19856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.426569938659668,
+      "learning_rate": 1.855811403508772e-06,
+      "loss": 1.8768,
+      "step": 19857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.558216094970703,
+      "learning_rate": 1.8548976608187137e-06,
+      "loss": 1.6476,
+      "step": 19858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3729074001312256,
+      "learning_rate": 1.853983918128655e-06,
+      "loss": 1.3116,
+      "step": 19859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6077213287353516,
+      "learning_rate": 1.8530701754385966e-06,
+      "loss": 1.7063,
+      "step": 19860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7026596069335938,
+      "learning_rate": 1.852156432748538e-06,
+      "loss": 1.7199,
+      "step": 19861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.016186237335205,
+      "learning_rate": 1.8512426900584796e-06,
+      "loss": 1.486,
+      "step": 19862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2470927238464355,
+      "learning_rate": 1.8503289473684214e-06,
+      "loss": 1.2499,
+      "step": 19863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6471147537231445,
+      "learning_rate": 1.8494152046783628e-06,
+      "loss": 1.4881,
+      "step": 19864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0395302772521973,
+      "learning_rate": 1.8485014619883044e-06,
+      "loss": 1.6856,
+      "step": 19865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.246893882751465,
+      "learning_rate": 1.8475877192982457e-06,
+      "loss": 1.5352,
+      "step": 19866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.579096794128418,
+      "learning_rate": 1.8466739766081873e-06,
+      "loss": 1.7747,
+      "step": 19867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.346615791320801,
+      "learning_rate": 1.845760233918129e-06,
+      "loss": 1.4072,
+      "step": 19868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.25903058052063,
+      "learning_rate": 1.8448464912280703e-06,
+      "loss": 1.5151,
+      "step": 19869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5698208808898926,
+      "learning_rate": 1.8439327485380119e-06,
+      "loss": 1.6081,
+      "step": 19870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.643810749053955,
+      "learning_rate": 1.8430190058479532e-06,
+      "loss": 1.6365,
+      "step": 19871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.572936534881592,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 1.6099,
+      "step": 19872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5807695388793945,
+      "learning_rate": 1.8411915204678362e-06,
+      "loss": 1.6116,
+      "step": 19873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.351318120956421,
+      "learning_rate": 1.840277777777778e-06,
+      "loss": 1.4229,
+      "step": 19874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.655529260635376,
+      "learning_rate": 1.8393640350877196e-06,
+      "loss": 1.738,
+      "step": 19875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.512619972229004,
+      "learning_rate": 1.838450292397661e-06,
+      "loss": 1.5589,
+      "step": 19876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.798971176147461,
+      "learning_rate": 1.8375365497076025e-06,
+      "loss": 1.5337,
+      "step": 19877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4920525550842285,
+      "learning_rate": 1.8366228070175441e-06,
+      "loss": 1.311,
+      "step": 19878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.815650224685669,
+      "learning_rate": 1.8357090643274855e-06,
+      "loss": 1.7933,
+      "step": 19879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3843092918395996,
+      "learning_rate": 1.834795321637427e-06,
+      "loss": 1.4137,
+      "step": 19880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4075403213500977,
+      "learning_rate": 1.8338815789473685e-06,
+      "loss": 1.5008,
+      "step": 19881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.111973285675049,
+      "learning_rate": 1.83296783625731e-06,
+      "loss": 1.3084,
+      "step": 19882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6917121410369873,
+      "learning_rate": 1.8320540935672514e-06,
+      "loss": 1.2515,
+      "step": 19883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9283599853515625,
+      "learning_rate": 1.831140350877193e-06,
+      "loss": 1.7061,
+      "step": 19884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7996139526367188,
+      "learning_rate": 1.8302266081871348e-06,
+      "loss": 1.8048,
+      "step": 19885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4482626914978027,
+      "learning_rate": 1.8293128654970762e-06,
+      "loss": 1.4483,
+      "step": 19886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3245790004730225,
+      "learning_rate": 1.8283991228070178e-06,
+      "loss": 1.537,
+      "step": 19887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0961554050445557,
+      "learning_rate": 1.8274853801169594e-06,
+      "loss": 1.4102,
+      "step": 19888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.822295665740967,
+      "learning_rate": 1.8265716374269007e-06,
+      "loss": 1.6652,
+      "step": 19889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.381643056869507,
+      "learning_rate": 1.8256578947368423e-06,
+      "loss": 1.2345,
+      "step": 19890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.326101779937744,
+      "learning_rate": 1.8247441520467837e-06,
+      "loss": 1.409,
+      "step": 19891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.356727123260498,
+      "learning_rate": 1.8238304093567253e-06,
+      "loss": 1.3683,
+      "step": 19892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.434614658355713,
+      "learning_rate": 1.8229166666666666e-06,
+      "loss": 1.329,
+      "step": 19893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8392670154571533,
+      "learning_rate": 1.8220029239766082e-06,
+      "loss": 1.5693,
+      "step": 19894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7774875164031982,
+      "learning_rate": 1.82108918128655e-06,
+      "loss": 1.7457,
+      "step": 19895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.607984781265259,
+      "learning_rate": 1.8201754385964914e-06,
+      "loss": 1.5061,
+      "step": 19896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7690839767456055,
+      "learning_rate": 1.819261695906433e-06,
+      "loss": 1.6865,
+      "step": 19897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.392378330230713,
+      "learning_rate": 1.8183479532163744e-06,
+      "loss": 1.506,
+      "step": 19898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.383460521697998,
+      "learning_rate": 1.817434210526316e-06,
+      "loss": 1.1626,
+      "step": 19899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.079751968383789,
+      "learning_rate": 1.8165204678362575e-06,
+      "loss": 1.9365,
+      "step": 19900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.072664976119995,
+      "learning_rate": 1.815606725146199e-06,
+      "loss": 1.3381,
+      "step": 19901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7262024879455566,
+      "learning_rate": 1.8146929824561405e-06,
+      "loss": 1.7243,
+      "step": 19902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7684569358825684,
+      "learning_rate": 1.8137792397660819e-06,
+      "loss": 1.3237,
+      "step": 19903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9274179935455322,
+      "learning_rate": 1.8128654970760235e-06,
+      "loss": 1.9546,
+      "step": 19904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7840349674224854,
+      "learning_rate": 1.8119517543859648e-06,
+      "loss": 1.5178,
+      "step": 19905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.620211601257324,
+      "learning_rate": 1.8110380116959066e-06,
+      "loss": 1.5121,
+      "step": 19906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.570798635482788,
+      "learning_rate": 1.8101242690058482e-06,
+      "loss": 1.4188,
+      "step": 19907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.00495982170105,
+      "learning_rate": 1.8092105263157896e-06,
+      "loss": 1.1436,
+      "step": 19908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3935165405273438,
+      "learning_rate": 1.8082967836257312e-06,
+      "loss": 1.3918,
+      "step": 19909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.495713233947754,
+      "learning_rate": 1.8073830409356728e-06,
+      "loss": 1.2696,
+      "step": 19910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3233659267425537,
+      "learning_rate": 1.8064692982456141e-06,
+      "loss": 2.1427,
+      "step": 19911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6860673427581787,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 1.5384,
+      "step": 19912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3616981506347656,
+      "learning_rate": 1.804641812865497e-06,
+      "loss": 1.4801,
+      "step": 19913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4693942070007324,
+      "learning_rate": 1.8037280701754387e-06,
+      "loss": 1.7227,
+      "step": 19914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.616316318511963,
+      "learning_rate": 1.80281432748538e-06,
+      "loss": 1.5553,
+      "step": 19915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.630763292312622,
+      "learning_rate": 1.8019005847953216e-06,
+      "loss": 1.5573,
+      "step": 19916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0211760997772217,
+      "learning_rate": 1.8009868421052634e-06,
+      "loss": 1.5537,
+      "step": 19917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.457059383392334,
+      "learning_rate": 1.8000730994152048e-06,
+      "loss": 1.4956,
+      "step": 19918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2833197116851807,
+      "learning_rate": 1.7991593567251464e-06,
+      "loss": 1.2446,
+      "step": 19919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7498862743377686,
+      "learning_rate": 1.798245614035088e-06,
+      "loss": 1.3883,
+      "step": 19920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.366515874862671,
+      "learning_rate": 1.7973318713450294e-06,
+      "loss": 1.1892,
+      "step": 19921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6745028495788574,
+      "learning_rate": 1.796418128654971e-06,
+      "loss": 1.7751,
+      "step": 19922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.639251470565796,
+      "learning_rate": 1.7955043859649123e-06,
+      "loss": 1.4157,
+      "step": 19923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.00327205657959,
+      "learning_rate": 1.794590643274854e-06,
+      "loss": 1.4327,
+      "step": 19924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.318009853363037,
+      "learning_rate": 1.7936769005847953e-06,
+      "loss": 1.212,
+      "step": 19925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2044975757598877,
+      "learning_rate": 1.7927631578947369e-06,
+      "loss": 1.1054,
+      "step": 19926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5744919776916504,
+      "learning_rate": 1.7918494152046787e-06,
+      "loss": 1.3239,
+      "step": 19927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3338804244995117,
+      "learning_rate": 1.79093567251462e-06,
+      "loss": 1.4279,
+      "step": 19928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3650357723236084,
+      "learning_rate": 1.7900219298245616e-06,
+      "loss": 1.1828,
+      "step": 19929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5427286624908447,
+      "learning_rate": 1.7891081871345032e-06,
+      "loss": 1.7745,
+      "step": 19930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0272295475006104,
+      "learning_rate": 1.7881944444444446e-06,
+      "loss": 1.296,
+      "step": 19931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.095349073410034,
+      "learning_rate": 1.7872807017543862e-06,
+      "loss": 1.3737,
+      "step": 19932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.461744785308838,
+      "learning_rate": 1.7863669590643276e-06,
+      "loss": 1.2906,
+      "step": 19933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5176727771759033,
+      "learning_rate": 1.7854532163742691e-06,
+      "loss": 1.5355,
+      "step": 19934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.326310396194458,
+      "learning_rate": 1.7845394736842105e-06,
+      "loss": 1.2702,
+      "step": 19935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.423088312149048,
+      "learning_rate": 1.783625730994152e-06,
+      "loss": 1.3359,
+      "step": 19936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5329430103302,
+      "learning_rate": 1.7827119883040937e-06,
+      "loss": 1.2943,
+      "step": 19937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.603870153427124,
+      "learning_rate": 1.7817982456140353e-06,
+      "loss": 1.5492,
+      "step": 19938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5149152278900146,
+      "learning_rate": 1.7808845029239769e-06,
+      "loss": 1.42,
+      "step": 19939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.432023286819458,
+      "learning_rate": 1.7799707602339184e-06,
+      "loss": 1.3005,
+      "step": 19940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1367626190185547,
+      "learning_rate": 1.7790570175438598e-06,
+      "loss": 1.2017,
+      "step": 19941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7825193405151367,
+      "learning_rate": 1.7781432748538014e-06,
+      "loss": 1.6219,
+      "step": 19942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8838908672332764,
+      "learning_rate": 1.7772295321637428e-06,
+      "loss": 1.6764,
+      "step": 19943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.723022699356079,
+      "learning_rate": 1.7763157894736844e-06,
+      "loss": 1.5461,
+      "step": 19944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4879136085510254,
+      "learning_rate": 1.7754020467836257e-06,
+      "loss": 1.6392,
+      "step": 19945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.85180401802063,
+      "learning_rate": 1.7744883040935673e-06,
+      "loss": 1.4413,
+      "step": 19946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.277902364730835,
+      "learning_rate": 1.773574561403509e-06,
+      "loss": 1.0926,
+      "step": 19947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.110309600830078,
+      "learning_rate": 1.7726608187134503e-06,
+      "loss": 1.1481,
+      "step": 19948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.645456075668335,
+      "learning_rate": 1.771747076023392e-06,
+      "loss": 1.4983,
+      "step": 19949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.438830614089966,
+      "learning_rate": 1.7708333333333337e-06,
+      "loss": 1.6045,
+      "step": 19950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.470383405685425,
+      "learning_rate": 1.769919590643275e-06,
+      "loss": 1.5442,
+      "step": 19951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.855301856994629,
+      "learning_rate": 1.7690058479532166e-06,
+      "loss": 1.4189,
+      "step": 19952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.506664752960205,
+      "learning_rate": 1.768092105263158e-06,
+      "loss": 1.3251,
+      "step": 19953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5775418281555176,
+      "learning_rate": 1.7671783625730996e-06,
+      "loss": 1.6408,
+      "step": 19954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.349433183670044,
+      "learning_rate": 1.766264619883041e-06,
+      "loss": 1.5771,
+      "step": 19955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.09384822845459,
+      "learning_rate": 1.7653508771929826e-06,
+      "loss": 1.233,
+      "step": 19956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5869102478027344,
+      "learning_rate": 1.764437134502924e-06,
+      "loss": 1.4163,
+      "step": 19957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7050387859344482,
+      "learning_rate": 1.7635233918128655e-06,
+      "loss": 1.5665,
+      "step": 19958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3045051097869873,
+      "learning_rate": 1.7626096491228073e-06,
+      "loss": 1.4182,
+      "step": 19959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.625080108642578,
+      "learning_rate": 1.7616959064327487e-06,
+      "loss": 1.5136,
+      "step": 19960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7634506225585938,
+      "learning_rate": 1.7607821637426903e-06,
+      "loss": 1.6075,
+      "step": 19961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.355830669403076,
+      "learning_rate": 1.7598684210526319e-06,
+      "loss": 1.2295,
+      "step": 19962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.689634323120117,
+      "learning_rate": 1.7589546783625732e-06,
+      "loss": 1.6148,
+      "step": 19963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7883089780807495,
+      "learning_rate": 1.7580409356725148e-06,
+      "loss": 1.1087,
+      "step": 19964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2247936725616455,
+      "learning_rate": 1.7571271929824562e-06,
+      "loss": 1.3888,
+      "step": 19965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1138572692871094,
+      "learning_rate": 1.7562134502923978e-06,
+      "loss": 1.3571,
+      "step": 19966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3735601902008057,
+      "learning_rate": 1.7552997076023392e-06,
+      "loss": 1.3,
+      "step": 19967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1340138912200928,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 1.0251,
+      "step": 19968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.754733085632324,
+      "learning_rate": 1.7534722222222223e-06,
+      "loss": 1.6322,
+      "step": 19969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.000494956970215,
+      "learning_rate": 1.752558479532164e-06,
+      "loss": 1.7287,
+      "step": 19970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.687638998031616,
+      "learning_rate": 1.7516447368421055e-06,
+      "loss": 1.4396,
+      "step": 19971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.256617307662964,
+      "learning_rate": 1.750730994152047e-06,
+      "loss": 1.2449,
+      "step": 19972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.907834529876709,
+      "learning_rate": 1.7498172514619885e-06,
+      "loss": 1.471,
+      "step": 19973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.438542366027832,
+      "learning_rate": 1.74890350877193e-06,
+      "loss": 1.4201,
+      "step": 19974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1988463401794434,
+      "learning_rate": 1.7479897660818714e-06,
+      "loss": 1.9208,
+      "step": 19975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.641932964324951,
+      "learning_rate": 1.747076023391813e-06,
+      "loss": 1.4431,
+      "step": 19976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2016806602478027,
+      "learning_rate": 1.7461622807017544e-06,
+      "loss": 1.2786,
+      "step": 19977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.488448143005371,
+      "learning_rate": 1.745248538011696e-06,
+      "loss": 1.7056,
+      "step": 19978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7025094032287598,
+      "learning_rate": 1.7443347953216376e-06,
+      "loss": 1.2464,
+      "step": 19979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7803103923797607,
+      "learning_rate": 1.743421052631579e-06,
+      "loss": 1.7339,
+      "step": 19980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3160767555236816,
+      "learning_rate": 1.7425073099415207e-06,
+      "loss": 1.3372,
+      "step": 19981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2389798164367676,
+      "learning_rate": 1.7415935672514623e-06,
+      "loss": 1.2047,
+      "step": 19982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.720153570175171,
+      "learning_rate": 1.7406798245614037e-06,
+      "loss": 1.4382,
+      "step": 19983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5557785034179688,
+      "learning_rate": 1.7397660818713453e-06,
+      "loss": 1.4858,
+      "step": 19984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.596540689468384,
+      "learning_rate": 1.7388523391812866e-06,
+      "loss": 1.5654,
+      "step": 19985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1867430210113525,
+      "learning_rate": 1.7379385964912282e-06,
+      "loss": 1.7624,
+      "step": 19986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1324596405029297,
+      "learning_rate": 1.7370248538011696e-06,
+      "loss": 1.3066,
+      "step": 19987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6514787673950195,
+      "learning_rate": 1.7361111111111112e-06,
+      "loss": 1.5054,
+      "step": 19988
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6842503547668457,
+      "learning_rate": 1.7351973684210528e-06,
+      "loss": 1.5918,
+      "step": 19989
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.422156810760498,
+      "learning_rate": 1.7342836257309942e-06,
+      "loss": 1.3977,
+      "step": 19990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3628695011138916,
+      "learning_rate": 1.733369883040936e-06,
+      "loss": 1.7553,
+      "step": 19991
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.466280937194824,
+      "learning_rate": 1.7324561403508775e-06,
+      "loss": 1.2706,
+      "step": 19992
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5134596824645996,
+      "learning_rate": 1.731542397660819e-06,
+      "loss": 1.5029,
+      "step": 19993
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.621767044067383,
+      "learning_rate": 1.7306286549707605e-06,
+      "loss": 1.4237,
+      "step": 19994
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.480067253112793,
+      "learning_rate": 1.7297149122807019e-06,
+      "loss": 1.2127,
+      "step": 19995
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.227774143218994,
+      "learning_rate": 1.7288011695906435e-06,
+      "loss": 1.1386,
+      "step": 19996
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0984063148498535,
+      "learning_rate": 1.7278874269005848e-06,
+      "loss": 1.059,
+      "step": 19997
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.730239152908325,
+      "learning_rate": 1.7269736842105264e-06,
+      "loss": 1.8044,
+      "step": 19998
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4094126224517822,
+      "learning_rate": 1.726059941520468e-06,
+      "loss": 1.25,
+      "step": 19999
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.334104537963867,
+      "learning_rate": 1.7251461988304094e-06,
+      "loss": 1.486,
+      "step": 20000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5050411224365234,
+      "learning_rate": 1.724232456140351e-06,
+      "loss": 1.701,
+      "step": 20001
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6111698150634766,
+      "learning_rate": 1.7233187134502928e-06,
+      "loss": 1.6512,
+      "step": 20002
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3046209812164307,
+      "learning_rate": 1.7224049707602341e-06,
+      "loss": 1.1671,
+      "step": 20003
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.87567138671875,
+      "learning_rate": 1.7214912280701757e-06,
+      "loss": 1.842,
+      "step": 20004
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5222373008728027,
+      "learning_rate": 1.720577485380117e-06,
+      "loss": 1.7383,
+      "step": 20005
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.208204984664917,
+      "learning_rate": 1.7196637426900587e-06,
+      "loss": 1.1805,
+      "step": 20006
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9406180381774902,
+      "learning_rate": 1.71875e-06,
+      "loss": 1.4793,
+      "step": 20007
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4683094024658203,
+      "learning_rate": 1.7178362573099416e-06,
+      "loss": 1.6035,
+      "step": 20008
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4279398918151855,
+      "learning_rate": 1.7169225146198832e-06,
+      "loss": 1.4258,
+      "step": 20009
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.181666374206543,
+      "learning_rate": 1.7160087719298246e-06,
+      "loss": 1.2127,
+      "step": 20010
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3600218296051025,
+      "learning_rate": 1.7150950292397662e-06,
+      "loss": 1.573,
+      "step": 20011
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.103400945663452,
+      "learning_rate": 1.7141812865497076e-06,
+      "loss": 1.9112,
+      "step": 20012
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6762278079986572,
+      "learning_rate": 1.7132675438596494e-06,
+      "loss": 1.6951,
+      "step": 20013
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2103400230407715,
+      "learning_rate": 1.712353801169591e-06,
+      "loss": 1.2405,
+      "step": 20014
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6740100383758545,
+      "learning_rate": 1.7114400584795323e-06,
+      "loss": 1.6405,
+      "step": 20015
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7163166999816895,
+      "learning_rate": 1.710526315789474e-06,
+      "loss": 1.4844,
+      "step": 20016
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.602797508239746,
+      "learning_rate": 1.7096125730994153e-06,
+      "loss": 1.5103,
+      "step": 20017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.12109637260437,
+      "learning_rate": 1.7086988304093569e-06,
+      "loss": 1.2157,
+      "step": 20018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2891793251037598,
+      "learning_rate": 1.7077850877192982e-06,
+      "loss": 1.4681,
+      "step": 20019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.198599100112915,
+      "learning_rate": 1.7068713450292398e-06,
+      "loss": 1.3286,
+      "step": 20020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.612374782562256,
+      "learning_rate": 1.7059576023391814e-06,
+      "loss": 1.4859,
+      "step": 20021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2705140113830566,
+      "learning_rate": 1.7050438596491228e-06,
+      "loss": 1.3493,
+      "step": 20022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.39821720123291,
+      "learning_rate": 1.7041301169590644e-06,
+      "loss": 1.4692,
+      "step": 20023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.045821189880371,
+      "learning_rate": 1.7032163742690062e-06,
+      "loss": 1.3675,
+      "step": 20024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9498848915100098,
+      "learning_rate": 1.7023026315789476e-06,
+      "loss": 1.6725,
+      "step": 20025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.320566177368164,
+      "learning_rate": 1.7013888888888891e-06,
+      "loss": 1.4297,
+      "step": 20026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.499345541000366,
+      "learning_rate": 1.7004751461988305e-06,
+      "loss": 1.3801,
+      "step": 20027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4180657863616943,
+      "learning_rate": 1.699561403508772e-06,
+      "loss": 1.4188,
+      "step": 20028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3695180416107178,
+      "learning_rate": 1.6986476608187135e-06,
+      "loss": 1.4177,
+      "step": 20029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3974523544311523,
+      "learning_rate": 1.697733918128655e-06,
+      "loss": 1.4451,
+      "step": 20030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.857241630554199,
+      "learning_rate": 1.6968201754385966e-06,
+      "loss": 1.6992,
+      "step": 20031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.524196147918701,
+      "learning_rate": 1.695906432748538e-06,
+      "loss": 1.6884,
+      "step": 20032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.384427785873413,
+      "learning_rate": 1.6949926900584796e-06,
+      "loss": 1.4,
+      "step": 20033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.92384934425354,
+      "learning_rate": 1.6940789473684214e-06,
+      "loss": 1.5818,
+      "step": 20034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.444309949874878,
+      "learning_rate": 1.6931652046783628e-06,
+      "loss": 1.5138,
+      "step": 20035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2925565242767334,
+      "learning_rate": 1.6922514619883044e-06,
+      "loss": 1.5957,
+      "step": 20036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2522082328796387,
+      "learning_rate": 1.6913377192982457e-06,
+      "loss": 1.2477,
+      "step": 20037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8902761936187744,
+      "learning_rate": 1.6904239766081873e-06,
+      "loss": 1.2889,
+      "step": 20038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.158351182937622,
+      "learning_rate": 1.6895102339181287e-06,
+      "loss": 1.1251,
+      "step": 20039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.323152780532837,
+      "learning_rate": 1.6885964912280703e-06,
+      "loss": 1.3477,
+      "step": 20040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2679853439331055,
+      "learning_rate": 1.6876827485380119e-06,
+      "loss": 1.9673,
+      "step": 20041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.475914478302002,
+      "learning_rate": 1.6867690058479532e-06,
+      "loss": 1.3113,
+      "step": 20042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.142484426498413,
+      "learning_rate": 1.6858552631578948e-06,
+      "loss": 1.2105,
+      "step": 20043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.377812385559082,
+      "learning_rate": 1.6849415204678362e-06,
+      "loss": 1.446,
+      "step": 20044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.637566328048706,
+      "learning_rate": 1.684027777777778e-06,
+      "loss": 1.5198,
+      "step": 20045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3535349369049072,
+      "learning_rate": 1.6831140350877196e-06,
+      "loss": 1.3372,
+      "step": 20046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.04349684715271,
+      "learning_rate": 1.682200292397661e-06,
+      "loss": 1.7263,
+      "step": 20047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.668931484222412,
+      "learning_rate": 1.6812865497076026e-06,
+      "loss": 1.484,
+      "step": 20048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.380774736404419,
+      "learning_rate": 1.680372807017544e-06,
+      "loss": 1.448,
+      "step": 20049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.522397518157959,
+      "learning_rate": 1.6794590643274855e-06,
+      "loss": 1.3619,
+      "step": 20050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.16108512878418,
+      "learning_rate": 1.678545321637427e-06,
+      "loss": 1.456,
+      "step": 20051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.600221633911133,
+      "learning_rate": 1.6776315789473685e-06,
+      "loss": 1.5706,
+      "step": 20052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.142374038696289,
+      "learning_rate": 1.67671783625731e-06,
+      "loss": 1.1729,
+      "step": 20053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5633158683776855,
+      "learning_rate": 1.6758040935672514e-06,
+      "loss": 1.2443,
+      "step": 20054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1180272102355957,
+      "learning_rate": 1.674890350877193e-06,
+      "loss": 1.333,
+      "step": 20055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3102166652679443,
+      "learning_rate": 1.6739766081871348e-06,
+      "loss": 1.2882,
+      "step": 20056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1307458877563477,
+      "learning_rate": 1.6730628654970762e-06,
+      "loss": 1.3581,
+      "step": 20057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2669878005981445,
+      "learning_rate": 1.6721491228070178e-06,
+      "loss": 1.2237,
+      "step": 20058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3042869567871094,
+      "learning_rate": 1.6712353801169592e-06,
+      "loss": 1.5643,
+      "step": 20059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.605135440826416,
+      "learning_rate": 1.6703216374269007e-06,
+      "loss": 1.8636,
+      "step": 20060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.576326370239258,
+      "learning_rate": 1.6694078947368423e-06,
+      "loss": 1.5601,
+      "step": 20061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2994954586029053,
+      "learning_rate": 1.6684941520467837e-06,
+      "loss": 1.5536,
+      "step": 20062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.137756824493408,
+      "learning_rate": 1.6675804093567253e-06,
+      "loss": 1.1388,
+      "step": 20063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.300443410873413,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2985,
+      "step": 20064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6372392177581787,
+      "learning_rate": 1.6657529239766082e-06,
+      "loss": 1.9916,
+      "step": 20065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.614779472351074,
+      "learning_rate": 1.66483918128655e-06,
+      "loss": 1.4409,
+      "step": 20066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6852123737335205,
+      "learning_rate": 1.6639254385964914e-06,
+      "loss": 1.6538,
+      "step": 20067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.503645658493042,
+      "learning_rate": 1.663011695906433e-06,
+      "loss": 1.5775,
+      "step": 20068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.939908504486084,
+      "learning_rate": 1.6620979532163744e-06,
+      "loss": 1.8081,
+      "step": 20069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5396132469177246,
+      "learning_rate": 1.661184210526316e-06,
+      "loss": 1.5067,
+      "step": 20070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.417794942855835,
+      "learning_rate": 1.6602704678362576e-06,
+      "loss": 1.2566,
+      "step": 20071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6467127799987793,
+      "learning_rate": 1.659356725146199e-06,
+      "loss": 1.76,
+      "step": 20072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1840949058532715,
+      "learning_rate": 1.6584429824561405e-06,
+      "loss": 1.426,
+      "step": 20073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2553207874298096,
+      "learning_rate": 1.6575292397660819e-06,
+      "loss": 1.38,
+      "step": 20074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.046780586242676,
+      "learning_rate": 1.6566154970760235e-06,
+      "loss": 1.224,
+      "step": 20075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.456315755844116,
+      "learning_rate": 1.6557017543859648e-06,
+      "loss": 1.2259,
+      "step": 20076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.320885181427002,
+      "learning_rate": 1.6547880116959066e-06,
+      "loss": 1.3456,
+      "step": 20077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.410672664642334,
+      "learning_rate": 1.6538742690058482e-06,
+      "loss": 1.2909,
+      "step": 20078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8573949337005615,
+      "learning_rate": 1.6529605263157896e-06,
+      "loss": 1.5043,
+      "step": 20079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.462165594100952,
+      "learning_rate": 1.6520467836257312e-06,
+      "loss": 1.4713,
+      "step": 20080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4943056106567383,
+      "learning_rate": 1.6511330409356726e-06,
+      "loss": 1.4226,
+      "step": 20081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.509012460708618,
+      "learning_rate": 1.6502192982456142e-06,
+      "loss": 1.4218,
+      "step": 20082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.434359312057495,
+      "learning_rate": 1.6493055555555557e-06,
+      "loss": 1.5055,
+      "step": 20083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2272424697875977,
+      "learning_rate": 1.6483918128654971e-06,
+      "loss": 1.2024,
+      "step": 20084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6418278217315674,
+      "learning_rate": 1.6474780701754387e-06,
+      "loss": 1.7185,
+      "step": 20085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4968457221984863,
+      "learning_rate": 1.64656432748538e-06,
+      "loss": 1.0225,
+      "step": 20086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1818082332611084,
+      "learning_rate": 1.6456505847953217e-06,
+      "loss": 1.3494,
+      "step": 20087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4431676864624023,
+      "learning_rate": 1.6447368421052635e-06,
+      "loss": 1.615,
+      "step": 20088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.375298023223877,
+      "learning_rate": 1.6438230994152048e-06,
+      "loss": 1.2913,
+      "step": 20089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8655943870544434,
+      "learning_rate": 1.6429093567251464e-06,
+      "loss": 1.8312,
+      "step": 20090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1605312824249268,
+      "learning_rate": 1.6419956140350878e-06,
+      "loss": 1.7129,
+      "step": 20091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2538952827453613,
+      "learning_rate": 1.6410818713450294e-06,
+      "loss": 1.2886,
+      "step": 20092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.698740243911743,
+      "learning_rate": 1.640168128654971e-06,
+      "loss": 1.7345,
+      "step": 20093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.635669469833374,
+      "learning_rate": 1.6392543859649123e-06,
+      "loss": 1.4247,
+      "step": 20094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0040576457977295,
+      "learning_rate": 1.638340643274854e-06,
+      "loss": 1.2133,
+      "step": 20095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9303040504455566,
+      "learning_rate": 1.6374269005847953e-06,
+      "loss": 1.6559,
+      "step": 20096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.728959560394287,
+      "learning_rate": 1.6365131578947369e-06,
+      "loss": 1.5065,
+      "step": 20097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4626452922821045,
+      "learning_rate": 1.6355994152046787e-06,
+      "loss": 1.4581,
+      "step": 20098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.672001838684082,
+      "learning_rate": 1.63468567251462e-06,
+      "loss": 1.405,
+      "step": 20099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1223347187042236,
+      "learning_rate": 1.6337719298245616e-06,
+      "loss": 1.6681,
+      "step": 20100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.836085319519043,
+      "learning_rate": 1.632858187134503e-06,
+      "loss": 1.3782,
+      "step": 20101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6756796836853027,
+      "learning_rate": 1.6319444444444446e-06,
+      "loss": 1.5621,
+      "step": 20102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3825273513793945,
+      "learning_rate": 1.6310307017543862e-06,
+      "loss": 1.3604,
+      "step": 20103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4777119159698486,
+      "learning_rate": 1.6301169590643276e-06,
+      "loss": 1.4281,
+      "step": 20104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.859905242919922,
+      "learning_rate": 1.6292032163742692e-06,
+      "loss": 1.8991,
+      "step": 20105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.437936782836914,
+      "learning_rate": 1.6282894736842105e-06,
+      "loss": 1.3893,
+      "step": 20106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7668755054473877,
+      "learning_rate": 1.6273757309941521e-06,
+      "loss": 1.6063,
+      "step": 20107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.562927722930908,
+      "learning_rate": 1.6264619883040935e-06,
+      "loss": 1.2513,
+      "step": 20108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3945302963256836,
+      "learning_rate": 1.6255482456140353e-06,
+      "loss": 1.5263,
+      "step": 20109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3392446041107178,
+      "learning_rate": 1.6246345029239769e-06,
+      "loss": 1.3699,
+      "step": 20110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.585690498352051,
+      "learning_rate": 1.6237207602339182e-06,
+      "loss": 1.3384,
+      "step": 20111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.619532346725464,
+      "learning_rate": 1.6228070175438598e-06,
+      "loss": 1.7244,
+      "step": 20112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2419614791870117,
+      "learning_rate": 1.6218932748538014e-06,
+      "loss": 1.3597,
+      "step": 20113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3963840007781982,
+      "learning_rate": 1.6209795321637428e-06,
+      "loss": 1.2161,
+      "step": 20114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.230980157852173,
+      "learning_rate": 1.6200657894736844e-06,
+      "loss": 1.2717,
+      "step": 20115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.438843011856079,
+      "learning_rate": 1.6191520467836258e-06,
+      "loss": 1.0701,
+      "step": 20116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.87176251411438,
+      "learning_rate": 1.6182383040935673e-06,
+      "loss": 1.6003,
+      "step": 20117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3404200077056885,
+      "learning_rate": 1.6173245614035087e-06,
+      "loss": 1.4179,
+      "step": 20118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3822789192199707,
+      "learning_rate": 1.6164108187134503e-06,
+      "loss": 1.3948,
+      "step": 20119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.584965467453003,
+      "learning_rate": 1.615497076023392e-06,
+      "loss": 1.5544,
+      "step": 20120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5727217197418213,
+      "learning_rate": 1.6145833333333335e-06,
+      "loss": 1.3954,
+      "step": 20121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.002720832824707,
+      "learning_rate": 1.613669590643275e-06,
+      "loss": 1.735,
+      "step": 20122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.500781297683716,
+      "learning_rate": 1.6127558479532166e-06,
+      "loss": 1.4953,
+      "step": 20123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8238682746887207,
+      "learning_rate": 1.611842105263158e-06,
+      "loss": 1.7118,
+      "step": 20124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.49503493309021,
+      "learning_rate": 1.6109283625730996e-06,
+      "loss": 1.5561,
+      "step": 20125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8478658199310303,
+      "learning_rate": 1.610014619883041e-06,
+      "loss": 1.5281,
+      "step": 20126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.466520071029663,
+      "learning_rate": 1.6091008771929826e-06,
+      "loss": 1.4441,
+      "step": 20127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9634613990783691,
+      "learning_rate": 1.608187134502924e-06,
+      "loss": 1.265,
+      "step": 20128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.595290184020996,
+      "learning_rate": 1.6072733918128655e-06,
+      "loss": 1.6721,
+      "step": 20129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.884474754333496,
+      "learning_rate": 1.6063596491228073e-06,
+      "loss": 1.3224,
+      "step": 20130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.69555401802063,
+      "learning_rate": 1.6054459064327487e-06,
+      "loss": 1.3113,
+      "step": 20131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.156895399093628,
+      "learning_rate": 1.6045321637426903e-06,
+      "loss": 1.4553,
+      "step": 20132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6350560188293457,
+      "learning_rate": 1.6036184210526319e-06,
+      "loss": 1.4121,
+      "step": 20133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.159168243408203,
+      "learning_rate": 1.6027046783625732e-06,
+      "loss": 1.2646,
+      "step": 20134
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3947246074676514,
+      "learning_rate": 1.6017909356725148e-06,
+      "loss": 1.5001,
+      "step": 20135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8476624488830566,
+      "learning_rate": 1.6008771929824562e-06,
+      "loss": 1.6035,
+      "step": 20136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1318483352661133,
+      "learning_rate": 1.5999634502923978e-06,
+      "loss": 1.1234,
+      "step": 20137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2241463661193848,
+      "learning_rate": 1.5990497076023392e-06,
+      "loss": 1.3179,
+      "step": 20138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.599757194519043,
+      "learning_rate": 1.5981359649122808e-06,
+      "loss": 1.7061,
+      "step": 20139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7430305480957031,
+      "learning_rate": 1.5972222222222221e-06,
+      "loss": 0.945,
+      "step": 20140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6493005752563477,
+      "learning_rate": 1.596308479532164e-06,
+      "loss": 1.3682,
+      "step": 20141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5432469844818115,
+      "learning_rate": 1.5953947368421055e-06,
+      "loss": 1.4138,
+      "step": 20142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0283150672912598,
+      "learning_rate": 1.5944809941520469e-06,
+      "loss": 1.0874,
+      "step": 20143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7023727893829346,
+      "learning_rate": 1.5935672514619885e-06,
+      "loss": 1.5269,
+      "step": 20144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8011709451675415,
+      "learning_rate": 1.59265350877193e-06,
+      "loss": 1.0615,
+      "step": 20145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.673762559890747,
+      "learning_rate": 1.5917397660818714e-06,
+      "loss": 1.4263,
+      "step": 20146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5971405506134033,
+      "learning_rate": 1.590826023391813e-06,
+      "loss": 1.2385,
+      "step": 20147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.327160120010376,
+      "learning_rate": 1.5899122807017544e-06,
+      "loss": 1.5664,
+      "step": 20148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7773616313934326,
+      "learning_rate": 1.588998538011696e-06,
+      "loss": 1.7245,
+      "step": 20149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.160775899887085,
+      "learning_rate": 1.5880847953216374e-06,
+      "loss": 1.3877,
+      "step": 20150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2078471183776855,
+      "learning_rate": 1.587171052631579e-06,
+      "loss": 1.0322,
+      "step": 20151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4020609855651855,
+      "learning_rate": 1.5862573099415207e-06,
+      "loss": 1.358,
+      "step": 20152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1259758472442627,
+      "learning_rate": 1.5853435672514621e-06,
+      "loss": 1.3072,
+      "step": 20153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.600874662399292,
+      "learning_rate": 1.5844298245614037e-06,
+      "loss": 1.2464,
+      "step": 20154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7657365798950195,
+      "learning_rate": 1.5835160818713453e-06,
+      "loss": 1.3787,
+      "step": 20155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.617204189300537,
+      "learning_rate": 1.5826023391812867e-06,
+      "loss": 1.3507,
+      "step": 20156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6670305728912354,
+      "learning_rate": 1.5816885964912282e-06,
+      "loss": 1.5509,
+      "step": 20157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3406221866607666,
+      "learning_rate": 1.5807748538011696e-06,
+      "loss": 1.3734,
+      "step": 20158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9640684127807617,
+      "learning_rate": 1.5798611111111112e-06,
+      "loss": 1.751,
+      "step": 20159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8741705417633057,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 1.5949,
+      "step": 20160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4604039192199707,
+      "learning_rate": 1.5780336257309942e-06,
+      "loss": 1.5125,
+      "step": 20161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5413761138916016,
+      "learning_rate": 1.577119883040936e-06,
+      "loss": 1.4898,
+      "step": 20162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5539650917053223,
+      "learning_rate": 1.5762061403508773e-06,
+      "loss": 1.415,
+      "step": 20163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.649824380874634,
+      "learning_rate": 1.575292397660819e-06,
+      "loss": 1.7754,
+      "step": 20164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.258908271789551,
+      "learning_rate": 1.5743786549707605e-06,
+      "loss": 1.3004,
+      "step": 20165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4041783809661865,
+      "learning_rate": 1.5734649122807019e-06,
+      "loss": 1.3855,
+      "step": 20166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.639739751815796,
+      "learning_rate": 1.5725511695906435e-06,
+      "loss": 1.197,
+      "step": 20167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6247966289520264,
+      "learning_rate": 1.5716374269005848e-06,
+      "loss": 1.2689,
+      "step": 20168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5037903785705566,
+      "learning_rate": 1.5707236842105264e-06,
+      "loss": 1.2346,
+      "step": 20169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1736834049224854,
+      "learning_rate": 1.5698099415204678e-06,
+      "loss": 1.1567,
+      "step": 20170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3677000999450684,
+      "learning_rate": 1.5688961988304094e-06,
+      "loss": 1.4895,
+      "step": 20171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0513429641723633,
+      "learning_rate": 1.567982456140351e-06,
+      "loss": 1.2641,
+      "step": 20172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3223965167999268,
+      "learning_rate": 1.5670687134502926e-06,
+      "loss": 1.2017,
+      "step": 20173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.244312286376953,
+      "learning_rate": 1.5661549707602342e-06,
+      "loss": 1.2868,
+      "step": 20174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8282299041748047,
+      "learning_rate": 1.5652412280701757e-06,
+      "loss": 1.9655,
+      "step": 20175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.227548837661743,
+      "learning_rate": 1.5643274853801171e-06,
+      "loss": 1.3537,
+      "step": 20176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.36071515083313,
+      "learning_rate": 1.5634137426900587e-06,
+      "loss": 1.2014,
+      "step": 20177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.58927321434021,
+      "learning_rate": 1.5625e-06,
+      "loss": 1.5302,
+      "step": 20178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.246425151824951,
+      "learning_rate": 1.5615862573099417e-06,
+      "loss": 1.432,
+      "step": 20179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2228317260742188,
+      "learning_rate": 1.560672514619883e-06,
+      "loss": 1.3253,
+      "step": 20180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.416900873184204,
+      "learning_rate": 1.5597587719298246e-06,
+      "loss": 1.3392,
+      "step": 20181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4585511684417725,
+      "learning_rate": 1.5588450292397662e-06,
+      "loss": 1.2102,
+      "step": 20182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8244380950927734,
+      "learning_rate": 1.5579312865497076e-06,
+      "loss": 1.4671,
+      "step": 20183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3087944984436035,
+      "learning_rate": 1.5570175438596494e-06,
+      "loss": 1.451,
+      "step": 20184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.473238229751587,
+      "learning_rate": 1.556103801169591e-06,
+      "loss": 1.2807,
+      "step": 20185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5180509090423584,
+      "learning_rate": 1.5551900584795323e-06,
+      "loss": 1.5521,
+      "step": 20186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.57995343208313,
+      "learning_rate": 1.554276315789474e-06,
+      "loss": 1.5832,
+      "step": 20187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.182029962539673,
+      "learning_rate": 1.5533625730994153e-06,
+      "loss": 1.0943,
+      "step": 20188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3731346130371094,
+      "learning_rate": 1.5524488304093569e-06,
+      "loss": 1.5703,
+      "step": 20189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3176844120025635,
+      "learning_rate": 1.5515350877192983e-06,
+      "loss": 1.3983,
+      "step": 20190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.184579610824585,
+      "learning_rate": 1.5506213450292398e-06,
+      "loss": 1.1743,
+      "step": 20191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1734862327575684,
+      "learning_rate": 1.5497076023391812e-06,
+      "loss": 1.2677,
+      "step": 20192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2582316398620605,
+      "learning_rate": 1.5487938596491228e-06,
+      "loss": 1.398,
+      "step": 20193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.276656150817871,
+      "learning_rate": 1.5478801169590644e-06,
+      "loss": 1.5951,
+      "step": 20194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3916866779327393,
+      "learning_rate": 1.5469663742690062e-06,
+      "loss": 1.6785,
+      "step": 20195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.932678699493408,
+      "learning_rate": 1.5460526315789476e-06,
+      "loss": 1.5948,
+      "step": 20196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9258816242218018,
+      "learning_rate": 1.5451388888888892e-06,
+      "loss": 1.702,
+      "step": 20197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2184112071990967,
+      "learning_rate": 1.5442251461988305e-06,
+      "loss": 1.2418,
+      "step": 20198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0199999809265137,
+      "learning_rate": 1.5433114035087721e-06,
+      "loss": 1.2196,
+      "step": 20199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.259617567062378,
+      "learning_rate": 1.5423976608187135e-06,
+      "loss": 1.3187,
+      "step": 20200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2095255851745605,
+      "learning_rate": 1.541483918128655e-06,
+      "loss": 1.2708,
+      "step": 20201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6427555084228516,
+      "learning_rate": 1.5405701754385964e-06,
+      "loss": 1.7796,
+      "step": 20202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.549551486968994,
+      "learning_rate": 1.539656432748538e-06,
+      "loss": 1.4362,
+      "step": 20203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.812631845474243,
+      "learning_rate": 1.5387426900584796e-06,
+      "loss": 1.8681,
+      "step": 20204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.287174940109253,
+      "learning_rate": 1.5378289473684212e-06,
+      "loss": 1.0581,
+      "step": 20205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2646842002868652,
+      "learning_rate": 1.5369152046783628e-06,
+      "loss": 1.4312,
+      "step": 20206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.576796054840088,
+      "learning_rate": 1.5360014619883044e-06,
+      "loss": 1.3469,
+      "step": 20207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2975640296936035,
+      "learning_rate": 1.5350877192982458e-06,
+      "loss": 1.1653,
+      "step": 20208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1586947441101074,
+      "learning_rate": 1.5341739766081873e-06,
+      "loss": 1.3995,
+      "step": 20209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7243053913116455,
+      "learning_rate": 1.5332602339181287e-06,
+      "loss": 1.4376,
+      "step": 20210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1773719787597656,
+      "learning_rate": 1.5323464912280703e-06,
+      "loss": 1.1151,
+      "step": 20211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6361355781555176,
+      "learning_rate": 1.5314327485380117e-06,
+      "loss": 1.9349,
+      "step": 20212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1474387645721436,
+      "learning_rate": 1.5305190058479533e-06,
+      "loss": 1.2055,
+      "step": 20213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3938000202178955,
+      "learning_rate": 1.5296052631578948e-06,
+      "loss": 1.5107,
+      "step": 20214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2432072162628174,
+      "learning_rate": 1.5286915204678362e-06,
+      "loss": 1.3076,
+      "step": 20215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4258038997650146,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 1.5662,
+      "step": 20216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0723178386688232,
+      "learning_rate": 1.5268640350877196e-06,
+      "loss": 1.2699,
+      "step": 20217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.598231077194214,
+      "learning_rate": 1.525950292397661e-06,
+      "loss": 1.6285,
+      "step": 20218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.861783981323242,
+      "learning_rate": 1.5250365497076026e-06,
+      "loss": 1.6134,
+      "step": 20219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3613460063934326,
+      "learning_rate": 1.524122807017544e-06,
+      "loss": 1.2932,
+      "step": 20220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.996638536453247,
+      "learning_rate": 1.5232090643274855e-06,
+      "loss": 1.6516,
+      "step": 20221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5276877880096436,
+      "learning_rate": 1.522295321637427e-06,
+      "loss": 1.4518,
+      "step": 20222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.710224151611328,
+      "learning_rate": 1.5213815789473685e-06,
+      "loss": 1.6628,
+      "step": 20223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3581695556640625,
+      "learning_rate": 1.52046783625731e-06,
+      "loss": 1.3587,
+      "step": 20224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4470415115356445,
+      "learning_rate": 1.5195540935672514e-06,
+      "loss": 1.6027,
+      "step": 20225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.216708183288574,
+      "learning_rate": 1.518640350877193e-06,
+      "loss": 1.2809,
+      "step": 20226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.760425329208374,
+      "learning_rate": 1.5177266081871348e-06,
+      "loss": 1.7042,
+      "step": 20227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9994529485702515,
+      "learning_rate": 1.5168128654970762e-06,
+      "loss": 1.312,
+      "step": 20228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9489753246307373,
+      "learning_rate": 1.5158991228070178e-06,
+      "loss": 1.6041,
+      "step": 20229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3554282188415527,
+      "learning_rate": 1.5149853801169592e-06,
+      "loss": 1.2212,
+      "step": 20230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.213906764984131,
+      "learning_rate": 1.5140716374269008e-06,
+      "loss": 1.2285,
+      "step": 20231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.224414110183716,
+      "learning_rate": 1.5131578947368421e-06,
+      "loss": 1.517,
+      "step": 20232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1971120834350586,
+      "learning_rate": 1.5122441520467837e-06,
+      "loss": 1.9237,
+      "step": 20233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.125171422958374,
+      "learning_rate": 1.5113304093567253e-06,
+      "loss": 1.0831,
+      "step": 20234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2015483379364014,
+      "learning_rate": 1.5104166666666667e-06,
+      "loss": 1.324,
+      "step": 20235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5036611557006836,
+      "learning_rate": 1.5095029239766083e-06,
+      "loss": 1.5898,
+      "step": 20236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.662318706512451,
+      "learning_rate": 1.50858918128655e-06,
+      "loss": 1.3825,
+      "step": 20237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.360002279281616,
+      "learning_rate": 1.5076754385964914e-06,
+      "loss": 1.5118,
+      "step": 20238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5196826457977295,
+      "learning_rate": 1.506761695906433e-06,
+      "loss": 1.4058,
+      "step": 20239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.476442813873291,
+      "learning_rate": 1.5058479532163744e-06,
+      "loss": 1.2603,
+      "step": 20240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6942315101623535,
+      "learning_rate": 1.504934210526316e-06,
+      "loss": 1.3978,
+      "step": 20241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.580744504928589,
+      "learning_rate": 1.5040204678362574e-06,
+      "loss": 1.3734,
+      "step": 20242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.038884401321411,
+      "learning_rate": 1.503106725146199e-06,
+      "loss": 1.7693,
+      "step": 20243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0422558784484863,
+      "learning_rate": 1.5021929824561405e-06,
+      "loss": 1.324,
+      "step": 20244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.540574312210083,
+      "learning_rate": 1.501279239766082e-06,
+      "loss": 1.2998,
+      "step": 20245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3169445991516113,
+      "learning_rate": 1.5003654970760235e-06,
+      "loss": 1.2927,
+      "step": 20246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.757744312286377,
+      "learning_rate": 1.4994517543859649e-06,
+      "loss": 1.4394,
+      "step": 20247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.800508737564087,
+      "learning_rate": 1.4985380116959067e-06,
+      "loss": 1.4448,
+      "step": 20248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.690891981124878,
+      "learning_rate": 1.4976242690058482e-06,
+      "loss": 1.4076,
+      "step": 20249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4346935749053955,
+      "learning_rate": 1.4967105263157896e-06,
+      "loss": 1.5541,
+      "step": 20250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.944612979888916,
+      "learning_rate": 1.4957967836257312e-06,
+      "loss": 1.7274,
+      "step": 20251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3089966773986816,
+      "learning_rate": 1.4948830409356726e-06,
+      "loss": 1.4368,
+      "step": 20252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.56599760055542,
+      "learning_rate": 1.4939692982456142e-06,
+      "loss": 1.1554,
+      "step": 20253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4148635864257812,
+      "learning_rate": 1.4930555555555555e-06,
+      "loss": 1.6468,
+      "step": 20254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4674320220947266,
+      "learning_rate": 1.4921418128654971e-06,
+      "loss": 1.4667,
+      "step": 20255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4834952354431152,
+      "learning_rate": 1.4912280701754387e-06,
+      "loss": 1.3475,
+      "step": 20256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.610206127166748,
+      "learning_rate": 1.49031432748538e-06,
+      "loss": 1.6489,
+      "step": 20257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.483353614807129,
+      "learning_rate": 1.4894005847953217e-06,
+      "loss": 1.3506,
+      "step": 20258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0563085079193115,
+      "learning_rate": 1.4884868421052635e-06,
+      "loss": 1.1963,
+      "step": 20259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6211531162261963,
+      "learning_rate": 1.4875730994152048e-06,
+      "loss": 1.4456,
+      "step": 20260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0787103176116943,
+      "learning_rate": 1.4866593567251464e-06,
+      "loss": 1.2941,
+      "step": 20261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.356549024581909,
+      "learning_rate": 1.4857456140350878e-06,
+      "loss": 1.2618,
+      "step": 20262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6544458866119385,
+      "learning_rate": 1.4848318713450294e-06,
+      "loss": 1.3183,
+      "step": 20263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.796549081802368,
+      "learning_rate": 1.4839181286549708e-06,
+      "loss": 1.7894,
+      "step": 20264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9497371912002563,
+      "learning_rate": 1.4830043859649124e-06,
+      "loss": 1.1233,
+      "step": 20265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9247610569000244,
+      "learning_rate": 1.482090643274854e-06,
+      "loss": 1.6339,
+      "step": 20266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.164456605911255,
+      "learning_rate": 1.4811769005847953e-06,
+      "loss": 1.194,
+      "step": 20267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3613436222076416,
+      "learning_rate": 1.480263157894737e-06,
+      "loss": 1.4478,
+      "step": 20268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4988372325897217,
+      "learning_rate": 1.4793494152046787e-06,
+      "loss": 1.438,
+      "step": 20269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.756746768951416,
+      "learning_rate": 1.47843567251462e-06,
+      "loss": 1.5788,
+      "step": 20270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2033305168151855,
+      "learning_rate": 1.4775219298245617e-06,
+      "loss": 1.1145,
+      "step": 20271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4491515159606934,
+      "learning_rate": 1.476608187134503e-06,
+      "loss": 1.1879,
+      "step": 20272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4879274368286133,
+      "learning_rate": 1.4756944444444446e-06,
+      "loss": 1.3713,
+      "step": 20273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.53523850440979,
+      "learning_rate": 1.474780701754386e-06,
+      "loss": 1.6781,
+      "step": 20274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7212109565734863,
+      "learning_rate": 1.4738669590643276e-06,
+      "loss": 1.2773,
+      "step": 20275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6737592220306396,
+      "learning_rate": 1.4729532163742692e-06,
+      "loss": 1.6647,
+      "step": 20276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2835378646850586,
+      "learning_rate": 1.4720394736842105e-06,
+      "loss": 1.6386,
+      "step": 20277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2759950160980225,
+      "learning_rate": 1.4711257309941521e-06,
+      "loss": 1.2436,
+      "step": 20278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8793394565582275,
+      "learning_rate": 1.4702119883040935e-06,
+      "loss": 1.7204,
+      "step": 20279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.505282163619995,
+      "learning_rate": 1.4692982456140353e-06,
+      "loss": 1.3999,
+      "step": 20280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.62079119682312,
+      "learning_rate": 1.4683845029239769e-06,
+      "loss": 1.3102,
+      "step": 20281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2142350673675537,
+      "learning_rate": 1.4674707602339183e-06,
+      "loss": 1.1996,
+      "step": 20282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.56406307220459,
+      "learning_rate": 1.4665570175438598e-06,
+      "loss": 1.4163,
+      "step": 20283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.128207206726074,
+      "learning_rate": 1.4656432748538012e-06,
+      "loss": 1.448,
+      "step": 20284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.42545223236084,
+      "learning_rate": 1.4647295321637428e-06,
+      "loss": 1.5105,
+      "step": 20285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.884032964706421,
+      "learning_rate": 1.4638157894736844e-06,
+      "loss": 1.8359,
+      "step": 20286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.502753257751465,
+      "learning_rate": 1.4629020467836258e-06,
+      "loss": 1.3116,
+      "step": 20287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4079065322875977,
+      "learning_rate": 1.4619883040935674e-06,
+      "loss": 1.4102,
+      "step": 20288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6884400844573975,
+      "learning_rate": 1.4610745614035087e-06,
+      "loss": 1.2027,
+      "step": 20289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.374774932861328,
+      "learning_rate": 1.4601608187134503e-06,
+      "loss": 1.7006,
+      "step": 20290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7091681957244873,
+      "learning_rate": 1.4592470760233921e-06,
+      "loss": 1.5436,
+      "step": 20291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.184535026550293,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 1.1268,
+      "step": 20292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3251097202301025,
+      "learning_rate": 1.457419590643275e-06,
+      "loss": 1.3439,
+      "step": 20293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.939928650856018,
+      "learning_rate": 1.4565058479532164e-06,
+      "loss": 1.1502,
+      "step": 20294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8894684314727783,
+      "learning_rate": 1.455592105263158e-06,
+      "loss": 1.6542,
+      "step": 20295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3007023334503174,
+      "learning_rate": 1.4546783625730996e-06,
+      "loss": 1.3424,
+      "step": 20296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5607190132141113,
+      "learning_rate": 1.453764619883041e-06,
+      "loss": 1.5297,
+      "step": 20297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.130650758743286,
+      "learning_rate": 1.4528508771929826e-06,
+      "loss": 1.2277,
+      "step": 20298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.925135850906372,
+      "learning_rate": 1.451937134502924e-06,
+      "loss": 2.038,
+      "step": 20299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.073198080062866,
+      "learning_rate": 1.4510233918128655e-06,
+      "loss": 1.9092,
+      "step": 20300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5509445667266846,
+      "learning_rate": 1.4501096491228073e-06,
+      "loss": 1.2976,
+      "step": 20301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.924452781677246,
+      "learning_rate": 1.4491959064327487e-06,
+      "loss": 1.2816,
+      "step": 20302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.275031328201294,
+      "learning_rate": 1.4482821637426903e-06,
+      "loss": 1.3101,
+      "step": 20303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.682858943939209,
+      "learning_rate": 1.4473684210526317e-06,
+      "loss": 1.6743,
+      "step": 20304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6296889781951904,
+      "learning_rate": 1.4464546783625733e-06,
+      "loss": 1.1927,
+      "step": 20305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.730456829071045,
+      "learning_rate": 1.4455409356725148e-06,
+      "loss": 1.5399,
+      "step": 20306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4286420345306396,
+      "learning_rate": 1.4446271929824562e-06,
+      "loss": 1.5623,
+      "step": 20307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8426687717437744,
+      "learning_rate": 1.4437134502923978e-06,
+      "loss": 1.4276,
+      "step": 20308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.481128692626953,
+      "learning_rate": 1.4427997076023392e-06,
+      "loss": 1.5343,
+      "step": 20309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3535335063934326,
+      "learning_rate": 1.4418859649122808e-06,
+      "loss": 1.0392,
+      "step": 20310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.542644500732422,
+      "learning_rate": 1.4409722222222221e-06,
+      "loss": 1.3019,
+      "step": 20311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.517061948776245,
+      "learning_rate": 1.440058479532164e-06,
+      "loss": 1.3542,
+      "step": 20312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.930094003677368,
+      "learning_rate": 1.4391447368421055e-06,
+      "loss": 1.2768,
+      "step": 20313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0558865070343018,
+      "learning_rate": 1.438230994152047e-06,
+      "loss": 1.7198,
+      "step": 20314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8769564628601074,
+      "learning_rate": 1.4373172514619885e-06,
+      "loss": 1.4861,
+      "step": 20315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6559336185455322,
+      "learning_rate": 1.4364035087719299e-06,
+      "loss": 1.6074,
+      "step": 20316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5559909343719482,
+      "learning_rate": 1.4354897660818714e-06,
+      "loss": 1.3283,
+      "step": 20317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5433731079101562,
+      "learning_rate": 1.434576023391813e-06,
+      "loss": 1.6487,
+      "step": 20318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.511812210083008,
+      "learning_rate": 1.4336622807017544e-06,
+      "loss": 1.3426,
+      "step": 20319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.426903247833252,
+      "learning_rate": 1.432748538011696e-06,
+      "loss": 1.5804,
+      "step": 20320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.099842071533203,
+      "learning_rate": 1.4318347953216374e-06,
+      "loss": 1.4634,
+      "step": 20321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.740499496459961,
+      "learning_rate": 1.430921052631579e-06,
+      "loss": 1.5361,
+      "step": 20322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.402468681335449,
+      "learning_rate": 1.4300073099415208e-06,
+      "loss": 1.5327,
+      "step": 20323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9820873737335205,
+      "learning_rate": 1.4290935672514621e-06,
+      "loss": 1.8782,
+      "step": 20324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.360135078430176,
+      "learning_rate": 1.4281798245614037e-06,
+      "loss": 1.5234,
+      "step": 20325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3049538135528564,
+      "learning_rate": 1.427266081871345e-06,
+      "loss": 1.9201,
+      "step": 20326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.278146982192993,
+      "learning_rate": 1.4263523391812867e-06,
+      "loss": 1.3574,
+      "step": 20327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2226994037628174,
+      "learning_rate": 1.4254385964912283e-06,
+      "loss": 1.1908,
+      "step": 20328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3475449085235596,
+      "learning_rate": 1.4245248538011696e-06,
+      "loss": 1.1586,
+      "step": 20329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.38411808013916,
+      "learning_rate": 1.4236111111111112e-06,
+      "loss": 1.3332,
+      "step": 20330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4765329360961914,
+      "learning_rate": 1.4226973684210526e-06,
+      "loss": 1.1868,
+      "step": 20331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.357858419418335,
+      "learning_rate": 1.4217836257309942e-06,
+      "loss": 1.263,
+      "step": 20332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3560101985931396,
+      "learning_rate": 1.420869883040936e-06,
+      "loss": 1.3035,
+      "step": 20333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3601043224334717,
+      "learning_rate": 1.4199561403508774e-06,
+      "loss": 1.3762,
+      "step": 20334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4212992191314697,
+      "learning_rate": 1.419042397660819e-06,
+      "loss": 1.1856,
+      "step": 20335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2245919704437256,
+      "learning_rate": 1.4181286549707603e-06,
+      "loss": 1.8522,
+      "step": 20336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.991922378540039,
+      "learning_rate": 1.417214912280702e-06,
+      "loss": 1.6695,
+      "step": 20337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3162107467651367,
+      "learning_rate": 1.4163011695906435e-06,
+      "loss": 1.2828,
+      "step": 20338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.890073537826538,
+      "learning_rate": 1.4153874269005849e-06,
+      "loss": 1.595,
+      "step": 20339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9043985605239868,
+      "learning_rate": 1.4144736842105264e-06,
+      "loss": 1.0544,
+      "step": 20340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.299917221069336,
+      "learning_rate": 1.4135599415204678e-06,
+      "loss": 1.4764,
+      "step": 20341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.602952480316162,
+      "learning_rate": 1.4126461988304094e-06,
+      "loss": 1.5139,
+      "step": 20342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1644177436828613,
+      "learning_rate": 1.4117324561403508e-06,
+      "loss": 1.1729,
+      "step": 20343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7538013458251953,
+      "learning_rate": 1.4108187134502926e-06,
+      "loss": 1.7753,
+      "step": 20344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.863754391670227,
+      "learning_rate": 1.4099049707602342e-06,
+      "loss": 1.2655,
+      "step": 20345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.079523801803589,
+      "learning_rate": 1.4089912280701755e-06,
+      "loss": 1.1573,
+      "step": 20346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1271467208862305,
+      "learning_rate": 1.4080774853801171e-06,
+      "loss": 1.1752,
+      "step": 20347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.924651861190796,
+      "learning_rate": 1.4071637426900587e-06,
+      "loss": 1.7069,
+      "step": 20348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1985766887664795,
+      "learning_rate": 1.40625e-06,
+      "loss": 1.1684,
+      "step": 20349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9433183670043945,
+      "learning_rate": 1.4053362573099417e-06,
+      "loss": 1.5538,
+      "step": 20350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7733421325683594,
+      "learning_rate": 1.404422514619883e-06,
+      "loss": 1.5988,
+      "step": 20351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.957825779914856,
+      "learning_rate": 1.4035087719298246e-06,
+      "loss": 1.1725,
+      "step": 20352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2978551387786865,
+      "learning_rate": 1.402595029239766e-06,
+      "loss": 1.924,
+      "step": 20353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.805572509765625,
+      "learning_rate": 1.4016812865497076e-06,
+      "loss": 1.7485,
+      "step": 20354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7203681468963623,
+      "learning_rate": 1.4007675438596494e-06,
+      "loss": 1.7097,
+      "step": 20355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.447751522064209,
+      "learning_rate": 1.3998538011695908e-06,
+      "loss": 1.4394,
+      "step": 20356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.888010263442993,
+      "learning_rate": 1.3989400584795324e-06,
+      "loss": 1.866,
+      "step": 20357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.507932186126709,
+      "learning_rate": 1.398026315789474e-06,
+      "loss": 1.4562,
+      "step": 20358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7299435138702393,
+      "learning_rate": 1.3971125730994153e-06,
+      "loss": 1.5563,
+      "step": 20359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4909541606903076,
+      "learning_rate": 1.396198830409357e-06,
+      "loss": 1.4358,
+      "step": 20360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.317575216293335,
+      "learning_rate": 1.3952850877192983e-06,
+      "loss": 1.4925,
+      "step": 20361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0311806201934814,
+      "learning_rate": 1.3943713450292399e-06,
+      "loss": 1.1846,
+      "step": 20362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.213408946990967,
+      "learning_rate": 1.3934576023391812e-06,
+      "loss": 1.2411,
+      "step": 20363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.526935338973999,
+      "learning_rate": 1.3925438596491228e-06,
+      "loss": 1.381,
+      "step": 20364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9804024696350098,
+      "learning_rate": 1.3916301169590644e-06,
+      "loss": 1.0014,
+      "step": 20365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.669010639190674,
+      "learning_rate": 1.390716374269006e-06,
+      "loss": 1.5781,
+      "step": 20366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0451416969299316,
+      "learning_rate": 1.3898026315789476e-06,
+      "loss": 1.5118,
+      "step": 20367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6631412506103516,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 1.1177,
+      "step": 20368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9873030185699463,
+      "learning_rate": 1.3879751461988305e-06,
+      "loss": 1.2775,
+      "step": 20369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5425493717193604,
+      "learning_rate": 1.3870614035087721e-06,
+      "loss": 1.3573,
+      "step": 20370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1201021671295166,
+      "learning_rate": 1.3861476608187135e-06,
+      "loss": 1.2852,
+      "step": 20371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9488964080810547,
+      "learning_rate": 1.385233918128655e-06,
+      "loss": 1.5442,
+      "step": 20372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.580979585647583,
+      "learning_rate": 1.3843201754385965e-06,
+      "loss": 1.3814,
+      "step": 20373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4175233840942383,
+      "learning_rate": 1.383406432748538e-06,
+      "loss": 1.5698,
+      "step": 20374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.32554292678833,
+      "learning_rate": 1.3824926900584794e-06,
+      "loss": 1.5535,
+      "step": 20375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.274559497833252,
+      "learning_rate": 1.3815789473684212e-06,
+      "loss": 1.1669,
+      "step": 20376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6012814044952393,
+      "learning_rate": 1.3806652046783628e-06,
+      "loss": 1.3944,
+      "step": 20377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2393901348114014,
+      "learning_rate": 1.3797514619883042e-06,
+      "loss": 1.2163,
+      "step": 20378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.954845666885376,
+      "learning_rate": 1.3788377192982458e-06,
+      "loss": 1.6661,
+      "step": 20379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7066259384155273,
+      "learning_rate": 1.3779239766081874e-06,
+      "loss": 1.3188,
+      "step": 20380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.208721876144409,
+      "learning_rate": 1.3770102339181287e-06,
+      "loss": 1.0979,
+      "step": 20381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.574756622314453,
+      "learning_rate": 1.3760964912280703e-06,
+      "loss": 1.6908,
+      "step": 20382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2751224040985107,
+      "learning_rate": 1.3751827485380117e-06,
+      "loss": 1.5107,
+      "step": 20383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.094106674194336,
+      "learning_rate": 1.3742690058479533e-06,
+      "loss": 1.6083,
+      "step": 20384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5108025074005127,
+      "learning_rate": 1.3733552631578946e-06,
+      "loss": 1.3515,
+      "step": 20385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.347409725189209,
+      "learning_rate": 1.3724415204678362e-06,
+      "loss": 1.1966,
+      "step": 20386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6377274990081787,
+      "learning_rate": 1.371527777777778e-06,
+      "loss": 1.7255,
+      "step": 20387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.486924171447754,
+      "learning_rate": 1.3706140350877194e-06,
+      "loss": 1.3156,
+      "step": 20388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.034646511077881,
+      "learning_rate": 1.369700292397661e-06,
+      "loss": 1.829,
+      "step": 20389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6508889198303223,
+      "learning_rate": 1.3687865497076026e-06,
+      "loss": 1.6234,
+      "step": 20390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.366882801055908,
+      "learning_rate": 1.367872807017544e-06,
+      "loss": 1.248,
+      "step": 20391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2222888469696045,
+      "learning_rate": 1.3669590643274855e-06,
+      "loss": 1.4395,
+      "step": 20392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7689287662506104,
+      "learning_rate": 1.366045321637427e-06,
+      "loss": 1.5029,
+      "step": 20393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5177001953125,
+      "learning_rate": 1.3651315789473685e-06,
+      "loss": 1.2,
+      "step": 20394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5173113346099854,
+      "learning_rate": 1.3642178362573099e-06,
+      "loss": 1.5252,
+      "step": 20395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0262563228607178,
+      "learning_rate": 1.3633040935672515e-06,
+      "loss": 1.0386,
+      "step": 20396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9675161838531494,
+      "learning_rate": 1.362390350877193e-06,
+      "loss": 1.8083,
+      "step": 20397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.718512773513794,
+      "learning_rate": 1.3614766081871346e-06,
+      "loss": 1.9252,
+      "step": 20398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1696114540100098,
+      "learning_rate": 1.3605628654970762e-06,
+      "loss": 1.284,
+      "step": 20399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.21382474899292,
+      "learning_rate": 1.3596491228070178e-06,
+      "loss": 1.3074,
+      "step": 20400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9950833320617676,
+      "learning_rate": 1.3587353801169592e-06,
+      "loss": 1.1926,
+      "step": 20401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1933529376983643,
+      "learning_rate": 1.3578216374269008e-06,
+      "loss": 1.2928,
+      "step": 20402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9036195278167725,
+      "learning_rate": 1.3569078947368421e-06,
+      "loss": 1.4528,
+      "step": 20403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.38588547706604,
+      "learning_rate": 1.3559941520467837e-06,
+      "loss": 1.1901,
+      "step": 20404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2524337768554688,
+      "learning_rate": 1.355080409356725e-06,
+      "loss": 1.4154,
+      "step": 20405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3387022018432617,
+      "learning_rate": 1.3541666666666667e-06,
+      "loss": 1.1419,
+      "step": 20406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6223580837249756,
+      "learning_rate": 1.3532529239766083e-06,
+      "loss": 1.5493,
+      "step": 20407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.097123146057129,
+      "learning_rate": 1.3523391812865499e-06,
+      "loss": 1.3471,
+      "step": 20408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8171284198760986,
+      "learning_rate": 1.3514254385964914e-06,
+      "loss": 1.6174,
+      "step": 20409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0396900177001953,
+      "learning_rate": 1.350511695906433e-06,
+      "loss": 0.8986,
+      "step": 20410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9967868328094482,
+      "learning_rate": 1.3495979532163744e-06,
+      "loss": 1.6265,
+      "step": 20411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3599181175231934,
+      "learning_rate": 1.348684210526316e-06,
+      "loss": 1.2782,
+      "step": 20412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3351542949676514,
+      "learning_rate": 1.3477704678362574e-06,
+      "loss": 1.1068,
+      "step": 20413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.771440029144287,
+      "learning_rate": 1.346856725146199e-06,
+      "loss": 1.3834,
+      "step": 20414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1454451084136963,
+      "learning_rate": 1.3459429824561403e-06,
+      "loss": 1.3359,
+      "step": 20415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.310683250427246,
+      "learning_rate": 1.345029239766082e-06,
+      "loss": 1.2908,
+      "step": 20416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.307332992553711,
+      "learning_rate": 1.3441154970760235e-06,
+      "loss": 1.5217,
+      "step": 20417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4207615852355957,
+      "learning_rate": 1.3432017543859649e-06,
+      "loss": 1.4629,
+      "step": 20418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1897530555725098,
+      "learning_rate": 1.3422880116959067e-06,
+      "loss": 1.3555,
+      "step": 20419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.438472032546997,
+      "learning_rate": 1.3413742690058483e-06,
+      "loss": 1.3227,
+      "step": 20420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.527651071548462,
+      "learning_rate": 1.3404605263157896e-06,
+      "loss": 1.5604,
+      "step": 20421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7083427906036377,
+      "learning_rate": 1.3395467836257312e-06,
+      "loss": 1.5928,
+      "step": 20422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.668677806854248,
+      "learning_rate": 1.3386330409356726e-06,
+      "loss": 1.5738,
+      "step": 20423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.335127353668213,
+      "learning_rate": 1.3377192982456142e-06,
+      "loss": 1.3177,
+      "step": 20424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1988673210144043,
+      "learning_rate": 1.3368055555555556e-06,
+      "loss": 1.2599,
+      "step": 20425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5123729705810547,
+      "learning_rate": 1.3358918128654971e-06,
+      "loss": 1.5519,
+      "step": 20426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3226137161254883,
+      "learning_rate": 1.3349780701754387e-06,
+      "loss": 1.5594,
+      "step": 20427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0855960845947266,
+      "learning_rate": 1.33406432748538e-06,
+      "loss": 1.1894,
+      "step": 20428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1722395420074463,
+      "learning_rate": 1.3331505847953217e-06,
+      "loss": 1.3181,
+      "step": 20429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3453195095062256,
+      "learning_rate": 1.3322368421052635e-06,
+      "loss": 1.4288,
+      "step": 20430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.937753915786743,
+      "learning_rate": 1.3313230994152049e-06,
+      "loss": 1.5076,
+      "step": 20431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9495298862457275,
+      "learning_rate": 1.3304093567251464e-06,
+      "loss": 1.0443,
+      "step": 20432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.24118971824646,
+      "learning_rate": 1.3294956140350878e-06,
+      "loss": 1.6876,
+      "step": 20433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.366758108139038,
+      "learning_rate": 1.3285818713450294e-06,
+      "loss": 1.3493,
+      "step": 20434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.205444097518921,
+      "learning_rate": 1.3276681286549708e-06,
+      "loss": 1.7204,
+      "step": 20435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7896764278411865,
+      "learning_rate": 1.3267543859649124e-06,
+      "loss": 1.6729,
+      "step": 20436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5018470287323,
+      "learning_rate": 1.3258406432748537e-06,
+      "loss": 1.2632,
+      "step": 20437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.499729871749878,
+      "learning_rate": 1.3249269005847953e-06,
+      "loss": 1.3317,
+      "step": 20438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7659740447998047,
+      "learning_rate": 1.324013157894737e-06,
+      "loss": 1.6064,
+      "step": 20439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5400524139404297,
+      "learning_rate": 1.3230994152046785e-06,
+      "loss": 1.4512,
+      "step": 20440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.506152868270874,
+      "learning_rate": 1.32218567251462e-06,
+      "loss": 1.3959,
+      "step": 20441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4312024116516113,
+      "learning_rate": 1.3212719298245617e-06,
+      "loss": 1.6904,
+      "step": 20442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6990537643432617,
+      "learning_rate": 1.320358187134503e-06,
+      "loss": 1.2341,
+      "step": 20443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1806864738464355,
+      "learning_rate": 1.3194444444444446e-06,
+      "loss": 1.1368,
+      "step": 20444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.895411252975464,
+      "learning_rate": 1.318530701754386e-06,
+      "loss": 1.8563,
+      "step": 20445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7521071434020996,
+      "learning_rate": 1.3176169590643276e-06,
+      "loss": 1.3159,
+      "step": 20446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.417938470840454,
+      "learning_rate": 1.316703216374269e-06,
+      "loss": 1.1618,
+      "step": 20447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.55582332611084,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 1.4603,
+      "step": 20448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.097808361053467,
+      "learning_rate": 1.3148757309941521e-06,
+      "loss": 1.2527,
+      "step": 20449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5481927394866943,
+      "learning_rate": 1.3139619883040935e-06,
+      "loss": 1.5262,
+      "step": 20450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.376199722290039,
+      "learning_rate": 1.3130482456140353e-06,
+      "loss": 1.3979,
+      "step": 20451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.351835250854492,
+      "learning_rate": 1.312134502923977e-06,
+      "loss": 1.8716,
+      "step": 20452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.528697967529297,
+      "learning_rate": 1.3112207602339183e-06,
+      "loss": 1.4926,
+      "step": 20453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2511637210845947,
+      "learning_rate": 1.3103070175438599e-06,
+      "loss": 1.346,
+      "step": 20454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9924795627593994,
+      "learning_rate": 1.3093932748538012e-06,
+      "loss": 1.5335,
+      "step": 20455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7216238975524902,
+      "learning_rate": 1.3084795321637428e-06,
+      "loss": 1.5233,
+      "step": 20456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9931771755218506,
+      "learning_rate": 1.3075657894736842e-06,
+      "loss": 1.5223,
+      "step": 20457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1697871685028076,
+      "learning_rate": 1.3066520467836258e-06,
+      "loss": 1.3981,
+      "step": 20458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3213911056518555,
+      "learning_rate": 1.3057383040935674e-06,
+      "loss": 1.3572,
+      "step": 20459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6268484592437744,
+      "learning_rate": 1.3048245614035087e-06,
+      "loss": 1.6878,
+      "step": 20460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9737799167633057,
+      "learning_rate": 1.3039108187134503e-06,
+      "loss": 1.8929,
+      "step": 20461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5433831214904785,
+      "learning_rate": 1.3029970760233921e-06,
+      "loss": 1.1918,
+      "step": 20462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2461414337158203,
+      "learning_rate": 1.3020833333333335e-06,
+      "loss": 1.2032,
+      "step": 20463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.648465394973755,
+      "learning_rate": 1.301169590643275e-06,
+      "loss": 1.4636,
+      "step": 20464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.730459213256836,
+      "learning_rate": 1.3002558479532165e-06,
+      "loss": 1.6321,
+      "step": 20465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3812642097473145,
+      "learning_rate": 1.299342105263158e-06,
+      "loss": 1.3987,
+      "step": 20466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.201781988143921,
+      "learning_rate": 1.2984283625730994e-06,
+      "loss": 1.4737,
+      "step": 20467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.433788299560547,
+      "learning_rate": 1.297514619883041e-06,
+      "loss": 1.584,
+      "step": 20468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5507028102874756,
+      "learning_rate": 1.2966008771929826e-06,
+      "loss": 1.3568,
+      "step": 20469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5358994007110596,
+      "learning_rate": 1.295687134502924e-06,
+      "loss": 1.619,
+      "step": 20470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.501600980758667,
+      "learning_rate": 1.2947733918128656e-06,
+      "loss": 1.3243,
+      "step": 20471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5456109046936035,
+      "learning_rate": 1.2938596491228074e-06,
+      "loss": 1.3121,
+      "step": 20472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6038286685943604,
+      "learning_rate": 1.2929459064327487e-06,
+      "loss": 1.2559,
+      "step": 20473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3527183532714844,
+      "learning_rate": 1.2920321637426903e-06,
+      "loss": 1.4533,
+      "step": 20474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.342407703399658,
+      "learning_rate": 1.2911184210526317e-06,
+      "loss": 1.8304,
+      "step": 20475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1573143005371094,
+      "learning_rate": 1.2902046783625733e-06,
+      "loss": 1.1568,
+      "step": 20476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0128097534179688,
+      "learning_rate": 1.2892909356725146e-06,
+      "loss": 1.2128,
+      "step": 20477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.395376443862915,
+      "learning_rate": 1.2883771929824562e-06,
+      "loss": 1.5201,
+      "step": 20478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.177694082260132,
+      "learning_rate": 1.2874634502923978e-06,
+      "loss": 1.8332,
+      "step": 20479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.602703332901001,
+      "learning_rate": 1.2865497076023392e-06,
+      "loss": 1.4825,
+      "step": 20480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.283766269683838,
+      "learning_rate": 1.2856359649122808e-06,
+      "loss": 1.2,
+      "step": 20481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4336140155792236,
+      "learning_rate": 1.2847222222222222e-06,
+      "loss": 1.2581,
+      "step": 20482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4013333320617676,
+      "learning_rate": 1.283808479532164e-06,
+      "loss": 1.3299,
+      "step": 20483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.743932008743286,
+      "learning_rate": 1.2828947368421055e-06,
+      "loss": 1.2665,
+      "step": 20484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.038966178894043,
+      "learning_rate": 1.281980994152047e-06,
+      "loss": 1.2021,
+      "step": 20485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2183640003204346,
+      "learning_rate": 1.2810672514619885e-06,
+      "loss": 1.4978,
+      "step": 20486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5293445587158203,
+      "learning_rate": 1.2801535087719299e-06,
+      "loss": 1.6892,
+      "step": 20487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1548726558685303,
+      "learning_rate": 1.2792397660818715e-06,
+      "loss": 1.3993,
+      "step": 20488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.776958703994751,
+      "learning_rate": 1.278326023391813e-06,
+      "loss": 1.5581,
+      "step": 20489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.209815263748169,
+      "learning_rate": 1.2774122807017544e-06,
+      "loss": 1.174,
+      "step": 20490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.519644260406494,
+      "learning_rate": 1.276498538011696e-06,
+      "loss": 1.5966,
+      "step": 20491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.148526191711426,
+      "learning_rate": 1.2755847953216374e-06,
+      "loss": 1.7822,
+      "step": 20492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4258534908294678,
+      "learning_rate": 1.274671052631579e-06,
+      "loss": 1.2752,
+      "step": 20493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.308022975921631,
+      "learning_rate": 1.2737573099415208e-06,
+      "loss": 1.4627,
+      "step": 20494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.758606195449829,
+      "learning_rate": 1.2728435672514621e-06,
+      "loss": 1.5837,
+      "step": 20495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6248245239257812,
+      "learning_rate": 1.2719298245614037e-06,
+      "loss": 1.3702,
+      "step": 20496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8075666427612305,
+      "learning_rate": 1.271016081871345e-06,
+      "loss": 1.4762,
+      "step": 20497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.005653142929077,
+      "learning_rate": 1.2701023391812867e-06,
+      "loss": 1.2465,
+      "step": 20498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.446826219558716,
+      "learning_rate": 1.269188596491228e-06,
+      "loss": 1.3793,
+      "step": 20499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8088619709014893,
+      "learning_rate": 1.2682748538011696e-06,
+      "loss": 1.615,
+      "step": 20500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1542365550994873,
+      "learning_rate": 1.2673611111111112e-06,
+      "loss": 1.4799,
+      "step": 20501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.907142162322998,
+      "learning_rate": 1.2664473684210526e-06,
+      "loss": 1.3889,
+      "step": 20502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.188117742538452,
+      "learning_rate": 1.2655336257309942e-06,
+      "loss": 1.2285,
+      "step": 20503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1984715461730957,
+      "learning_rate": 1.264619883040936e-06,
+      "loss": 1.1605,
+      "step": 20504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.420891523361206,
+      "learning_rate": 1.2637061403508774e-06,
+      "loss": 1.4075,
+      "step": 20505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.467350721359253,
+      "learning_rate": 1.262792397660819e-06,
+      "loss": 1.6284,
+      "step": 20506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.146219253540039,
+      "learning_rate": 1.2618786549707603e-06,
+      "loss": 1.4225,
+      "step": 20507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.427487373352051,
+      "learning_rate": 1.260964912280702e-06,
+      "loss": 1.372,
+      "step": 20508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8868377208709717,
+      "learning_rate": 1.2600511695906433e-06,
+      "loss": 1.6968,
+      "step": 20509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.185018539428711,
+      "learning_rate": 1.2591374269005849e-06,
+      "loss": 1.2165,
+      "step": 20510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.488445520401001,
+      "learning_rate": 1.2582236842105265e-06,
+      "loss": 1.5454,
+      "step": 20511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3167834281921387,
+      "learning_rate": 1.2573099415204678e-06,
+      "loss": 1.2962,
+      "step": 20512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.353905439376831,
+      "learning_rate": 1.2563961988304094e-06,
+      "loss": 1.5792,
+      "step": 20513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.171010732650757,
+      "learning_rate": 1.2554824561403508e-06,
+      "loss": 1.6854,
+      "step": 20514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.247546672821045,
+      "learning_rate": 1.2545687134502926e-06,
+      "loss": 1.1846,
+      "step": 20515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8902010917663574,
+      "learning_rate": 1.2536549707602342e-06,
+      "loss": 1.0626,
+      "step": 20516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9161455631256104,
+      "learning_rate": 1.2527412280701756e-06,
+      "loss": 1.8587,
+      "step": 20517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.328061819076538,
+      "learning_rate": 1.2518274853801171e-06,
+      "loss": 1.3886,
+      "step": 20518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.360795497894287,
+      "learning_rate": 1.2509137426900585e-06,
+      "loss": 1.3791,
+      "step": 20519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6596105098724365,
+      "learning_rate": 1.25e-06,
+      "loss": 1.5381,
+      "step": 20520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3827943801879883,
+      "learning_rate": 1.2490862573099417e-06,
+      "loss": 1.5058,
+      "step": 20521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2948009967803955,
+      "learning_rate": 1.248172514619883e-06,
+      "loss": 1.2413,
+      "step": 20522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0129969120025635,
+      "learning_rate": 1.2472587719298246e-06,
+      "loss": 2.03,
+      "step": 20523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.420844554901123,
+      "learning_rate": 1.2463450292397662e-06,
+      "loss": 1.3617,
+      "step": 20524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.240432024002075,
+      "learning_rate": 1.2454312865497076e-06,
+      "loss": 1.2407,
+      "step": 20525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9051593542099,
+      "learning_rate": 1.2445175438596492e-06,
+      "loss": 1.0619,
+      "step": 20526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3058812618255615,
+      "learning_rate": 1.2436038011695908e-06,
+      "loss": 1.3405,
+      "step": 20527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.608487367630005,
+      "learning_rate": 1.2426900584795322e-06,
+      "loss": 1.4845,
+      "step": 20528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6248843669891357,
+      "learning_rate": 1.2417763157894737e-06,
+      "loss": 1.3397,
+      "step": 20529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9370810985565186,
+      "learning_rate": 1.2408625730994153e-06,
+      "loss": 1.534,
+      "step": 20530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.007347583770752,
+      "learning_rate": 1.239948830409357e-06,
+      "loss": 1.6159,
+      "step": 20531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7059433460235596,
+      "learning_rate": 1.2390350877192983e-06,
+      "loss": 1.4032,
+      "step": 20532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2781736850738525,
+      "learning_rate": 1.2381213450292399e-06,
+      "loss": 1.3085,
+      "step": 20533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.695737600326538,
+      "learning_rate": 1.2372076023391815e-06,
+      "loss": 1.5398,
+      "step": 20534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4081690311431885,
+      "learning_rate": 1.2362938596491228e-06,
+      "loss": 1.4087,
+      "step": 20535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.459916830062866,
+      "learning_rate": 1.2353801169590644e-06,
+      "loss": 1.4334,
+      "step": 20536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5820722579956055,
+      "learning_rate": 1.234466374269006e-06,
+      "loss": 1.6189,
+      "step": 20537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2944111824035645,
+      "learning_rate": 1.2335526315789474e-06,
+      "loss": 1.2162,
+      "step": 20538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5086958408355713,
+      "learning_rate": 1.232638888888889e-06,
+      "loss": 1.4306,
+      "step": 20539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5175180435180664,
+      "learning_rate": 1.2317251461988306e-06,
+      "loss": 1.3626,
+      "step": 20540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6313352584838867,
+      "learning_rate": 1.2308114035087721e-06,
+      "loss": 1.6984,
+      "step": 20541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.479443073272705,
+      "learning_rate": 1.2298976608187135e-06,
+      "loss": 1.3991,
+      "step": 20542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3644425868988037,
+      "learning_rate": 1.228983918128655e-06,
+      "loss": 1.5063,
+      "step": 20543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9104387760162354,
+      "learning_rate": 1.2280701754385965e-06,
+      "loss": 1.4882,
+      "step": 20544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.775334119796753,
+      "learning_rate": 1.227156432748538e-06,
+      "loss": 1.8265,
+      "step": 20545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2642171382904053,
+      "learning_rate": 1.2262426900584796e-06,
+      "loss": 1.3942,
+      "step": 20546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.758932113647461,
+      "learning_rate": 1.2253289473684212e-06,
+      "loss": 1.4175,
+      "step": 20547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.455190658569336,
+      "learning_rate": 1.2244152046783626e-06,
+      "loss": 1.3101,
+      "step": 20548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4386956691741943,
+      "learning_rate": 1.2235014619883042e-06,
+      "loss": 1.4088,
+      "step": 20549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3692357540130615,
+      "learning_rate": 1.2225877192982458e-06,
+      "loss": 1.4664,
+      "step": 20550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.501002788543701,
+      "learning_rate": 1.2216739766081872e-06,
+      "loss": 1.3729,
+      "step": 20551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.371396780014038,
+      "learning_rate": 1.2207602339181287e-06,
+      "loss": 1.7577,
+      "step": 20552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6445090770721436,
+      "learning_rate": 1.2198464912280703e-06,
+      "loss": 1.5115,
+      "step": 20553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.738304615020752,
+      "learning_rate": 1.2189327485380117e-06,
+      "loss": 1.8303,
+      "step": 20554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2957630157470703,
+      "learning_rate": 1.2180190058479533e-06,
+      "loss": 1.3204,
+      "step": 20555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.211815118789673,
+      "learning_rate": 1.2171052631578949e-06,
+      "loss": 1.6494,
+      "step": 20556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7024781703948975,
+      "learning_rate": 1.2161915204678365e-06,
+      "loss": 1.3222,
+      "step": 20557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 48.00735855102539,
+      "learning_rate": 1.2152777777777778e-06,
+      "loss": 1.2378,
+      "step": 20558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6484904289245605,
+      "learning_rate": 1.2143640350877194e-06,
+      "loss": 1.145,
+      "step": 20559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5893590450286865,
+      "learning_rate": 1.2134502923976608e-06,
+      "loss": 1.4557,
+      "step": 20560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5894062519073486,
+      "learning_rate": 1.2125365497076024e-06,
+      "loss": 1.46,
+      "step": 20561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1907081604003906,
+      "learning_rate": 1.211622807017544e-06,
+      "loss": 1.2582,
+      "step": 20562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.185070276260376,
+      "learning_rate": 1.2107090643274856e-06,
+      "loss": 1.3227,
+      "step": 20563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.542738914489746,
+      "learning_rate": 1.209795321637427e-06,
+      "loss": 1.4675,
+      "step": 20564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4736220836639404,
+      "learning_rate": 1.2088815789473685e-06,
+      "loss": 1.3023,
+      "step": 20565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.437798500061035,
+      "learning_rate": 1.20796783625731e-06,
+      "loss": 1.3572,
+      "step": 20566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1873271465301514,
+      "learning_rate": 1.2070540935672517e-06,
+      "loss": 1.6352,
+      "step": 20567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3035285472869873,
+      "learning_rate": 1.206140350877193e-06,
+      "loss": 1.2907,
+      "step": 20568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2531657218933105,
+      "learning_rate": 1.2052266081871346e-06,
+      "loss": 1.9984,
+      "step": 20569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.042837381362915,
+      "learning_rate": 1.204312865497076e-06,
+      "loss": 1.1756,
+      "step": 20570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.373091220855713,
+      "learning_rate": 1.2033991228070176e-06,
+      "loss": 1.2307,
+      "step": 20571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0790419578552246,
+      "learning_rate": 1.2024853801169592e-06,
+      "loss": 1.2898,
+      "step": 20572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.252312183380127,
+      "learning_rate": 1.2015716374269008e-06,
+      "loss": 1.2751,
+      "step": 20573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.203254461288452,
+      "learning_rate": 1.2006578947368422e-06,
+      "loss": 1.3571,
+      "step": 20574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6765687465667725,
+      "learning_rate": 1.1997441520467837e-06,
+      "loss": 1.4482,
+      "step": 20575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.547577142715454,
+      "learning_rate": 1.1988304093567251e-06,
+      "loss": 1.3567,
+      "step": 20576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4459636211395264,
+      "learning_rate": 1.197916666666667e-06,
+      "loss": 1.0214,
+      "step": 20577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4606544971466064,
+      "learning_rate": 1.1970029239766083e-06,
+      "loss": 1.3862,
+      "step": 20578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.363001585006714,
+      "learning_rate": 1.1960891812865499e-06,
+      "loss": 1.445,
+      "step": 20579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.554027795791626,
+      "learning_rate": 1.1951754385964912e-06,
+      "loss": 1.6352,
+      "step": 20580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5958003997802734,
+      "learning_rate": 1.1942616959064328e-06,
+      "loss": 1.5607,
+      "step": 20581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7512171268463135,
+      "learning_rate": 1.1933479532163744e-06,
+      "loss": 1.4703,
+      "step": 20582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.918109655380249,
+      "learning_rate": 1.192434210526316e-06,
+      "loss": 1.6427,
+      "step": 20583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.289907217025757,
+      "learning_rate": 1.1915204678362574e-06,
+      "loss": 1.257,
+      "step": 20584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2759416103363037,
+      "learning_rate": 1.190606725146199e-06,
+      "loss": 1.4701,
+      "step": 20585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7467153072357178,
+      "learning_rate": 1.1896929824561403e-06,
+      "loss": 1.7009,
+      "step": 20586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.439542055130005,
+      "learning_rate": 1.188779239766082e-06,
+      "loss": 1.51,
+      "step": 20587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8947417736053467,
+      "learning_rate": 1.1878654970760235e-06,
+      "loss": 1.3517,
+      "step": 20588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0400757789611816,
+      "learning_rate": 1.186951754385965e-06,
+      "loss": 1.0999,
+      "step": 20589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.064190149307251,
+      "learning_rate": 1.1860380116959065e-06,
+      "loss": 1.168,
+      "step": 20590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4116153717041016,
+      "learning_rate": 1.185124269005848e-06,
+      "loss": 1.5858,
+      "step": 20591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.074772596359253,
+      "learning_rate": 1.1842105263157894e-06,
+      "loss": 1.0612,
+      "step": 20592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8480398654937744,
+      "learning_rate": 1.1832967836257312e-06,
+      "loss": 1.7033,
+      "step": 20593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6363950967788696,
+      "learning_rate": 1.1823830409356726e-06,
+      "loss": 0.9411,
+      "step": 20594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.351238489151001,
+      "learning_rate": 1.1814692982456142e-06,
+      "loss": 1.495,
+      "step": 20595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.186088800430298,
+      "learning_rate": 1.1805555555555556e-06,
+      "loss": 1.0981,
+      "step": 20596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2587428092956543,
+      "learning_rate": 1.1796418128654972e-06,
+      "loss": 1.2205,
+      "step": 20597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.222618341445923,
+      "learning_rate": 1.1787280701754387e-06,
+      "loss": 1.2447,
+      "step": 20598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.837827444076538,
+      "learning_rate": 1.1778143274853803e-06,
+      "loss": 1.4792,
+      "step": 20599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1929194927215576,
+      "learning_rate": 1.1769005847953217e-06,
+      "loss": 1.6122,
+      "step": 20600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.016967296600342,
+      "learning_rate": 1.1759868421052633e-06,
+      "loss": 1.4703,
+      "step": 20601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2252352237701416,
+      "learning_rate": 1.1750730994152047e-06,
+      "loss": 1.1781,
+      "step": 20602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.306302547454834,
+      "learning_rate": 1.1741593567251465e-06,
+      "loss": 1.3865,
+      "step": 20603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7418434619903564,
+      "learning_rate": 1.1732456140350878e-06,
+      "loss": 1.5036,
+      "step": 20604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3791191577911377,
+      "learning_rate": 1.1723318713450294e-06,
+      "loss": 1.475,
+      "step": 20605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6270976066589355,
+      "learning_rate": 1.1714181286549708e-06,
+      "loss": 2.0306,
+      "step": 20606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2887392044067383,
+      "learning_rate": 1.1705043859649124e-06,
+      "loss": 1.409,
+      "step": 20607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0978240966796875,
+      "learning_rate": 1.1695906432748538e-06,
+      "loss": 0.9988,
+      "step": 20608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2707009315490723,
+      "learning_rate": 1.1686769005847956e-06,
+      "loss": 1.6042,
+      "step": 20609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.947720527648926,
+      "learning_rate": 1.167763157894737e-06,
+      "loss": 1.3227,
+      "step": 20610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8150079250335693,
+      "learning_rate": 1.1668494152046785e-06,
+      "loss": 1.5237,
+      "step": 20611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8121362924575806,
+      "learning_rate": 1.1659356725146199e-06,
+      "loss": 0.9876,
+      "step": 20612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2599430084228516,
+      "learning_rate": 1.1650219298245615e-06,
+      "loss": 1.2085,
+      "step": 20613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.422480344772339,
+      "learning_rate": 1.164108187134503e-06,
+      "loss": 1.4495,
+      "step": 20614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.469738721847534,
+      "learning_rate": 1.1631944444444446e-06,
+      "loss": 1.1799,
+      "step": 20615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.285856246948242,
+      "learning_rate": 1.162280701754386e-06,
+      "loss": 1.1861,
+      "step": 20616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.604724884033203,
+      "learning_rate": 1.1613669590643276e-06,
+      "loss": 1.2842,
+      "step": 20617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.988171339035034,
+      "learning_rate": 1.160453216374269e-06,
+      "loss": 1.5095,
+      "step": 20618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2768514156341553,
+      "learning_rate": 1.1595394736842108e-06,
+      "loss": 1.0314,
+      "step": 20619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9616076946258545,
+      "learning_rate": 1.1586257309941522e-06,
+      "loss": 1.2789,
+      "step": 20620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.442629337310791,
+      "learning_rate": 1.1577119883040937e-06,
+      "loss": 1.5799,
+      "step": 20621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3574471473693848,
+      "learning_rate": 1.1567982456140351e-06,
+      "loss": 1.1733,
+      "step": 20622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.385429859161377,
+      "learning_rate": 1.1558845029239767e-06,
+      "loss": 1.378,
+      "step": 20623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.422384023666382,
+      "learning_rate": 1.154970760233918e-06,
+      "loss": 1.6267,
+      "step": 20624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0079472064971924,
+      "learning_rate": 1.1540570175438599e-06,
+      "loss": 1.8319,
+      "step": 20625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.829061508178711,
+      "learning_rate": 1.1531432748538012e-06,
+      "loss": 1.5663,
+      "step": 20626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8319029808044434,
+      "learning_rate": 1.1522295321637428e-06,
+      "loss": 1.3664,
+      "step": 20627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.081594228744507,
+      "learning_rate": 1.1513157894736842e-06,
+      "loss": 2.1871,
+      "step": 20628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0331242084503174,
+      "learning_rate": 1.1504020467836258e-06,
+      "loss": 1.1861,
+      "step": 20629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9161314964294434,
+      "learning_rate": 1.1494883040935674e-06,
+      "loss": 1.6229,
+      "step": 20630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.644045114517212,
+      "learning_rate": 1.148574561403509e-06,
+      "loss": 1.6068,
+      "step": 20631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.962050437927246,
+      "learning_rate": 1.1476608187134503e-06,
+      "loss": 1.1953,
+      "step": 20632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3015353679656982,
+      "learning_rate": 1.146747076023392e-06,
+      "loss": 1.3686,
+      "step": 20633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.177790403366089,
+      "learning_rate": 1.1458333333333333e-06,
+      "loss": 1.5709,
+      "step": 20634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.404043674468994,
+      "learning_rate": 1.144919590643275e-06,
+      "loss": 1.4454,
+      "step": 20635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.579623222351074,
+      "learning_rate": 1.1440058479532165e-06,
+      "loss": 1.4311,
+      "step": 20636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.489290475845337,
+      "learning_rate": 1.143092105263158e-06,
+      "loss": 1.5337,
+      "step": 20637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4669597148895264,
+      "learning_rate": 1.1421783625730994e-06,
+      "loss": 1.3878,
+      "step": 20638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2566263675689697,
+      "learning_rate": 1.141264619883041e-06,
+      "loss": 1.2143,
+      "step": 20639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.687297821044922,
+      "learning_rate": 1.1403508771929824e-06,
+      "loss": 1.7491,
+      "step": 20640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9578279256820679,
+      "learning_rate": 1.1394371345029242e-06,
+      "loss": 1.1556,
+      "step": 20641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5865962505340576,
+      "learning_rate": 1.1385233918128656e-06,
+      "loss": 1.7111,
+      "step": 20642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.416862964630127,
+      "learning_rate": 1.1376096491228072e-06,
+      "loss": 1.4067,
+      "step": 20643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.285799026489258,
+      "learning_rate": 1.1366959064327485e-06,
+      "loss": 1.2877,
+      "step": 20644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4172050952911377,
+      "learning_rate": 1.1357821637426901e-06,
+      "loss": 1.0656,
+      "step": 20645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7401621341705322,
+      "learning_rate": 1.1348684210526317e-06,
+      "loss": 1.4967,
+      "step": 20646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.062432050704956,
+      "learning_rate": 1.1339546783625733e-06,
+      "loss": 1.8557,
+      "step": 20647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.775646924972534,
+      "learning_rate": 1.1330409356725147e-06,
+      "loss": 1.6423,
+      "step": 20648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9845592975616455,
+      "learning_rate": 1.1321271929824562e-06,
+      "loss": 1.7623,
+      "step": 20649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.332710027694702,
+      "learning_rate": 1.1312134502923976e-06,
+      "loss": 1.3714,
+      "step": 20650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5234384536743164,
+      "learning_rate": 1.1302997076023394e-06,
+      "loss": 1.3706,
+      "step": 20651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5950772762298584,
+      "learning_rate": 1.1293859649122808e-06,
+      "loss": 1.3205,
+      "step": 20652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9759702682495117,
+      "learning_rate": 1.1284722222222224e-06,
+      "loss": 1.2851,
+      "step": 20653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3827850818634033,
+      "learning_rate": 1.1275584795321638e-06,
+      "loss": 1.3133,
+      "step": 20654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.306647539138794,
+      "learning_rate": 1.1266447368421053e-06,
+      "loss": 1.4439,
+      "step": 20655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6772258281707764,
+      "learning_rate": 1.1257309941520467e-06,
+      "loss": 1.2326,
+      "step": 20656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.91418719291687,
+      "learning_rate": 1.1248172514619885e-06,
+      "loss": 1.7194,
+      "step": 20657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1816611289978027,
+      "learning_rate": 1.1239035087719299e-06,
+      "loss": 1.1535,
+      "step": 20658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.195249557495117,
+      "learning_rate": 1.1229897660818715e-06,
+      "loss": 1.2663,
+      "step": 20659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.993861198425293,
+      "learning_rate": 1.1220760233918129e-06,
+      "loss": 1.2621,
+      "step": 20660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1331379413604736,
+      "learning_rate": 1.1211622807017544e-06,
+      "loss": 1.1015,
+      "step": 20661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4895496368408203,
+      "learning_rate": 1.120248538011696e-06,
+      "loss": 1.2346,
+      "step": 20662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.295715570449829,
+      "learning_rate": 1.1193347953216376e-06,
+      "loss": 1.3724,
+      "step": 20663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3045566082000732,
+      "learning_rate": 1.118421052631579e-06,
+      "loss": 1.6337,
+      "step": 20664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.906067132949829,
+      "learning_rate": 1.1175073099415206e-06,
+      "loss": 1.528,
+      "step": 20665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9051891565322876,
+      "learning_rate": 1.116593567251462e-06,
+      "loss": 1.2484,
+      "step": 20666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4302802085876465,
+      "learning_rate": 1.1156798245614037e-06,
+      "loss": 1.402,
+      "step": 20667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3100709915161133,
+      "learning_rate": 1.1147660818713451e-06,
+      "loss": 1.3628,
+      "step": 20668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5775299072265625,
+      "learning_rate": 1.1138523391812867e-06,
+      "loss": 1.6847,
+      "step": 20669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0583770275115967,
+      "learning_rate": 1.112938596491228e-06,
+      "loss": 1.2031,
+      "step": 20670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.94946551322937,
+      "learning_rate": 1.1120248538011697e-06,
+      "loss": 1.4494,
+      "step": 20671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2473530769348145,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.2222,
+      "step": 20672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3192341327667236,
+      "learning_rate": 1.1101973684210528e-06,
+      "loss": 1.5897,
+      "step": 20673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3556816577911377,
+      "learning_rate": 1.1092836257309942e-06,
+      "loss": 1.4158,
+      "step": 20674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2841386795043945,
+      "learning_rate": 1.1083698830409358e-06,
+      "loss": 1.3471,
+      "step": 20675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.641160249710083,
+      "learning_rate": 1.1074561403508772e-06,
+      "loss": 1.3786,
+      "step": 20676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7075722217559814,
+      "learning_rate": 1.1065423976608188e-06,
+      "loss": 1.3458,
+      "step": 20677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.254807949066162,
+      "learning_rate": 1.1056286549707603e-06,
+      "loss": 1.125,
+      "step": 20678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4778151512145996,
+      "learning_rate": 1.104714912280702e-06,
+      "loss": 1.1845,
+      "step": 20679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.505032777786255,
+      "learning_rate": 1.1038011695906433e-06,
+      "loss": 1.4417,
+      "step": 20680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8986238241195679,
+      "learning_rate": 1.1028874269005849e-06,
+      "loss": 1.1673,
+      "step": 20681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0239062309265137,
+      "learning_rate": 1.1019736842105263e-06,
+      "loss": 1.1114,
+      "step": 20682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8646364212036133,
+      "learning_rate": 1.101059941520468e-06,
+      "loss": 1.5148,
+      "step": 20683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.468082904815674,
+      "learning_rate": 1.1001461988304094e-06,
+      "loss": 1.3244,
+      "step": 20684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.605764865875244,
+      "learning_rate": 1.099232456140351e-06,
+      "loss": 1.4808,
+      "step": 20685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2967309951782227,
+      "learning_rate": 1.0983187134502924e-06,
+      "loss": 1.2711,
+      "step": 20686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.23276424407959,
+      "learning_rate": 1.097404970760234e-06,
+      "loss": 1.0918,
+      "step": 20687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4390034675598145,
+      "learning_rate": 1.0964912280701756e-06,
+      "loss": 1.4638,
+      "step": 20688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.846367835998535,
+      "learning_rate": 1.0955774853801172e-06,
+      "loss": 1.4864,
+      "step": 20689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.879781723022461,
+      "learning_rate": 1.0946637426900585e-06,
+      "loss": 1.7195,
+      "step": 20690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3553779125213623,
+      "learning_rate": 1.0937500000000001e-06,
+      "loss": 1.201,
+      "step": 20691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.504394292831421,
+      "learning_rate": 1.0928362573099415e-06,
+      "loss": 1.3692,
+      "step": 20692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2846381664276123,
+      "learning_rate": 1.091922514619883e-06,
+      "loss": 1.1093,
+      "step": 20693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.52193546295166,
+      "learning_rate": 1.0910087719298247e-06,
+      "loss": 1.6874,
+      "step": 20694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5976858139038086,
+      "learning_rate": 1.0900950292397662e-06,
+      "loss": 1.4243,
+      "step": 20695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.332655668258667,
+      "learning_rate": 1.0891812865497076e-06,
+      "loss": 1.3729,
+      "step": 20696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.689444065093994,
+      "learning_rate": 1.0882675438596492e-06,
+      "loss": 1.8235,
+      "step": 20697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6590240001678467,
+      "learning_rate": 1.0873538011695906e-06,
+      "loss": 1.7963,
+      "step": 20698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4367482662200928,
+      "learning_rate": 1.0864400584795322e-06,
+      "loss": 1.1212,
+      "step": 20699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.419933319091797,
+      "learning_rate": 1.0855263157894738e-06,
+      "loss": 1.3976,
+      "step": 20700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.398014783859253,
+      "learning_rate": 1.0846125730994153e-06,
+      "loss": 1.2432,
+      "step": 20701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.391129493713379,
+      "learning_rate": 1.0836988304093567e-06,
+      "loss": 1.2234,
+      "step": 20702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5045807361602783,
+      "learning_rate": 1.0827850877192983e-06,
+      "loss": 1.503,
+      "step": 20703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.245288610458374,
+      "learning_rate": 1.0818713450292399e-06,
+      "loss": 1.2246,
+      "step": 20704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3034768104553223,
+      "learning_rate": 1.0809576023391815e-06,
+      "loss": 1.3339,
+      "step": 20705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2722649574279785,
+      "learning_rate": 1.0800438596491228e-06,
+      "loss": 1.1713,
+      "step": 20706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0607645511627197,
+      "learning_rate": 1.0791301169590644e-06,
+      "loss": 1.2216,
+      "step": 20707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.463123321533203,
+      "learning_rate": 1.0782163742690058e-06,
+      "loss": 1.2741,
+      "step": 20708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.949946880340576,
+      "learning_rate": 1.0773026315789474e-06,
+      "loss": 1.3091,
+      "step": 20709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5465378761291504,
+      "learning_rate": 1.076388888888889e-06,
+      "loss": 1.5848,
+      "step": 20710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5021469593048096,
+      "learning_rate": 1.0754751461988306e-06,
+      "loss": 1.1741,
+      "step": 20711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7059388160705566,
+      "learning_rate": 1.074561403508772e-06,
+      "loss": 1.7032,
+      "step": 20712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.17314076423645,
+      "learning_rate": 1.0736476608187135e-06,
+      "loss": 0.9988,
+      "step": 20713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3305630683898926,
+      "learning_rate": 1.0727339181286551e-06,
+      "loss": 1.3866,
+      "step": 20714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.264719486236572,
+      "learning_rate": 1.0718201754385965e-06,
+      "loss": 1.5232,
+      "step": 20715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.301774024963379,
+      "learning_rate": 1.070906432748538e-06,
+      "loss": 1.3708,
+      "step": 20716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7144436836242676,
+      "learning_rate": 1.0699926900584797e-06,
+      "loss": 1.4712,
+      "step": 20717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6099822521209717,
+      "learning_rate": 1.069078947368421e-06,
+      "loss": 1.4473,
+      "step": 20718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1304032802581787,
+      "learning_rate": 1.0681652046783626e-06,
+      "loss": 1.1511,
+      "step": 20719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4278132915496826,
+      "learning_rate": 1.0672514619883042e-06,
+      "loss": 1.603,
+      "step": 20720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4203989505767822,
+      "learning_rate": 1.0663377192982458e-06,
+      "loss": 1.5342,
+      "step": 20721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.901585817337036,
+      "learning_rate": 1.0654239766081872e-06,
+      "loss": 1.6445,
+      "step": 20722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.485708236694336,
+      "learning_rate": 1.0645102339181288e-06,
+      "loss": 1.486,
+      "step": 20723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.821828603744507,
+      "learning_rate": 1.0635964912280703e-06,
+      "loss": 1.4803,
+      "step": 20724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4868764877319336,
+      "learning_rate": 1.0626827485380117e-06,
+      "loss": 1.3766,
+      "step": 20725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2343673706054688,
+      "learning_rate": 1.0617690058479533e-06,
+      "loss": 1.2683,
+      "step": 20726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.605725049972534,
+      "learning_rate": 1.0608552631578949e-06,
+      "loss": 1.6622,
+      "step": 20727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.098726272583008,
+      "learning_rate": 1.0599415204678363e-06,
+      "loss": 1.3025,
+      "step": 20728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2427783012390137,
+      "learning_rate": 1.0590277777777778e-06,
+      "loss": 1.2791,
+      "step": 20729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.20320463180542,
+      "learning_rate": 1.0581140350877194e-06,
+      "loss": 1.1509,
+      "step": 20730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.676135301589966,
+      "learning_rate": 1.0572002923976608e-06,
+      "loss": 1.3999,
+      "step": 20731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8676517009735107,
+      "learning_rate": 1.0562865497076024e-06,
+      "loss": 1.4614,
+      "step": 20732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.63041353225708,
+      "learning_rate": 1.055372807017544e-06,
+      "loss": 1.3874,
+      "step": 20733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.401336193084717,
+      "learning_rate": 1.0544590643274854e-06,
+      "loss": 1.2096,
+      "step": 20734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.101470947265625,
+      "learning_rate": 1.053545321637427e-06,
+      "loss": 1.0653,
+      "step": 20735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.608302354812622,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 1.4144,
+      "step": 20736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.902595043182373,
+      "learning_rate": 1.0517178362573101e-06,
+      "loss": 1.4587,
+      "step": 20737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4913406372070312,
+      "learning_rate": 1.0508040935672515e-06,
+      "loss": 1.4509,
+      "step": 20738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.855865716934204,
+      "learning_rate": 1.049890350877193e-06,
+      "loss": 1.7026,
+      "step": 20739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.975665807723999,
+      "learning_rate": 1.0489766081871347e-06,
+      "loss": 1.8467,
+      "step": 20740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.22654128074646,
+      "learning_rate": 1.048062865497076e-06,
+      "loss": 1.4239,
+      "step": 20741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.087409257888794,
+      "learning_rate": 1.0471491228070176e-06,
+      "loss": 1.2631,
+      "step": 20742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2454938888549805,
+      "learning_rate": 1.0462353801169592e-06,
+      "loss": 1.2854,
+      "step": 20743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6987972259521484,
+      "learning_rate": 1.0453216374269006e-06,
+      "loss": 1.6574,
+      "step": 20744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6468918323516846,
+      "learning_rate": 1.0444078947368422e-06,
+      "loss": 1.4739,
+      "step": 20745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7977025508880615,
+      "learning_rate": 1.0434941520467838e-06,
+      "loss": 1.4755,
+      "step": 20746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.112593173980713,
+      "learning_rate": 1.0425804093567251e-06,
+      "loss": 0.9696,
+      "step": 20747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.033491849899292,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 1.7682,
+      "step": 20748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.770364999771118,
+      "learning_rate": 1.0407529239766083e-06,
+      "loss": 1.22,
+      "step": 20749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.341764450073242,
+      "learning_rate": 1.0398391812865499e-06,
+      "loss": 1.2222,
+      "step": 20750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.187699556350708,
+      "learning_rate": 1.0389254385964913e-06,
+      "loss": 1.4811,
+      "step": 20751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4536640644073486,
+      "learning_rate": 1.0380116959064328e-06,
+      "loss": 1.559,
+      "step": 20752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.598027467727661,
+      "learning_rate": 1.0370979532163744e-06,
+      "loss": 1.4965,
+      "step": 20753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5482277870178223,
+      "learning_rate": 1.0361842105263158e-06,
+      "loss": 1.703,
+      "step": 20754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9263617992401123,
+      "learning_rate": 1.0352704678362574e-06,
+      "loss": 1.3492,
+      "step": 20755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.278836727142334,
+      "learning_rate": 1.034356725146199e-06,
+      "loss": 1.293,
+      "step": 20756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.803826093673706,
+      "learning_rate": 1.0334429824561404e-06,
+      "loss": 1.467,
+      "step": 20757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.474757432937622,
+      "learning_rate": 1.032529239766082e-06,
+      "loss": 1.7041,
+      "step": 20758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1292028427124023,
+      "learning_rate": 1.0316154970760235e-06,
+      "loss": 1.2806,
+      "step": 20759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2391130924224854,
+      "learning_rate": 1.030701754385965e-06,
+      "loss": 1.5104,
+      "step": 20760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.920952081680298,
+      "learning_rate": 1.0297880116959065e-06,
+      "loss": 1.4774,
+      "step": 20761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1042943000793457,
+      "learning_rate": 1.028874269005848e-06,
+      "loss": 1.6555,
+      "step": 20762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2021729946136475,
+      "learning_rate": 1.0279605263157895e-06,
+      "loss": 1.7408,
+      "step": 20763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.261610507965088,
+      "learning_rate": 1.027046783625731e-06,
+      "loss": 1.2924,
+      "step": 20764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.06192946434021,
+      "learning_rate": 1.0261330409356726e-06,
+      "loss": 1.2635,
+      "step": 20765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1520888805389404,
+      "learning_rate": 1.0252192982456142e-06,
+      "loss": 1.3245,
+      "step": 20766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3986098766326904,
+      "learning_rate": 1.0243055555555556e-06,
+      "loss": 1.3276,
+      "step": 20767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.236372947692871,
+      "learning_rate": 1.0233918128654972e-06,
+      "loss": 1.3368,
+      "step": 20768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.33914852142334,
+      "learning_rate": 1.0224780701754388e-06,
+      "loss": 1.1004,
+      "step": 20769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.962890863418579,
+      "learning_rate": 1.0215643274853801e-06,
+      "loss": 1.9866,
+      "step": 20770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.519780397415161,
+      "learning_rate": 1.0206505847953217e-06,
+      "loss": 1.3485,
+      "step": 20771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3831064701080322,
+      "learning_rate": 1.0197368421052633e-06,
+      "loss": 1.3975,
+      "step": 20772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0679779052734375,
+      "learning_rate": 1.0188230994152047e-06,
+      "loss": 1.7923,
+      "step": 20773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9959789514541626,
+      "learning_rate": 1.0179093567251463e-06,
+      "loss": 1.0797,
+      "step": 20774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6722872257232666,
+      "learning_rate": 1.0169956140350878e-06,
+      "loss": 1.6229,
+      "step": 20775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9917253255844116,
+      "learning_rate": 1.0160818713450294e-06,
+      "loss": 1.2355,
+      "step": 20776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9200704097747803,
+      "learning_rate": 1.0151681286549708e-06,
+      "loss": 1.5726,
+      "step": 20777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8858473300933838,
+      "learning_rate": 1.0142543859649124e-06,
+      "loss": 1.1323,
+      "step": 20778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.843437433242798,
+      "learning_rate": 1.0133406432748538e-06,
+      "loss": 1.5699,
+      "step": 20779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0854837894439697,
+      "learning_rate": 1.0124269005847954e-06,
+      "loss": 1.2594,
+      "step": 20780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5889899730682373,
+      "learning_rate": 1.011513157894737e-06,
+      "loss": 1.5477,
+      "step": 20781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.565394163131714,
+      "learning_rate": 1.0105994152046785e-06,
+      "loss": 1.6321,
+      "step": 20782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8395333290100098,
+      "learning_rate": 1.00968567251462e-06,
+      "loss": 1.5999,
+      "step": 20783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.967153310775757,
+      "learning_rate": 1.0087719298245615e-06,
+      "loss": 1.43,
+      "step": 20784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.597482919692993,
+      "learning_rate": 1.007858187134503e-06,
+      "loss": 1.673,
+      "step": 20785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.599855661392212,
+      "learning_rate": 1.0069444444444447e-06,
+      "loss": 1.395,
+      "step": 20786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4237213134765625,
+      "learning_rate": 1.006030701754386e-06,
+      "loss": 1.5336,
+      "step": 20787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1465084552764893,
+      "learning_rate": 1.0051169590643276e-06,
+      "loss": 1.2942,
+      "step": 20788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1272740364074707,
+      "learning_rate": 1.004203216374269e-06,
+      "loss": 1.2204,
+      "step": 20789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2558705806732178,
+      "learning_rate": 1.0032894736842106e-06,
+      "loss": 1.3225,
+      "step": 20790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.448838472366333,
+      "learning_rate": 1.0023757309941522e-06,
+      "loss": 1.463,
+      "step": 20791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.750401258468628,
+      "learning_rate": 1.0014619883040938e-06,
+      "loss": 1.6771,
+      "step": 20792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.468517303466797,
+      "learning_rate": 1.0005482456140351e-06,
+      "loss": 1.5658,
+      "step": 20793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.164308547973633,
+      "learning_rate": 9.996345029239767e-07,
+      "loss": 1.8656,
+      "step": 20794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.82129168510437,
+      "learning_rate": 9.98720760233918e-07,
+      "loss": 1.4775,
+      "step": 20795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3456690311431885,
+      "learning_rate": 9.978070175438597e-07,
+      "loss": 1.4507,
+      "step": 20796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2039577960968018,
+      "learning_rate": 9.968932748538013e-07,
+      "loss": 1.5964,
+      "step": 20797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3017807006835938,
+      "learning_rate": 9.959795321637428e-07,
+      "loss": 1.3228,
+      "step": 20798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6841251850128174,
+      "learning_rate": 9.950657894736842e-07,
+      "loss": 1.4993,
+      "step": 20799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2723867893218994,
+      "learning_rate": 9.941520467836258e-07,
+      "loss": 1.3917,
+      "step": 20800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.638096332550049,
+      "learning_rate": 9.932383040935674e-07,
+      "loss": 1.6602,
+      "step": 20801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.560375452041626,
+      "learning_rate": 9.92324561403509e-07,
+      "loss": 1.521,
+      "step": 20802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.716956853866577,
+      "learning_rate": 9.914108187134504e-07,
+      "loss": 1.7098,
+      "step": 20803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7253358364105225,
+      "learning_rate": 9.90497076023392e-07,
+      "loss": 1.6906,
+      "step": 20804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9359105825424194,
+      "learning_rate": 9.895833333333333e-07,
+      "loss": 0.9254,
+      "step": 20805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4177067279815674,
+      "learning_rate": 9.88669590643275e-07,
+      "loss": 1.194,
+      "step": 20806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.193592071533203,
+      "learning_rate": 9.877558479532165e-07,
+      "loss": 1.0389,
+      "step": 20807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0330090522766113,
+      "learning_rate": 9.86842105263158e-07,
+      "loss": 1.8864,
+      "step": 20808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6232850551605225,
+      "learning_rate": 9.859283625730995e-07,
+      "loss": 1.7032,
+      "step": 20809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.659404754638672,
+      "learning_rate": 9.85014619883041e-07,
+      "loss": 1.6523,
+      "step": 20810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.364732503890991,
+      "learning_rate": 9.841008771929824e-07,
+      "loss": 1.2611,
+      "step": 20811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.045844078063965,
+      "learning_rate": 9.831871345029242e-07,
+      "loss": 1.3357,
+      "step": 20812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5436062812805176,
+      "learning_rate": 9.822733918128656e-07,
+      "loss": 1.2758,
+      "step": 20813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.81923770904541,
+      "learning_rate": 9.813596491228072e-07,
+      "loss": 1.5929,
+      "step": 20814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.33555269241333,
+      "learning_rate": 9.804459064327485e-07,
+      "loss": 1.3268,
+      "step": 20815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7781686782836914,
+      "learning_rate": 9.795321637426901e-07,
+      "loss": 1.6262,
+      "step": 20816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6947014331817627,
+      "learning_rate": 9.786184210526317e-07,
+      "loss": 1.9128,
+      "step": 20817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.939884662628174,
+      "learning_rate": 9.777046783625733e-07,
+      "loss": 1.6863,
+      "step": 20818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5472252368927,
+      "learning_rate": 9.767909356725147e-07,
+      "loss": 1.4313,
+      "step": 20819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1212103366851807,
+      "learning_rate": 9.758771929824563e-07,
+      "loss": 1.1823,
+      "step": 20820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5354771614074707,
+      "learning_rate": 9.749634502923976e-07,
+      "loss": 1.5058,
+      "step": 20821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2835395336151123,
+      "learning_rate": 9.740497076023392e-07,
+      "loss": 1.5388,
+      "step": 20822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6652379035949707,
+      "learning_rate": 9.731359649122808e-07,
+      "loss": 1.2074,
+      "step": 20823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.578552484512329,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 1.6352,
+      "step": 20824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4869916439056396,
+      "learning_rate": 9.713084795321638e-07,
+      "loss": 1.1976,
+      "step": 20825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7030296325683594,
+      "learning_rate": 9.703947368421054e-07,
+      "loss": 1.4718,
+      "step": 20826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.277953863143921,
+      "learning_rate": 9.694809941520467e-07,
+      "loss": 1.3349,
+      "step": 20827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.227679491043091,
+      "learning_rate": 9.685672514619885e-07,
+      "loss": 1.2444,
+      "step": 20828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2502782344818115,
+      "learning_rate": 9.6765350877193e-07,
+      "loss": 1.2751,
+      "step": 20829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.115434408187866,
+      "learning_rate": 9.667397660818715e-07,
+      "loss": 1.2602,
+      "step": 20830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.097217559814453,
+      "learning_rate": 9.658260233918129e-07,
+      "loss": 1.2849,
+      "step": 20831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.858062744140625,
+      "learning_rate": 9.649122807017545e-07,
+      "loss": 1.7032,
+      "step": 20832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.513345956802368,
+      "learning_rate": 9.63998538011696e-07,
+      "loss": 1.2428,
+      "step": 20833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2976787090301514,
+      "learning_rate": 9.630847953216376e-07,
+      "loss": 1.4346,
+      "step": 20834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.324061393737793,
+      "learning_rate": 9.62171052631579e-07,
+      "loss": 1.3376,
+      "step": 20835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.547664165496826,
+      "learning_rate": 9.612573099415206e-07,
+      "loss": 1.4949,
+      "step": 20836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3930037021636963,
+      "learning_rate": 9.60343567251462e-07,
+      "loss": 1.5075,
+      "step": 20837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.522843599319458,
+      "learning_rate": 9.594298245614038e-07,
+      "loss": 1.4532,
+      "step": 20838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.760291814804077,
+      "learning_rate": 9.585160818713451e-07,
+      "loss": 1.4056,
+      "step": 20839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3388831615448,
+      "learning_rate": 9.576023391812867e-07,
+      "loss": 1.3756,
+      "step": 20840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8348233699798584,
+      "learning_rate": 9.56688596491228e-07,
+      "loss": 1.5358,
+      "step": 20841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4910223484039307,
+      "learning_rate": 9.557748538011697e-07,
+      "loss": 1.5424,
+      "step": 20842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.831214189529419,
+      "learning_rate": 9.54861111111111e-07,
+      "loss": 2.031,
+      "step": 20843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.103799819946289,
+      "learning_rate": 9.539473684210528e-07,
+      "loss": 1.5705,
+      "step": 20844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6251139640808105,
+      "learning_rate": 9.530336257309942e-07,
+      "loss": 1.4455,
+      "step": 20845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.74672532081604,
+      "learning_rate": 9.521198830409357e-07,
+      "loss": 1.4576,
+      "step": 20846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4366037845611572,
+      "learning_rate": 9.512061403508773e-07,
+      "loss": 1.3152,
+      "step": 20847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2997498512268066,
+      "learning_rate": 9.502923976608188e-07,
+      "loss": 1.2821,
+      "step": 20848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6083285808563232,
+      "learning_rate": 9.493786549707604e-07,
+      "loss": 1.3524,
+      "step": 20849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.62420916557312,
+      "learning_rate": 9.484649122807018e-07,
+      "loss": 1.6492,
+      "step": 20850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.276334762573242,
+      "learning_rate": 9.475511695906433e-07,
+      "loss": 1.4442,
+      "step": 20851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.426132917404175,
+      "learning_rate": 9.466374269005849e-07,
+      "loss": 1.4399,
+      "step": 20852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5227928161621094,
+      "learning_rate": 9.457236842105264e-07,
+      "loss": 1.3731,
+      "step": 20853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.348598003387451,
+      "learning_rate": 9.44809941520468e-07,
+      "loss": 1.4112,
+      "step": 20854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8058629035949707,
+      "learning_rate": 9.438961988304094e-07,
+      "loss": 1.4847,
+      "step": 20855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.328615665435791,
+      "learning_rate": 9.429824561403509e-07,
+      "loss": 1.2257,
+      "step": 20856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.685652494430542,
+      "learning_rate": 9.420687134502925e-07,
+      "loss": 1.5578,
+      "step": 20857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1873812675476074,
+      "learning_rate": 9.41154970760234e-07,
+      "loss": 1.0112,
+      "step": 20858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4250402450561523,
+      "learning_rate": 9.402412280701755e-07,
+      "loss": 1.3895,
+      "step": 20859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1109867095947266,
+      "learning_rate": 9.393274853801171e-07,
+      "loss": 1.3195,
+      "step": 20860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3241701126098633,
+      "learning_rate": 9.384137426900585e-07,
+      "loss": 1.3385,
+      "step": 20861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4583628177642822,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 1.5888,
+      "step": 20862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8090078830718994,
+      "learning_rate": 9.365862573099416e-07,
+      "loss": 1.6462,
+      "step": 20863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3828511238098145,
+      "learning_rate": 9.356725146198831e-07,
+      "loss": 1.3164,
+      "step": 20864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7136337757110596,
+      "learning_rate": 9.347587719298247e-07,
+      "loss": 1.5394,
+      "step": 20865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0596091747283936,
+      "learning_rate": 9.338450292397662e-07,
+      "loss": 0.9881,
+      "step": 20866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.263291597366333,
+      "learning_rate": 9.329312865497076e-07,
+      "loss": 1.5066,
+      "step": 20867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.157575845718384,
+      "learning_rate": 9.320175438596492e-07,
+      "loss": 1.2708,
+      "step": 20868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.221771240234375,
+      "learning_rate": 9.311038011695907e-07,
+      "loss": 1.2157,
+      "step": 20869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2594685554504395,
+      "learning_rate": 9.301900584795322e-07,
+      "loss": 1.7813,
+      "step": 20870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.749260663986206,
+      "learning_rate": 9.292763157894738e-07,
+      "loss": 1.93,
+      "step": 20871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.647181987762451,
+      "learning_rate": 9.283625730994153e-07,
+      "loss": 1.5733,
+      "step": 20872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5749197006225586,
+      "learning_rate": 9.274488304093568e-07,
+      "loss": 1.5481,
+      "step": 20873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5455048084259033,
+      "learning_rate": 9.265350877192983e-07,
+      "loss": 1.4533,
+      "step": 20874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.009113073348999,
+      "learning_rate": 9.256213450292398e-07,
+      "loss": 1.1553,
+      "step": 20875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2241718769073486,
+      "learning_rate": 9.247076023391814e-07,
+      "loss": 1.4072,
+      "step": 20876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.592529296875,
+      "learning_rate": 9.237938596491229e-07,
+      "loss": 1.2294,
+      "step": 20877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7744834423065186,
+      "learning_rate": 9.228801169590644e-07,
+      "loss": 1.8363,
+      "step": 20878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2180280685424805,
+      "learning_rate": 9.219663742690059e-07,
+      "loss": 1.6931,
+      "step": 20879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.525836944580078,
+      "learning_rate": 9.210526315789474e-07,
+      "loss": 1.6381,
+      "step": 20880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6743366718292236,
+      "learning_rate": 9.20138888888889e-07,
+      "loss": 1.4595,
+      "step": 20881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.878993272781372,
+      "learning_rate": 9.192251461988305e-07,
+      "loss": 1.6383,
+      "step": 20882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.725677013397217,
+      "learning_rate": 9.183114035087721e-07,
+      "loss": 1.5226,
+      "step": 20883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.99322509765625,
+      "learning_rate": 9.173976608187135e-07,
+      "loss": 1.764,
+      "step": 20884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.268240213394165,
+      "learning_rate": 9.16483918128655e-07,
+      "loss": 1.1718,
+      "step": 20885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.123723030090332,
+      "learning_rate": 9.155701754385965e-07,
+      "loss": 1.226,
+      "step": 20886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4942715167999268,
+      "learning_rate": 9.146564327485381e-07,
+      "loss": 1.4742,
+      "step": 20887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.317780017852783,
+      "learning_rate": 9.137426900584797e-07,
+      "loss": 1.2591,
+      "step": 20888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6029045581817627,
+      "learning_rate": 9.128289473684212e-07,
+      "loss": 1.3275,
+      "step": 20889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.828636407852173,
+      "learning_rate": 9.119152046783626e-07,
+      "loss": 1.1309,
+      "step": 20890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.336386203765869,
+      "learning_rate": 9.110014619883041e-07,
+      "loss": 1.6436,
+      "step": 20891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5596351623535156,
+      "learning_rate": 9.100877192982457e-07,
+      "loss": 1.4338,
+      "step": 20892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.387897253036499,
+      "learning_rate": 9.091739766081872e-07,
+      "loss": 1.433,
+      "step": 20893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1354477405548096,
+      "learning_rate": 9.082602339181288e-07,
+      "loss": 1.6284,
+      "step": 20894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.413587808609009,
+      "learning_rate": 9.073464912280703e-07,
+      "loss": 1.2829,
+      "step": 20895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.684274196624756,
+      "learning_rate": 9.064327485380117e-07,
+      "loss": 1.3379,
+      "step": 20896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1281256675720215,
+      "learning_rate": 9.055190058479533e-07,
+      "loss": 1.0557,
+      "step": 20897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.527975082397461,
+      "learning_rate": 9.046052631578948e-07,
+      "loss": 1.2611,
+      "step": 20898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3877339363098145,
+      "learning_rate": 9.036915204678364e-07,
+      "loss": 1.2776,
+      "step": 20899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4112045764923096,
+      "learning_rate": 9.027777777777779e-07,
+      "loss": 1.2472,
+      "step": 20900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2846009731292725,
+      "learning_rate": 9.018640350877193e-07,
+      "loss": 1.1123,
+      "step": 20901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5073564052581787,
+      "learning_rate": 9.009502923976608e-07,
+      "loss": 1.4076,
+      "step": 20902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1408181190490723,
+      "learning_rate": 9.000365497076024e-07,
+      "loss": 1.2103,
+      "step": 20903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1046981811523438,
+      "learning_rate": 8.99122807017544e-07,
+      "loss": 1.5102,
+      "step": 20904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3928310871124268,
+      "learning_rate": 8.982090643274855e-07,
+      "loss": 1.4804,
+      "step": 20905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3386294841766357,
+      "learning_rate": 8.97295321637427e-07,
+      "loss": 1.4341,
+      "step": 20906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.695258855819702,
+      "learning_rate": 8.963815789473684e-07,
+      "loss": 1.3766,
+      "step": 20907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1533145904541016,
+      "learning_rate": 8.9546783625731e-07,
+      "loss": 1.0017,
+      "step": 20908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.540273904800415,
+      "learning_rate": 8.945540935672516e-07,
+      "loss": 1.1348,
+      "step": 20909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.26162052154541,
+      "learning_rate": 8.936403508771931e-07,
+      "loss": 1.3508,
+      "step": 20910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.741267681121826,
+      "learning_rate": 8.927266081871346e-07,
+      "loss": 1.4542,
+      "step": 20911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6389172077178955,
+      "learning_rate": 8.91812865497076e-07,
+      "loss": 1.6039,
+      "step": 20912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9009604454040527,
+      "learning_rate": 8.908991228070176e-07,
+      "loss": 1.8019,
+      "step": 20913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.582183837890625,
+      "learning_rate": 8.899853801169592e-07,
+      "loss": 1.2488,
+      "step": 20914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.682068347930908,
+      "learning_rate": 8.890716374269007e-07,
+      "loss": 1.6122,
+      "step": 20915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.333186626434326,
+      "learning_rate": 8.881578947368422e-07,
+      "loss": 1.3898,
+      "step": 20916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.542029619216919,
+      "learning_rate": 8.872441520467837e-07,
+      "loss": 1.5853,
+      "step": 20917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.349181890487671,
+      "learning_rate": 8.863304093567251e-07,
+      "loss": 1.4158,
+      "step": 20918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2034189701080322,
+      "learning_rate": 8.854166666666668e-07,
+      "loss": 1.1469,
+      "step": 20919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7364864349365234,
+      "learning_rate": 8.845029239766083e-07,
+      "loss": 1.6989,
+      "step": 20920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.621079206466675,
+      "learning_rate": 8.835891812865498e-07,
+      "loss": 1.8825,
+      "step": 20921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.177879571914673,
+      "learning_rate": 8.826754385964913e-07,
+      "loss": 1.3975,
+      "step": 20922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.232760429382324,
+      "learning_rate": 8.817616959064328e-07,
+      "loss": 1.5044,
+      "step": 20923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.808964490890503,
+      "learning_rate": 8.808479532163743e-07,
+      "loss": 0.9827,
+      "step": 20924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2539918422698975,
+      "learning_rate": 8.799342105263159e-07,
+      "loss": 1.4441,
+      "step": 20925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9617719650268555,
+      "learning_rate": 8.790204678362574e-07,
+      "loss": 1.6303,
+      "step": 20926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.83156156539917,
+      "learning_rate": 8.781067251461989e-07,
+      "loss": 1.5517,
+      "step": 20927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7417378425598145,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 1.6626,
+      "step": 20928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.677279233932495,
+      "learning_rate": 8.76279239766082e-07,
+      "loss": 1.4248,
+      "step": 20929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.366384983062744,
+      "learning_rate": 8.753654970760235e-07,
+      "loss": 1.5193,
+      "step": 20930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.195406675338745,
+      "learning_rate": 8.74451754385965e-07,
+      "loss": 1.4235,
+      "step": 20931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2429709434509277,
+      "learning_rate": 8.735380116959065e-07,
+      "loss": 1.2572,
+      "step": 20932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0713067054748535,
+      "learning_rate": 8.72624269005848e-07,
+      "loss": 1.6905,
+      "step": 20933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2864115238189697,
+      "learning_rate": 8.717105263157895e-07,
+      "loss": 1.6875,
+      "step": 20934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2845234870910645,
+      "learning_rate": 8.707967836257312e-07,
+      "loss": 1.3147,
+      "step": 20935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.583547830581665,
+      "learning_rate": 8.698830409356726e-07,
+      "loss": 1.3193,
+      "step": 20936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7373974323272705,
+      "learning_rate": 8.689692982456141e-07,
+      "loss": 1.7005,
+      "step": 20937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6022026538848877,
+      "learning_rate": 8.680555555555556e-07,
+      "loss": 1.6152,
+      "step": 20938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.100905656814575,
+      "learning_rate": 8.671418128654971e-07,
+      "loss": 1.1452,
+      "step": 20939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.812673330307007,
+      "learning_rate": 8.662280701754388e-07,
+      "loss": 1.3754,
+      "step": 20940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4149577617645264,
+      "learning_rate": 8.653143274853802e-07,
+      "loss": 1.5502,
+      "step": 20941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.462540864944458,
+      "learning_rate": 8.644005847953217e-07,
+      "loss": 1.3232,
+      "step": 20942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8793096542358398,
+      "learning_rate": 8.634868421052632e-07,
+      "loss": 1.0988,
+      "step": 20943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.629988431930542,
+      "learning_rate": 8.625730994152047e-07,
+      "loss": 1.4254,
+      "step": 20944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.69452166557312,
+      "learning_rate": 8.616593567251464e-07,
+      "loss": 1.4354,
+      "step": 20945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4928553104400635,
+      "learning_rate": 8.607456140350879e-07,
+      "loss": 1.6301,
+      "step": 20946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1475861072540283,
+      "learning_rate": 8.598318713450293e-07,
+      "loss": 1.2374,
+      "step": 20947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7717974185943604,
+      "learning_rate": 8.589181286549708e-07,
+      "loss": 1.2611,
+      "step": 20948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7161052227020264,
+      "learning_rate": 8.580043859649123e-07,
+      "loss": 1.5269,
+      "step": 20949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2071621417999268,
+      "learning_rate": 8.570906432748538e-07,
+      "loss": 1.2984,
+      "step": 20950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7375032901763916,
+      "learning_rate": 8.561769005847955e-07,
+      "loss": 1.2156,
+      "step": 20951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4012205600738525,
+      "learning_rate": 8.55263157894737e-07,
+      "loss": 1.4666,
+      "step": 20952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6616647243499756,
+      "learning_rate": 8.543494152046784e-07,
+      "loss": 1.4716,
+      "step": 20953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1555233001708984,
+      "learning_rate": 8.534356725146199e-07,
+      "loss": 1.4308,
+      "step": 20954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9662333726882935,
+      "learning_rate": 8.525219298245614e-07,
+      "loss": 1.0162,
+      "step": 20955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2004551887512207,
+      "learning_rate": 8.516081871345031e-07,
+      "loss": 1.2503,
+      "step": 20956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.380347490310669,
+      "learning_rate": 8.506944444444446e-07,
+      "loss": 1.2599,
+      "step": 20957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.533217430114746,
+      "learning_rate": 8.49780701754386e-07,
+      "loss": 1.5104,
+      "step": 20958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.456559181213379,
+      "learning_rate": 8.488669590643275e-07,
+      "loss": 1.3018,
+      "step": 20959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4194576740264893,
+      "learning_rate": 8.47953216374269e-07,
+      "loss": 1.2548,
+      "step": 20960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2280211448669434,
+      "learning_rate": 8.470394736842107e-07,
+      "loss": 1.0359,
+      "step": 20961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.521088123321533,
+      "learning_rate": 8.461257309941522e-07,
+      "loss": 1.3389,
+      "step": 20962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.314307928085327,
+      "learning_rate": 8.452119883040937e-07,
+      "loss": 1.2691,
+      "step": 20963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3494174480438232,
+      "learning_rate": 8.442982456140351e-07,
+      "loss": 1.1684,
+      "step": 20964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.281170129776001,
+      "learning_rate": 8.433845029239766e-07,
+      "loss": 1.303,
+      "step": 20965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.529407024383545,
+      "learning_rate": 8.424707602339181e-07,
+      "loss": 1.4138,
+      "step": 20966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.203511953353882,
+      "learning_rate": 8.415570175438598e-07,
+      "loss": 1.4395,
+      "step": 20967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.33140230178833,
+      "learning_rate": 8.406432748538013e-07,
+      "loss": 1.6039,
+      "step": 20968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.069943904876709,
+      "learning_rate": 8.397295321637428e-07,
+      "loss": 1.867,
+      "step": 20969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3934595584869385,
+      "learning_rate": 8.388157894736842e-07,
+      "loss": 1.1883,
+      "step": 20970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8284919261932373,
+      "learning_rate": 8.379020467836257e-07,
+      "loss": 1.6045,
+      "step": 20971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3746755123138428,
+      "learning_rate": 8.369883040935674e-07,
+      "loss": 1.4091,
+      "step": 20972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2511398792266846,
+      "learning_rate": 8.360745614035089e-07,
+      "loss": 1.2165,
+      "step": 20973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5840272903442383,
+      "learning_rate": 8.351608187134504e-07,
+      "loss": 1.2165,
+      "step": 20974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.304821252822876,
+      "learning_rate": 8.342470760233919e-07,
+      "loss": 1.1757,
+      "step": 20975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.349648952484131,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 1.2832,
+      "step": 20976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5634689331054688,
+      "learning_rate": 8.32419590643275e-07,
+      "loss": 1.4387,
+      "step": 20977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0986602306365967,
+      "learning_rate": 8.315058479532165e-07,
+      "loss": 1.911,
+      "step": 20978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8474669456481934,
+      "learning_rate": 8.30592105263158e-07,
+      "loss": 1.6326,
+      "step": 20979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4716861248016357,
+      "learning_rate": 8.296783625730995e-07,
+      "loss": 1.5767,
+      "step": 20980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2784175872802734,
+      "learning_rate": 8.287646198830409e-07,
+      "loss": 1.1849,
+      "step": 20981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1117348670959473,
+      "learning_rate": 8.278508771929824e-07,
+      "loss": 1.2783,
+      "step": 20982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3335177898406982,
+      "learning_rate": 8.269371345029241e-07,
+      "loss": 1.5602,
+      "step": 20983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6317412853240967,
+      "learning_rate": 8.260233918128656e-07,
+      "loss": 1.4935,
+      "step": 20984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0066068172454834,
+      "learning_rate": 8.251096491228071e-07,
+      "loss": 1.293,
+      "step": 20985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2253379821777344,
+      "learning_rate": 8.241959064327486e-07,
+      "loss": 1.5058,
+      "step": 20986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4775991439819336,
+      "learning_rate": 8.2328216374269e-07,
+      "loss": 1.4556,
+      "step": 20987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.587299346923828,
+      "learning_rate": 8.223684210526317e-07,
+      "loss": 1.9078,
+      "step": 20988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5256004333496094,
+      "learning_rate": 8.214546783625732e-07,
+      "loss": 1.7312,
+      "step": 20989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.249467611312866,
+      "learning_rate": 8.205409356725147e-07,
+      "loss": 1.4284,
+      "step": 20990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.173058032989502,
+      "learning_rate": 8.196271929824562e-07,
+      "loss": 1.1368,
+      "step": 20991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.591754913330078,
+      "learning_rate": 8.187134502923977e-07,
+      "loss": 1.4774,
+      "step": 20992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4991562366485596,
+      "learning_rate": 8.177997076023393e-07,
+      "loss": 1.0967,
+      "step": 20993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4235095977783203,
+      "learning_rate": 8.168859649122808e-07,
+      "loss": 1.4645,
+      "step": 20994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8479833602905273,
+      "learning_rate": 8.159722222222223e-07,
+      "loss": 1.4347,
+      "step": 20995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6038055419921875,
+      "learning_rate": 8.150584795321638e-07,
+      "loss": 1.4128,
+      "step": 20996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5447640419006348,
+      "learning_rate": 8.141447368421053e-07,
+      "loss": 1.5602,
+      "step": 20997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.796785593032837,
+      "learning_rate": 8.132309941520467e-07,
+      "loss": 1.4134,
+      "step": 20998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8976047039031982,
+      "learning_rate": 8.123172514619884e-07,
+      "loss": 1.7785,
+      "step": 20999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.178868532180786,
+      "learning_rate": 8.114035087719299e-07,
+      "loss": 1.2272,
+      "step": 21000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.76419734954834,
+      "learning_rate": 8.104897660818714e-07,
+      "loss": 1.745,
+      "step": 21001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.683194875717163,
+      "learning_rate": 8.095760233918129e-07,
+      "loss": 1.4742,
+      "step": 21002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.846841335296631,
+      "learning_rate": 8.086622807017544e-07,
+      "loss": 1.4623,
+      "step": 21003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4955904483795166,
+      "learning_rate": 8.07748538011696e-07,
+      "loss": 1.5287,
+      "step": 21004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7148008346557617,
+      "learning_rate": 8.068347953216375e-07,
+      "loss": 1.7183,
+      "step": 21005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3212544918060303,
+      "learning_rate": 8.05921052631579e-07,
+      "loss": 1.5651,
+      "step": 21006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.445429801940918,
+      "learning_rate": 8.050073099415205e-07,
+      "loss": 1.3095,
+      "step": 21007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9392340183258057,
+      "learning_rate": 8.04093567251462e-07,
+      "loss": 1.8056,
+      "step": 21008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3711636066436768,
+      "learning_rate": 8.031798245614037e-07,
+      "loss": 1.3714,
+      "step": 21009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.048388719558716,
+      "learning_rate": 8.022660818713451e-07,
+      "loss": 1.5165,
+      "step": 21010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9703176021575928,
+      "learning_rate": 8.013523391812866e-07,
+      "loss": 1.8087,
+      "step": 21011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.68696928024292,
+      "learning_rate": 8.004385964912281e-07,
+      "loss": 1.5626,
+      "step": 21012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4198131561279297,
+      "learning_rate": 7.995248538011696e-07,
+      "loss": 1.6383,
+      "step": 21013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0359342098236084,
+      "learning_rate": 7.986111111111111e-07,
+      "loss": 1.648,
+      "step": 21014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6800570487976074,
+      "learning_rate": 7.976973684210528e-07,
+      "loss": 1.5434,
+      "step": 21015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.61934757232666,
+      "learning_rate": 7.967836257309942e-07,
+      "loss": 1.4284,
+      "step": 21016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5685043334960938,
+      "learning_rate": 7.958698830409357e-07,
+      "loss": 1.5849,
+      "step": 21017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4230458736419678,
+      "learning_rate": 7.949561403508772e-07,
+      "loss": 1.5865,
+      "step": 21018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.22578501701355,
+      "learning_rate": 7.940423976608187e-07,
+      "loss": 1.5487,
+      "step": 21019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1666009426116943,
+      "learning_rate": 7.931286549707604e-07,
+      "loss": 1.42,
+      "step": 21020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9685457944869995,
+      "learning_rate": 7.922149122807019e-07,
+      "loss": 1.1314,
+      "step": 21021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0788116455078125,
+      "learning_rate": 7.913011695906433e-07,
+      "loss": 1.2428,
+      "step": 21022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9990506172180176,
+      "learning_rate": 7.903874269005848e-07,
+      "loss": 1.9396,
+      "step": 21023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6506149768829346,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 1.912,
+      "step": 21024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.16040301322937,
+      "learning_rate": 7.88559941520468e-07,
+      "loss": 1.0244,
+      "step": 21025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0595996379852295,
+      "learning_rate": 7.876461988304095e-07,
+      "loss": 1.3042,
+      "step": 21026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1057934761047363,
+      "learning_rate": 7.867324561403509e-07,
+      "loss": 1.1306,
+      "step": 21027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6998484134674072,
+      "learning_rate": 7.858187134502924e-07,
+      "loss": 1.3891,
+      "step": 21028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9757044315338135,
+      "learning_rate": 7.849049707602339e-07,
+      "loss": 1.6137,
+      "step": 21029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7790720462799072,
+      "learning_rate": 7.839912280701755e-07,
+      "loss": 1.6666,
+      "step": 21030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4417355060577393,
+      "learning_rate": 7.830774853801171e-07,
+      "loss": 1.3904,
+      "step": 21031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.318220853805542,
+      "learning_rate": 7.821637426900586e-07,
+      "loss": 1.2357,
+      "step": 21032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.373553991317749,
+      "learning_rate": 7.8125e-07,
+      "loss": 1.3259,
+      "step": 21033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.522205114364624,
+      "learning_rate": 7.803362573099415e-07,
+      "loss": 1.5292,
+      "step": 21034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.36093807220459,
+      "learning_rate": 7.794225146198831e-07,
+      "loss": 1.2668,
+      "step": 21035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4863500595092773,
+      "learning_rate": 7.785087719298247e-07,
+      "loss": 1.4469,
+      "step": 21036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3687117099761963,
+      "learning_rate": 7.775950292397662e-07,
+      "loss": 1.4073,
+      "step": 21037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4695231914520264,
+      "learning_rate": 7.766812865497077e-07,
+      "loss": 1.4455,
+      "step": 21038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6209373474121094,
+      "learning_rate": 7.757675438596491e-07,
+      "loss": 1.6414,
+      "step": 21039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.086655855178833,
+      "learning_rate": 7.748538011695906e-07,
+      "loss": 1.4067,
+      "step": 21040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.703355312347412,
+      "learning_rate": 7.739400584795322e-07,
+      "loss": 1.2243,
+      "step": 21041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2222707271575928,
+      "learning_rate": 7.730263157894738e-07,
+      "loss": 1.0915,
+      "step": 21042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.974520206451416,
+      "learning_rate": 7.721125730994153e-07,
+      "loss": 1.6254,
+      "step": 21043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0642898082733154,
+      "learning_rate": 7.711988304093567e-07,
+      "loss": 1.1366,
+      "step": 21044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.733551502227783,
+      "learning_rate": 7.702850877192982e-07,
+      "loss": 1.4444,
+      "step": 21045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9528818130493164,
+      "learning_rate": 7.693713450292398e-07,
+      "loss": 1.9085,
+      "step": 21046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2075693607330322,
+      "learning_rate": 7.684576023391814e-07,
+      "loss": 1.3515,
+      "step": 21047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3381690979003906,
+      "learning_rate": 7.675438596491229e-07,
+      "loss": 1.3848,
+      "step": 21048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.365901231765747,
+      "learning_rate": 7.666301169590644e-07,
+      "loss": 1.4904,
+      "step": 21049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.446636199951172,
+      "learning_rate": 7.657163742690058e-07,
+      "loss": 1.431,
+      "step": 21050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.156643867492676,
+      "learning_rate": 7.648026315789474e-07,
+      "loss": 1.595,
+      "step": 21051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.202885627746582,
+      "learning_rate": 7.63888888888889e-07,
+      "loss": 1.2665,
+      "step": 21052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2569594383239746,
+      "learning_rate": 7.629751461988305e-07,
+      "loss": 1.2923,
+      "step": 21053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1495096683502197,
+      "learning_rate": 7.62061403508772e-07,
+      "loss": 1.3209,
+      "step": 21054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3972318172454834,
+      "learning_rate": 7.611476608187135e-07,
+      "loss": 1.6112,
+      "step": 21055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.356074571609497,
+      "learning_rate": 7.60233918128655e-07,
+      "loss": 1.4675,
+      "step": 21056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9718862771987915,
+      "learning_rate": 7.593201754385965e-07,
+      "loss": 1.1389,
+      "step": 21057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.313908815383911,
+      "learning_rate": 7.584064327485381e-07,
+      "loss": 1.5045,
+      "step": 21058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5311622619628906,
+      "learning_rate": 7.574926900584796e-07,
+      "loss": 1.3301,
+      "step": 21059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2749714851379395,
+      "learning_rate": 7.565789473684211e-07,
+      "loss": 1.3524,
+      "step": 21060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3074898719787598,
+      "learning_rate": 7.556652046783627e-07,
+      "loss": 1.4372,
+      "step": 21061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.756124258041382,
+      "learning_rate": 7.547514619883041e-07,
+      "loss": 1.7647,
+      "step": 21062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6396238803863525,
+      "learning_rate": 7.538377192982457e-07,
+      "loss": 1.4565,
+      "step": 21063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.526177406311035,
+      "learning_rate": 7.529239766081872e-07,
+      "loss": 1.2221,
+      "step": 21064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.105978488922119,
+      "learning_rate": 7.520102339181287e-07,
+      "loss": 1.4855,
+      "step": 21065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1663336753845215,
+      "learning_rate": 7.510964912280703e-07,
+      "loss": 1.341,
+      "step": 21066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2002058029174805,
+      "learning_rate": 7.501827485380117e-07,
+      "loss": 1.2152,
+      "step": 21067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4309070110321045,
+      "learning_rate": 7.492690058479533e-07,
+      "loss": 1.4018,
+      "step": 21068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.135885238647461,
+      "learning_rate": 7.483552631578948e-07,
+      "loss": 1.2983,
+      "step": 21069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1128270626068115,
+      "learning_rate": 7.474415204678363e-07,
+      "loss": 0.9994,
+      "step": 21070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4728264808654785,
+      "learning_rate": 7.465277777777778e-07,
+      "loss": 1.4727,
+      "step": 21071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.854464054107666,
+      "learning_rate": 7.456140350877194e-07,
+      "loss": 1.8197,
+      "step": 21072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9276838302612305,
+      "learning_rate": 7.447002923976608e-07,
+      "loss": 1.6756,
+      "step": 21073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7634854316711426,
+      "learning_rate": 7.437865497076024e-07,
+      "loss": 1.556,
+      "step": 21074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.834277391433716,
+      "learning_rate": 7.428728070175439e-07,
+      "loss": 1.4676,
+      "step": 21075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7210874557495117,
+      "learning_rate": 7.419590643274854e-07,
+      "loss": 1.8179,
+      "step": 21076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.219244956970215,
+      "learning_rate": 7.41045321637427e-07,
+      "loss": 1.3072,
+      "step": 21077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.392595052719116,
+      "learning_rate": 7.401315789473685e-07,
+      "loss": 1.442,
+      "step": 21078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9178218841552734,
+      "learning_rate": 7.3921783625731e-07,
+      "loss": 1.211,
+      "step": 21079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8953804969787598,
+      "learning_rate": 7.383040935672515e-07,
+      "loss": 1.3151,
+      "step": 21080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.780047655105591,
+      "learning_rate": 7.37390350877193e-07,
+      "loss": 1.5177,
+      "step": 21081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.881598949432373,
+      "learning_rate": 7.364766081871346e-07,
+      "loss": 1.394,
+      "step": 21082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2524588108062744,
+      "learning_rate": 7.355628654970761e-07,
+      "loss": 1.5495,
+      "step": 21083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.382707118988037,
+      "learning_rate": 7.346491228070177e-07,
+      "loss": 1.1945,
+      "step": 21084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3832478523254395,
+      "learning_rate": 7.337353801169591e-07,
+      "loss": 1.47,
+      "step": 21085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.742050886154175,
+      "learning_rate": 7.328216374269006e-07,
+      "loss": 1.5505,
+      "step": 21086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.213484287261963,
+      "learning_rate": 7.319078947368422e-07,
+      "loss": 1.2591,
+      "step": 21087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6915345191955566,
+      "learning_rate": 7.309941520467837e-07,
+      "loss": 1.3809,
+      "step": 21088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6349294185638428,
+      "learning_rate": 7.300804093567252e-07,
+      "loss": 1.3747,
+      "step": 21089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6134445667266846,
+      "learning_rate": 7.291666666666667e-07,
+      "loss": 1.5357,
+      "step": 21090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1409542560577393,
+      "learning_rate": 7.282529239766082e-07,
+      "loss": 1.3024,
+      "step": 21091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.854304552078247,
+      "learning_rate": 7.273391812865498e-07,
+      "loss": 1.6788,
+      "step": 21092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.323638439178467,
+      "learning_rate": 7.264254385964913e-07,
+      "loss": 1.5072,
+      "step": 21093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2367727756500244,
+      "learning_rate": 7.255116959064328e-07,
+      "loss": 1.4198,
+      "step": 21094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4160537719726562,
+      "learning_rate": 7.245979532163744e-07,
+      "loss": 1.2566,
+      "step": 21095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7310092449188232,
+      "learning_rate": 7.236842105263158e-07,
+      "loss": 1.4951,
+      "step": 21096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8115732669830322,
+      "learning_rate": 7.227704678362574e-07,
+      "loss": 1.4666,
+      "step": 21097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2891976833343506,
+      "learning_rate": 7.218567251461989e-07,
+      "loss": 1.1598,
+      "step": 21098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2843875885009766,
+      "learning_rate": 7.209429824561404e-07,
+      "loss": 1.0504,
+      "step": 21099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3476133346557617,
+      "learning_rate": 7.20029239766082e-07,
+      "loss": 1.5334,
+      "step": 21100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5307369232177734,
+      "learning_rate": 7.191154970760235e-07,
+      "loss": 1.5886,
+      "step": 21101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4784293174743652,
+      "learning_rate": 7.182017543859649e-07,
+      "loss": 1.3193,
+      "step": 21102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.321533441543579,
+      "learning_rate": 7.172880116959065e-07,
+      "loss": 1.4022,
+      "step": 21103
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8315134048461914,
+      "learning_rate": 7.16374269005848e-07,
+      "loss": 1.5685,
+      "step": 21104
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.980767011642456,
+      "learning_rate": 7.154605263157895e-07,
+      "loss": 1.6705,
+      "step": 21105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.748222589492798,
+      "learning_rate": 7.145467836257311e-07,
+      "loss": 1.4259,
+      "step": 21106
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0264341831207275,
+      "learning_rate": 7.136330409356725e-07,
+      "loss": 0.9917,
+      "step": 21107
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2243456840515137,
+      "learning_rate": 7.127192982456141e-07,
+      "loss": 1.2792,
+      "step": 21108
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1697747707366943,
+      "learning_rate": 7.118055555555556e-07,
+      "loss": 1.2311,
+      "step": 21109
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.232814073562622,
+      "learning_rate": 7.108918128654971e-07,
+      "loss": 1.222,
+      "step": 21110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.234464168548584,
+      "learning_rate": 7.099780701754387e-07,
+      "loss": 1.3257,
+      "step": 21111
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.708265542984009,
+      "learning_rate": 7.090643274853802e-07,
+      "loss": 1.1197,
+      "step": 21112
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6030426025390625,
+      "learning_rate": 7.081505847953217e-07,
+      "loss": 1.5927,
+      "step": 21113
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7932045459747314,
+      "learning_rate": 7.072368421052632e-07,
+      "loss": 1.8625,
+      "step": 21114
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2935585975646973,
+      "learning_rate": 7.063230994152047e-07,
+      "loss": 1.2814,
+      "step": 21115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.206212043762207,
+      "learning_rate": 7.054093567251463e-07,
+      "loss": 1.5445,
+      "step": 21116
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3150503635406494,
+      "learning_rate": 7.044956140350878e-07,
+      "loss": 1.2851,
+      "step": 21117
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.864011526107788,
+      "learning_rate": 7.035818713450294e-07,
+      "loss": 1.4941,
+      "step": 21118
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7622311115264893,
+      "learning_rate": 7.026681286549708e-07,
+      "loss": 1.4396,
+      "step": 21119
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.349591016769409,
+      "learning_rate": 7.017543859649123e-07,
+      "loss": 1.2699,
+      "step": 21120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9179675579071045,
+      "learning_rate": 7.008406432748538e-07,
+      "loss": 2.0611,
+      "step": 21121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0208868980407715,
+      "learning_rate": 6.999269005847954e-07,
+      "loss": 1.1732,
+      "step": 21122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2393102645874023,
+      "learning_rate": 6.99013157894737e-07,
+      "loss": 1.051,
+      "step": 21123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2394466400146484,
+      "learning_rate": 6.980994152046785e-07,
+      "loss": 1.8341,
+      "step": 21124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.661515951156616,
+      "learning_rate": 6.971856725146199e-07,
+      "loss": 1.6802,
+      "step": 21125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1161417961120605,
+      "learning_rate": 6.962719298245614e-07,
+      "loss": 1.3496,
+      "step": 21126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.996495008468628,
+      "learning_rate": 6.95358187134503e-07,
+      "loss": 1.4918,
+      "step": 21127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5286521911621094,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 1.2622,
+      "step": 21128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7834630012512207,
+      "learning_rate": 6.935307017543861e-07,
+      "loss": 1.1376,
+      "step": 21129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7059357166290283,
+      "learning_rate": 6.926169590643275e-07,
+      "loss": 1.9102,
+      "step": 21130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5524628162384033,
+      "learning_rate": 6.91703216374269e-07,
+      "loss": 1.5907,
+      "step": 21131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0538251399993896,
+      "learning_rate": 6.907894736842106e-07,
+      "loss": 0.9063,
+      "step": 21132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.274021625518799,
+      "learning_rate": 6.898757309941521e-07,
+      "loss": 1.2122,
+      "step": 21133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5767574310302734,
+      "learning_rate": 6.889619883040937e-07,
+      "loss": 1.7449,
+      "step": 21134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.876603126525879,
+      "learning_rate": 6.880482456140352e-07,
+      "loss": 1.7356,
+      "step": 21135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5485854148864746,
+      "learning_rate": 6.871345029239766e-07,
+      "loss": 1.5688,
+      "step": 21136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.35440731048584,
+      "learning_rate": 6.862207602339181e-07,
+      "loss": 1.2121,
+      "step": 21137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.264244318008423,
+      "learning_rate": 6.853070175438597e-07,
+      "loss": 1.3111,
+      "step": 21138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9023444652557373,
+      "learning_rate": 6.843932748538013e-07,
+      "loss": 1.4729,
+      "step": 21139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8736026287078857,
+      "learning_rate": 6.834795321637428e-07,
+      "loss": 1.6446,
+      "step": 21140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.218703269958496,
+      "learning_rate": 6.825657894736843e-07,
+      "loss": 1.3368,
+      "step": 21141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7286925315856934,
+      "learning_rate": 6.816520467836257e-07,
+      "loss": 1.2779,
+      "step": 21142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.140192985534668,
+      "learning_rate": 6.807383040935673e-07,
+      "loss": 1.3481,
+      "step": 21143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.092881679534912,
+      "learning_rate": 6.798245614035089e-07,
+      "loss": 1.787,
+      "step": 21144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.750336170196533,
+      "learning_rate": 6.789108187134504e-07,
+      "loss": 1.7693,
+      "step": 21145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.805021286010742,
+      "learning_rate": 6.779970760233919e-07,
+      "loss": 1.3954,
+      "step": 21146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.013972282409668,
+      "learning_rate": 6.770833333333333e-07,
+      "loss": 1.5637,
+      "step": 21147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3734426498413086,
+      "learning_rate": 6.761695906432749e-07,
+      "loss": 1.3088,
+      "step": 21148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8142590522766113,
+      "learning_rate": 6.752558479532165e-07,
+      "loss": 1.3208,
+      "step": 21149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.82541823387146,
+      "learning_rate": 6.74342105263158e-07,
+      "loss": 1.5811,
+      "step": 21150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3073782920837402,
+      "learning_rate": 6.734283625730995e-07,
+      "loss": 1.3462,
+      "step": 21151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6487538814544678,
+      "learning_rate": 6.72514619883041e-07,
+      "loss": 1.4097,
+      "step": 21152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0568912029266357,
+      "learning_rate": 6.716008771929824e-07,
+      "loss": 1.8026,
+      "step": 21153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.523716688156128,
+      "learning_rate": 6.706871345029241e-07,
+      "loss": 1.2735,
+      "step": 21154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.107107162475586,
+      "learning_rate": 6.697733918128656e-07,
+      "loss": 1.2647,
+      "step": 21155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4720349311828613,
+      "learning_rate": 6.688596491228071e-07,
+      "loss": 1.3596,
+      "step": 21156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.377366304397583,
+      "learning_rate": 6.679459064327486e-07,
+      "loss": 1.225,
+      "step": 21157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6291539669036865,
+      "learning_rate": 6.6703216374269e-07,
+      "loss": 1.4584,
+      "step": 21158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.647319793701172,
+      "learning_rate": 6.661184210526317e-07,
+      "loss": 1.6204,
+      "step": 21159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5198159217834473,
+      "learning_rate": 6.652046783625732e-07,
+      "loss": 1.5504,
+      "step": 21160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8492789268493652,
+      "learning_rate": 6.642909356725147e-07,
+      "loss": 1.893,
+      "step": 21161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8837358951568604,
+      "learning_rate": 6.633771929824562e-07,
+      "loss": 1.3618,
+      "step": 21162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.441681146621704,
+      "learning_rate": 6.624634502923977e-07,
+      "loss": 1.5457,
+      "step": 21163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2587227821350098,
+      "learning_rate": 6.615497076023393e-07,
+      "loss": 1.4548,
+      "step": 21164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.323054075241089,
+      "learning_rate": 6.606359649122808e-07,
+      "loss": 1.1789,
+      "step": 21165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.656174659729004,
+      "learning_rate": 6.597222222222223e-07,
+      "loss": 1.4844,
+      "step": 21166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4092094898223877,
+      "learning_rate": 6.588084795321638e-07,
+      "loss": 1.4407,
+      "step": 21167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.272402763366699,
+      "learning_rate": 6.578947368421053e-07,
+      "loss": 1.3962,
+      "step": 21168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1330690383911133,
+      "learning_rate": 6.569809941520468e-07,
+      "loss": 1.3443,
+      "step": 21169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1311211585998535,
+      "learning_rate": 6.560672514619885e-07,
+      "loss": 1.6906,
+      "step": 21170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2096824645996094,
+      "learning_rate": 6.551535087719299e-07,
+      "loss": 1.1773,
+      "step": 21171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3700599670410156,
+      "learning_rate": 6.542397660818714e-07,
+      "loss": 1.1798,
+      "step": 21172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.438857316970825,
+      "learning_rate": 6.533260233918129e-07,
+      "loss": 1.3809,
+      "step": 21173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7382171154022217,
+      "learning_rate": 6.524122807017544e-07,
+      "loss": 1.6309,
+      "step": 21174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5645148754119873,
+      "learning_rate": 6.514985380116961e-07,
+      "loss": 1.4889,
+      "step": 21175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6725423336029053,
+      "learning_rate": 6.505847953216375e-07,
+      "loss": 1.6259,
+      "step": 21176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7224371433258057,
+      "learning_rate": 6.49671052631579e-07,
+      "loss": 1.7362,
+      "step": 21177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7180843353271484,
+      "learning_rate": 6.487573099415205e-07,
+      "loss": 1.5753,
+      "step": 21178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.169435739517212,
+      "learning_rate": 6.47843567251462e-07,
+      "loss": 1.2886,
+      "step": 21179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.202136993408203,
+      "learning_rate": 6.469298245614037e-07,
+      "loss": 1.0441,
+      "step": 21180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.453677177429199,
+      "learning_rate": 6.460160818713452e-07,
+      "loss": 1.4505,
+      "step": 21181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1552155017852783,
+      "learning_rate": 6.451023391812866e-07,
+      "loss": 1.631,
+      "step": 21182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5664448738098145,
+      "learning_rate": 6.441885964912281e-07,
+      "loss": 1.4776,
+      "step": 21183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4323415756225586,
+      "learning_rate": 6.432748538011696e-07,
+      "loss": 1.2847,
+      "step": 21184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6525144577026367,
+      "learning_rate": 6.423611111111111e-07,
+      "loss": 1.6097,
+      "step": 21185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5927011966705322,
+      "learning_rate": 6.414473684210528e-07,
+      "loss": 1.6461,
+      "step": 21186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2885870933532715,
+      "learning_rate": 6.405336257309943e-07,
+      "loss": 1.2727,
+      "step": 21187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.756056785583496,
+      "learning_rate": 6.396198830409357e-07,
+      "loss": 1.7118,
+      "step": 21188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2943098545074463,
+      "learning_rate": 6.387061403508772e-07,
+      "loss": 1.2692,
+      "step": 21189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.555119037628174,
+      "learning_rate": 6.377923976608187e-07,
+      "loss": 1.5742,
+      "step": 21190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.540508985519409,
+      "learning_rate": 6.368786549707604e-07,
+      "loss": 1.7175,
+      "step": 21191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.76694917678833,
+      "learning_rate": 6.359649122807019e-07,
+      "loss": 1.5165,
+      "step": 21192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.205244779586792,
+      "learning_rate": 6.350511695906433e-07,
+      "loss": 1.9853,
+      "step": 21193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4810421466827393,
+      "learning_rate": 6.341374269005848e-07,
+      "loss": 1.4128,
+      "step": 21194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1672093868255615,
+      "learning_rate": 6.332236842105263e-07,
+      "loss": 1.0877,
+      "step": 21195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.07942795753479,
+      "learning_rate": 6.32309941520468e-07,
+      "loss": 1.3307,
+      "step": 21196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3344037532806396,
+      "learning_rate": 6.313961988304095e-07,
+      "loss": 1.3736,
+      "step": 21197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7295644283294678,
+      "learning_rate": 6.30482456140351e-07,
+      "loss": 1.6631,
+      "step": 21198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.563922643661499,
+      "learning_rate": 6.295687134502924e-07,
+      "loss": 1.7085,
+      "step": 21199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4102632999420166,
+      "learning_rate": 6.286549707602339e-07,
+      "loss": 1.3278,
+      "step": 21200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0519068241119385,
+      "learning_rate": 6.277412280701754e-07,
+      "loss": 1.3619,
+      "step": 21201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0681419372558594,
+      "learning_rate": 6.268274853801171e-07,
+      "loss": 1.0956,
+      "step": 21202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4268710613250732,
+      "learning_rate": 6.259137426900586e-07,
+      "loss": 1.5631,
+      "step": 21203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9106578826904297,
+      "learning_rate": 6.25e-07,
+      "loss": 1.7056,
+      "step": 21204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.825223922729492,
+      "learning_rate": 6.240862573099415e-07,
+      "loss": 1.7892,
+      "step": 21205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3863930702209473,
+      "learning_rate": 6.231725146198831e-07,
+      "loss": 1.2898,
+      "step": 21206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5047800540924072,
+      "learning_rate": 6.222587719298246e-07,
+      "loss": 1.0862,
+      "step": 21207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1058361530303955,
+      "learning_rate": 6.213450292397661e-07,
+      "loss": 1.509,
+      "step": 21208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3964998722076416,
+      "learning_rate": 6.204312865497077e-07,
+      "loss": 1.4015,
+      "step": 21209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.713468551635742,
+      "learning_rate": 6.195175438596491e-07,
+      "loss": 1.5764,
+      "step": 21210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.607667922973633,
+      "learning_rate": 6.186038011695907e-07,
+      "loss": 1.5315,
+      "step": 21211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8999247550964355,
+      "learning_rate": 6.176900584795322e-07,
+      "loss": 1.7773,
+      "step": 21212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.58418345451355,
+      "learning_rate": 6.167763157894737e-07,
+      "loss": 1.749,
+      "step": 21213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3105428218841553,
+      "learning_rate": 6.158625730994153e-07,
+      "loss": 1.5655,
+      "step": 21214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3676347732543945,
+      "learning_rate": 6.149488304093568e-07,
+      "loss": 1.4302,
+      "step": 21215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.717668056488037,
+      "learning_rate": 6.140350877192982e-07,
+      "loss": 1.4552,
+      "step": 21216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.301077127456665,
+      "learning_rate": 6.131213450292398e-07,
+      "loss": 1.1257,
+      "step": 21217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.975855827331543,
+      "learning_rate": 6.122076023391813e-07,
+      "loss": 1.763,
+      "step": 21218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.576913833618164,
+      "learning_rate": 6.112938596491229e-07,
+      "loss": 1.2246,
+      "step": 21219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.519007921218872,
+      "learning_rate": 6.103801169590644e-07,
+      "loss": 1.6079,
+      "step": 21220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.832864284515381,
+      "learning_rate": 6.094663742690059e-07,
+      "loss": 1.8974,
+      "step": 21221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.517113208770752,
+      "learning_rate": 6.085526315789474e-07,
+      "loss": 1.4676,
+      "step": 21222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.087354898452759,
+      "learning_rate": 6.076388888888889e-07,
+      "loss": 1.4515,
+      "step": 21223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.078111410140991,
+      "learning_rate": 6.067251461988304e-07,
+      "loss": 1.4528,
+      "step": 21224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1605145931243896,
+      "learning_rate": 6.05811403508772e-07,
+      "loss": 1.4018,
+      "step": 21225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.635345220565796,
+      "learning_rate": 6.048976608187135e-07,
+      "loss": 1.4747,
+      "step": 21226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.202852964401245,
+      "learning_rate": 6.03983918128655e-07,
+      "loss": 1.4323,
+      "step": 21227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9003355503082275,
+      "learning_rate": 6.030701754385965e-07,
+      "loss": 1.4611,
+      "step": 21228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.178453207015991,
+      "learning_rate": 6.02156432748538e-07,
+      "loss": 1.1929,
+      "step": 21229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.134307622909546,
+      "learning_rate": 6.012426900584796e-07,
+      "loss": 1.1173,
+      "step": 21230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.035311460494995,
+      "learning_rate": 6.003289473684211e-07,
+      "loss": 1.2626,
+      "step": 21231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4050474166870117,
+      "learning_rate": 5.994152046783626e-07,
+      "loss": 1.4477,
+      "step": 21232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8767571449279785,
+      "learning_rate": 5.985014619883041e-07,
+      "loss": 1.8189,
+      "step": 21233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5268428325653076,
+      "learning_rate": 5.975877192982456e-07,
+      "loss": 1.5774,
+      "step": 21234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.199615955352783,
+      "learning_rate": 5.966739766081872e-07,
+      "loss": 1.2298,
+      "step": 21235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.927607774734497,
+      "learning_rate": 5.957602339181287e-07,
+      "loss": 1.1475,
+      "step": 21236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1233880519866943,
+      "learning_rate": 5.948464912280702e-07,
+      "loss": 1.5491,
+      "step": 21237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1040456295013428,
+      "learning_rate": 5.939327485380118e-07,
+      "loss": 1.1518,
+      "step": 21238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.697209596633911,
+      "learning_rate": 5.930190058479532e-07,
+      "loss": 1.7786,
+      "step": 21239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.804856061935425,
+      "learning_rate": 5.921052631578947e-07,
+      "loss": 1.8786,
+      "step": 21240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4551029205322266,
+      "learning_rate": 5.911915204678363e-07,
+      "loss": 1.7641,
+      "step": 21241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.502943515777588,
+      "learning_rate": 5.902777777777778e-07,
+      "loss": 1.2281,
+      "step": 21242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2333390712738037,
+      "learning_rate": 5.893640350877194e-07,
+      "loss": 1.3395,
+      "step": 21243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9894769191741943,
+      "learning_rate": 5.884502923976609e-07,
+      "loss": 1.2771,
+      "step": 21244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.415576696395874,
+      "learning_rate": 5.875365497076023e-07,
+      "loss": 1.3697,
+      "step": 21245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4738664627075195,
+      "learning_rate": 5.866228070175439e-07,
+      "loss": 1.6284,
+      "step": 21246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3064796924591064,
+      "learning_rate": 5.857090643274854e-07,
+      "loss": 1.5678,
+      "step": 21247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.173105239868164,
+      "learning_rate": 5.847953216374269e-07,
+      "loss": 1.5894,
+      "step": 21248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5321524143218994,
+      "learning_rate": 5.838815789473685e-07,
+      "loss": 1.5545,
+      "step": 21249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4030487537384033,
+      "learning_rate": 5.829678362573099e-07,
+      "loss": 1.4516,
+      "step": 21250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3736345767974854,
+      "learning_rate": 5.820540935672515e-07,
+      "loss": 1.3813,
+      "step": 21251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6330859661102295,
+      "learning_rate": 5.81140350877193e-07,
+      "loss": 1.5476,
+      "step": 21252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.584423065185547,
+      "learning_rate": 5.802266081871345e-07,
+      "loss": 1.9429,
+      "step": 21253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.993549346923828,
+      "learning_rate": 5.793128654970761e-07,
+      "loss": 1.5635,
+      "step": 21254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.756087303161621,
+      "learning_rate": 5.783991228070176e-07,
+      "loss": 1.5357,
+      "step": 21255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.132019281387329,
+      "learning_rate": 5.77485380116959e-07,
+      "loss": 1.5358,
+      "step": 21256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.51279354095459,
+      "learning_rate": 5.765716374269006e-07,
+      "loss": 1.439,
+      "step": 21257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3766164779663086,
+      "learning_rate": 5.756578947368421e-07,
+      "loss": 1.6815,
+      "step": 21258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7775826454162598,
+      "learning_rate": 5.747441520467837e-07,
+      "loss": 1.2902,
+      "step": 21259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.66326904296875,
+      "learning_rate": 5.738304093567252e-07,
+      "loss": 1.63,
+      "step": 21260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.625518560409546,
+      "learning_rate": 5.729166666666667e-07,
+      "loss": 1.59,
+      "step": 21261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.728339195251465,
+      "learning_rate": 5.720029239766082e-07,
+      "loss": 1.7636,
+      "step": 21262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7109298706054688,
+      "learning_rate": 5.710891812865497e-07,
+      "loss": 1.5477,
+      "step": 21263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7533743381500244,
+      "learning_rate": 5.701754385964912e-07,
+      "loss": 1.7577,
+      "step": 21264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5814261436462402,
+      "learning_rate": 5.692616959064328e-07,
+      "loss": 1.4727,
+      "step": 21265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4781181812286377,
+      "learning_rate": 5.683479532163743e-07,
+      "loss": 1.3966,
+      "step": 21266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4028663635253906,
+      "learning_rate": 5.674342105263159e-07,
+      "loss": 1.3061,
+      "step": 21267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7797436714172363,
+      "learning_rate": 5.665204678362573e-07,
+      "loss": 1.7361,
+      "step": 21268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.802527904510498,
+      "learning_rate": 5.656067251461988e-07,
+      "loss": 1.4208,
+      "step": 21269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1940906047821045,
+      "learning_rate": 5.646929824561404e-07,
+      "loss": 1.3985,
+      "step": 21270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.477145195007324,
+      "learning_rate": 5.637792397660819e-07,
+      "loss": 1.3882,
+      "step": 21271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2845184803009033,
+      "learning_rate": 5.628654970760234e-07,
+      "loss": 1.3611,
+      "step": 21272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.223020076751709,
+      "learning_rate": 5.619517543859649e-07,
+      "loss": 1.5651,
+      "step": 21273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2369868755340576,
+      "learning_rate": 5.610380116959064e-07,
+      "loss": 1.8628,
+      "step": 21274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5907418727874756,
+      "learning_rate": 5.60124269005848e-07,
+      "loss": 1.513,
+      "step": 21275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1230781078338623,
+      "learning_rate": 5.592105263157895e-07,
+      "loss": 1.5472,
+      "step": 21276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7369229793548584,
+      "learning_rate": 5.58296783625731e-07,
+      "loss": 1.5754,
+      "step": 21277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.440980911254883,
+      "learning_rate": 5.573830409356726e-07,
+      "loss": 1.352,
+      "step": 21278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0544323921203613,
+      "learning_rate": 5.56469298245614e-07,
+      "loss": 1.2131,
+      "step": 21279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3433218002319336,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 1.4613,
+      "step": 21280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8163068294525146,
+      "learning_rate": 5.546418128654971e-07,
+      "loss": 1.5395,
+      "step": 21281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1712141036987305,
+      "learning_rate": 5.537280701754386e-07,
+      "loss": 1.2855,
+      "step": 21282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4281227588653564,
+      "learning_rate": 5.528143274853802e-07,
+      "loss": 1.6236,
+      "step": 21283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2229297161102295,
+      "learning_rate": 5.519005847953217e-07,
+      "loss": 1.194,
+      "step": 21284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.346478223800659,
+      "learning_rate": 5.509868421052631e-07,
+      "loss": 1.1425,
+      "step": 21285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3411524295806885,
+      "learning_rate": 5.500730994152047e-07,
+      "loss": 1.5585,
+      "step": 21286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5008468627929688,
+      "learning_rate": 5.491593567251462e-07,
+      "loss": 1.5152,
+      "step": 21287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9162003993988037,
+      "learning_rate": 5.482456140350878e-07,
+      "loss": 1.4428,
+      "step": 21288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3443400859832764,
+      "learning_rate": 5.473318713450293e-07,
+      "loss": 1.237,
+      "step": 21289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3542728424072266,
+      "learning_rate": 5.464181286549707e-07,
+      "loss": 1.5959,
+      "step": 21290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.521462917327881,
+      "learning_rate": 5.455043859649123e-07,
+      "loss": 1.5285,
+      "step": 21291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2048227787017822,
+      "learning_rate": 5.445906432748538e-07,
+      "loss": 1.2346,
+      "step": 21292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.186466932296753,
+      "learning_rate": 5.436769005847953e-07,
+      "loss": 1.2705,
+      "step": 21293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6075170040130615,
+      "learning_rate": 5.427631578947369e-07,
+      "loss": 1.7517,
+      "step": 21294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.924424409866333,
+      "learning_rate": 5.418494152046784e-07,
+      "loss": 1.6082,
+      "step": 21295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8414360284805298,
+      "learning_rate": 5.409356725146199e-07,
+      "loss": 1.4802,
+      "step": 21296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0871119499206543,
+      "learning_rate": 5.400219298245614e-07,
+      "loss": 1.157,
+      "step": 21297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9287991523742676,
+      "learning_rate": 5.391081871345029e-07,
+      "loss": 1.2402,
+      "step": 21298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.407108783721924,
+      "learning_rate": 5.381944444444445e-07,
+      "loss": 1.1721,
+      "step": 21299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7660562992095947,
+      "learning_rate": 5.37280701754386e-07,
+      "loss": 1.4712,
+      "step": 21300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.276533603668213,
+      "learning_rate": 5.363669590643276e-07,
+      "loss": 1.542,
+      "step": 21301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9757120609283447,
+      "learning_rate": 5.35453216374269e-07,
+      "loss": 1.769,
+      "step": 21302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1524946689605713,
+      "learning_rate": 5.345394736842105e-07,
+      "loss": 1.5278,
+      "step": 21303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.123539686203003,
+      "learning_rate": 5.336257309941521e-07,
+      "loss": 1.2376,
+      "step": 21304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5855915546417236,
+      "learning_rate": 5.327119883040936e-07,
+      "loss": 1.3828,
+      "step": 21305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.553408622741699,
+      "learning_rate": 5.317982456140352e-07,
+      "loss": 1.4197,
+      "step": 21306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.449066400527954,
+      "learning_rate": 5.308845029239767e-07,
+      "loss": 1.1929,
+      "step": 21307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.897707939147949,
+      "learning_rate": 5.299707602339181e-07,
+      "loss": 1.8193,
+      "step": 21308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1289820671081543,
+      "learning_rate": 5.290570175438597e-07,
+      "loss": 1.0402,
+      "step": 21309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2334847450256348,
+      "learning_rate": 5.281432748538012e-07,
+      "loss": 1.1548,
+      "step": 21310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.753495216369629,
+      "learning_rate": 5.272295321637427e-07,
+      "loss": 1.5659,
+      "step": 21311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.070319652557373,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 1.6064,
+      "step": 21312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2672126293182373,
+      "learning_rate": 5.254020467836257e-07,
+      "loss": 1.6717,
+      "step": 21313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0927481651306152,
+      "learning_rate": 5.244883040935673e-07,
+      "loss": 1.2328,
+      "step": 21314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.993322491645813,
+      "learning_rate": 5.235745614035088e-07,
+      "loss": 1.0321,
+      "step": 21315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.333232879638672,
+      "learning_rate": 5.226608187134503e-07,
+      "loss": 1.4391,
+      "step": 21316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9468649625778198,
+      "learning_rate": 5.217470760233919e-07,
+      "loss": 0.9694,
+      "step": 21317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5923357009887695,
+      "learning_rate": 5.208333333333334e-07,
+      "loss": 1.7485,
+      "step": 21318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.575582981109619,
+      "learning_rate": 5.199195906432749e-07,
+      "loss": 1.4998,
+      "step": 21319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4434428215026855,
+      "learning_rate": 5.190058479532164e-07,
+      "loss": 1.0431,
+      "step": 21320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.092252016067505,
+      "learning_rate": 5.180921052631579e-07,
+      "loss": 1.2572,
+      "step": 21321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8246819972991943,
+      "learning_rate": 5.171783625730995e-07,
+      "loss": 1.5075,
+      "step": 21322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1640796661376953,
+      "learning_rate": 5.16264619883041e-07,
+      "loss": 1.2193,
+      "step": 21323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0359182357788086,
+      "learning_rate": 5.153508771929825e-07,
+      "loss": 1.1723,
+      "step": 21324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.686480760574341,
+      "learning_rate": 5.14437134502924e-07,
+      "loss": 1.4643,
+      "step": 21325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.371462106704712,
+      "learning_rate": 5.135233918128655e-07,
+      "loss": 1.5523,
+      "step": 21326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.380967617034912,
+      "learning_rate": 5.126096491228071e-07,
+      "loss": 1.4725,
+      "step": 21327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.913560628890991,
+      "learning_rate": 5.116959064327486e-07,
+      "loss": 1.6983,
+      "step": 21328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3226492404937744,
+      "learning_rate": 5.107821637426901e-07,
+      "loss": 1.308,
+      "step": 21329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4198169708251953,
+      "learning_rate": 5.098684210526317e-07,
+      "loss": 1.3718,
+      "step": 21330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.759115695953369,
+      "learning_rate": 5.089546783625731e-07,
+      "loss": 1.6364,
+      "step": 21331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.057020902633667,
+      "learning_rate": 5.080409356725147e-07,
+      "loss": 1.1292,
+      "step": 21332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.641892433166504,
+      "learning_rate": 5.071271929824562e-07,
+      "loss": 1.6945,
+      "step": 21333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6207499504089355,
+      "learning_rate": 5.062134502923977e-07,
+      "loss": 1.5469,
+      "step": 21334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.620837450027466,
+      "learning_rate": 5.052997076023393e-07,
+      "loss": 1.5423,
+      "step": 21335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.169389009475708,
+      "learning_rate": 5.043859649122807e-07,
+      "loss": 0.9575,
+      "step": 21336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.076967239379883,
+      "learning_rate": 5.034722222222223e-07,
+      "loss": 1.2637,
+      "step": 21337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2163496017456055,
+      "learning_rate": 5.025584795321638e-07,
+      "loss": 1.2859,
+      "step": 21338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2205002307891846,
+      "learning_rate": 5.016447368421053e-07,
+      "loss": 1.2709,
+      "step": 21339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.358802556991577,
+      "learning_rate": 5.007309941520469e-07,
+      "loss": 1.3443,
+      "step": 21340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.749424457550049,
+      "learning_rate": 4.998172514619884e-07,
+      "loss": 1.5996,
+      "step": 21341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.101240634918213,
+      "learning_rate": 4.989035087719298e-07,
+      "loss": 1.573,
+      "step": 21342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.206268548965454,
+      "learning_rate": 4.979897660818714e-07,
+      "loss": 1.2254,
+      "step": 21343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5201833248138428,
+      "learning_rate": 4.970760233918129e-07,
+      "loss": 1.5553,
+      "step": 21344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.461540699005127,
+      "learning_rate": 4.961622807017545e-07,
+      "loss": 1.2671,
+      "step": 21345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.231093645095825,
+      "learning_rate": 4.95248538011696e-07,
+      "loss": 1.3086,
+      "step": 21346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.687133550643921,
+      "learning_rate": 4.943347953216375e-07,
+      "loss": 1.4318,
+      "step": 21347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2663869857788086,
+      "learning_rate": 4.93421052631579e-07,
+      "loss": 1.4313,
+      "step": 21348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.47636079788208,
+      "learning_rate": 4.925073099415205e-07,
+      "loss": 1.6658,
+      "step": 21349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.639512300491333,
+      "learning_rate": 4.915935672514621e-07,
+      "loss": 1.379,
+      "step": 21350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8482943773269653,
+      "learning_rate": 4.906798245614036e-07,
+      "loss": 1.0949,
+      "step": 21351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.417785882949829,
+      "learning_rate": 4.897660818713451e-07,
+      "loss": 1.3603,
+      "step": 21352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5154340267181396,
+      "learning_rate": 4.888523391812867e-07,
+      "loss": 1.5561,
+      "step": 21353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5813817977905273,
+      "learning_rate": 4.879385964912281e-07,
+      "loss": 1.422,
+      "step": 21354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.921717643737793,
+      "learning_rate": 4.870248538011696e-07,
+      "loss": 1.5667,
+      "step": 21355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1171085834503174,
+      "learning_rate": 4.861111111111112e-07,
+      "loss": 1.178,
+      "step": 21356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.028843879699707,
+      "learning_rate": 4.851973684210527e-07,
+      "loss": 1.6814,
+      "step": 21357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7907817363739014,
+      "learning_rate": 4.842836257309943e-07,
+      "loss": 1.4656,
+      "step": 21358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1998071670532227,
+      "learning_rate": 4.833698830409357e-07,
+      "loss": 1.2545,
+      "step": 21359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.153019428253174,
+      "learning_rate": 4.824561403508772e-07,
+      "loss": 1.4465,
+      "step": 21360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3625407218933105,
+      "learning_rate": 4.815423976608188e-07,
+      "loss": 1.1807,
+      "step": 21361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7972559928894043,
+      "learning_rate": 4.806286549707603e-07,
+      "loss": 1.836,
+      "step": 21362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.264033555984497,
+      "learning_rate": 4.797149122807019e-07,
+      "loss": 1.2631,
+      "step": 21363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5406062602996826,
+      "learning_rate": 4.788011695906434e-07,
+      "loss": 1.7846,
+      "step": 21364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6139519214630127,
+      "learning_rate": 4.778874269005848e-07,
+      "loss": 1.4188,
+      "step": 21365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.253114938735962,
+      "learning_rate": 4.769736842105264e-07,
+      "loss": 1.4341,
+      "step": 21366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9644761085510254,
+      "learning_rate": 4.7605994152046785e-07,
+      "loss": 2.1597,
+      "step": 21367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9458250999450684,
+      "learning_rate": 4.751461988304094e-07,
+      "loss": 1.1462,
+      "step": 21368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5429468154907227,
+      "learning_rate": 4.742324561403509e-07,
+      "loss": 1.4248,
+      "step": 21369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.737222909927368,
+      "learning_rate": 4.7331871345029245e-07,
+      "loss": 1.5575,
+      "step": 21370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5573644638061523,
+      "learning_rate": 4.72404970760234e-07,
+      "loss": 1.4416,
+      "step": 21371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0474843978881836,
+      "learning_rate": 4.7149122807017547e-07,
+      "loss": 1.7746,
+      "step": 21372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5310940742492676,
+      "learning_rate": 4.70577485380117e-07,
+      "loss": 1.6958,
+      "step": 21373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.304851770401001,
+      "learning_rate": 4.6966374269005853e-07,
+      "loss": 1.4092,
+      "step": 21374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.897650957107544,
+      "learning_rate": 4.6875000000000006e-07,
+      "loss": 1.7421,
+      "step": 21375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9599082469940186,
+      "learning_rate": 4.6783625730994155e-07,
+      "loss": 1.2202,
+      "step": 21376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.445321559906006,
+      "learning_rate": 4.669225146198831e-07,
+      "loss": 1.6289,
+      "step": 21377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5954153537750244,
+      "learning_rate": 4.660087719298246e-07,
+      "loss": 1.7224,
+      "step": 21378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7045388221740723,
+      "learning_rate": 4.650950292397661e-07,
+      "loss": 1.3045,
+      "step": 21379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4525864124298096,
+      "learning_rate": 4.641812865497076e-07,
+      "loss": 1.1973,
+      "step": 21380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1427459716796875,
+      "learning_rate": 4.6326754385964916e-07,
+      "loss": 1.4008,
+      "step": 21381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4391703605651855,
+      "learning_rate": 4.623538011695907e-07,
+      "loss": 1.5018,
+      "step": 21382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7215330600738525,
+      "learning_rate": 4.614400584795322e-07,
+      "loss": 1.3754,
+      "step": 21383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6212809085845947,
+      "learning_rate": 4.605263157894737e-07,
+      "loss": 2.4254,
+      "step": 21384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.60142183303833,
+      "learning_rate": 4.5961257309941524e-07,
+      "loss": 1.4994,
+      "step": 21385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.282442808151245,
+      "learning_rate": 4.5869883040935677e-07,
+      "loss": 1.5251,
+      "step": 21386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4779841899871826,
+      "learning_rate": 4.5778508771929825e-07,
+      "loss": 1.4014,
+      "step": 21387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3626699447631836,
+      "learning_rate": 4.5687134502923984e-07,
+      "loss": 1.4151,
+      "step": 21388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4747474193573,
+      "learning_rate": 4.559576023391813e-07,
+      "loss": 1.2878,
+      "step": 21389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4712822437286377,
+      "learning_rate": 4.5504385964912285e-07,
+      "loss": 1.5612,
+      "step": 21390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.149538278579712,
+      "learning_rate": 4.541301169590644e-07,
+      "loss": 1.0581,
+      "step": 21391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0842740535736084,
+      "learning_rate": 4.5321637426900587e-07,
+      "loss": 1.6011,
+      "step": 21392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4301912784576416,
+      "learning_rate": 4.523026315789474e-07,
+      "loss": 1.8835,
+      "step": 21393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4279110431671143,
+      "learning_rate": 4.5138888888888893e-07,
+      "loss": 1.3297,
+      "step": 21394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.323992967605591,
+      "learning_rate": 4.504751461988304e-07,
+      "loss": 1.4544,
+      "step": 21395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3362839221954346,
+      "learning_rate": 4.49561403508772e-07,
+      "loss": 1.3196,
+      "step": 21396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6397390365600586,
+      "learning_rate": 4.486476608187135e-07,
+      "loss": 1.4737,
+      "step": 21397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.505159854888916,
+      "learning_rate": 4.47733918128655e-07,
+      "loss": 1.5897,
+      "step": 21398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9446861743927,
+      "learning_rate": 4.4682017543859654e-07,
+      "loss": 1.6027,
+      "step": 21399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6863820552825928,
+      "learning_rate": 4.45906432748538e-07,
+      "loss": 1.7499,
+      "step": 21400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4368550777435303,
+      "learning_rate": 4.449926900584796e-07,
+      "loss": 1.3044,
+      "step": 21401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0853633880615234,
+      "learning_rate": 4.440789473684211e-07,
+      "loss": 1.0257,
+      "step": 21402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6459720134735107,
+      "learning_rate": 4.4316520467836257e-07,
+      "loss": 1.2904,
+      "step": 21403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.350707530975342,
+      "learning_rate": 4.4225146198830416e-07,
+      "loss": 1.3886,
+      "step": 21404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.503654956817627,
+      "learning_rate": 4.4133771929824564e-07,
+      "loss": 1.2908,
+      "step": 21405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4214818477630615,
+      "learning_rate": 4.4042397660818717e-07,
+      "loss": 1.3918,
+      "step": 21406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7179768085479736,
+      "learning_rate": 4.395102339181287e-07,
+      "loss": 1.5781,
+      "step": 21407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6033406257629395,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 1.4237,
+      "step": 21408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3348071575164795,
+      "learning_rate": 4.3768274853801177e-07,
+      "loss": 1.3787,
+      "step": 21409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6547460556030273,
+      "learning_rate": 4.3676900584795325e-07,
+      "loss": 1.6663,
+      "step": 21410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.213264226913452,
+      "learning_rate": 4.3585526315789473e-07,
+      "loss": 1.3564,
+      "step": 21411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.111706256866455,
+      "learning_rate": 4.349415204678363e-07,
+      "loss": 1.1819,
+      "step": 21412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3184943199157715,
+      "learning_rate": 4.340277777777778e-07,
+      "loss": 1.4762,
+      "step": 21413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0893561840057373,
+      "learning_rate": 4.331140350877194e-07,
+      "loss": 1.1097,
+      "step": 21414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.720472574234009,
+      "learning_rate": 4.3220029239766087e-07,
+      "loss": 1.2836,
+      "step": 21415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1884241104125977,
+      "learning_rate": 4.3128654970760235e-07,
+      "loss": 1.3427,
+      "step": 21416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.313781261444092,
+      "learning_rate": 4.3037280701754393e-07,
+      "loss": 1.5831,
+      "step": 21417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0393035411834717,
+      "learning_rate": 4.294590643274854e-07,
+      "loss": 1.7652,
+      "step": 21418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7104103565216064,
+      "learning_rate": 4.285453216374269e-07,
+      "loss": 1.3468,
+      "step": 21419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5875844955444336,
+      "learning_rate": 4.276315789473685e-07,
+      "loss": 1.4622,
+      "step": 21420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.263181447982788,
+      "learning_rate": 4.2671783625730996e-07,
+      "loss": 1.2788,
+      "step": 21421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4853200912475586,
+      "learning_rate": 4.2580409356725154e-07,
+      "loss": 1.3999,
+      "step": 21422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0248308181762695,
+      "learning_rate": 4.24890350877193e-07,
+      "loss": 1.1605,
+      "step": 21423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4134953022003174,
+      "learning_rate": 4.239766081871345e-07,
+      "loss": 1.2748,
+      "step": 21424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.461707592010498,
+      "learning_rate": 4.230628654970761e-07,
+      "loss": 1.4987,
+      "step": 21425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4300191402435303,
+      "learning_rate": 4.2214912280701757e-07,
+      "loss": 1.5049,
+      "step": 21426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1975910663604736,
+      "learning_rate": 4.2123538011695905e-07,
+      "loss": 1.432,
+      "step": 21427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.80708909034729,
+      "learning_rate": 4.2032163742690064e-07,
+      "loss": 1.6993,
+      "step": 21428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.55376935005188,
+      "learning_rate": 4.194078947368421e-07,
+      "loss": 1.3388,
+      "step": 21429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.279430389404297,
+      "learning_rate": 4.184941520467837e-07,
+      "loss": 1.4807,
+      "step": 21430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3685193061828613,
+      "learning_rate": 4.175804093567252e-07,
+      "loss": 1.3948,
+      "step": 21431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7064759731292725,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 1.5019,
+      "step": 21432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.95697021484375,
+      "learning_rate": 4.1575292397660825e-07,
+      "loss": 1.4666,
+      "step": 21433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.016218662261963,
+      "learning_rate": 4.1483918128654973e-07,
+      "loss": 1.136,
+      "step": 21434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7137787342071533,
+      "learning_rate": 4.139254385964912e-07,
+      "loss": 1.5985,
+      "step": 21435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0861189365386963,
+      "learning_rate": 4.130116959064328e-07,
+      "loss": 1.3585,
+      "step": 21436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.394803524017334,
+      "learning_rate": 4.120979532163743e-07,
+      "loss": 1.4851,
+      "step": 21437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5647473335266113,
+      "learning_rate": 4.1118421052631586e-07,
+      "loss": 1.3987,
+      "step": 21438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5200693607330322,
+      "learning_rate": 4.1027046783625735e-07,
+      "loss": 1.3437,
+      "step": 21439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3138952255249023,
+      "learning_rate": 4.093567251461988e-07,
+      "loss": 1.3008,
+      "step": 21440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1249828338623047,
+      "learning_rate": 4.084429824561404e-07,
+      "loss": 1.9427,
+      "step": 21441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4760403633117676,
+      "learning_rate": 4.075292397660819e-07,
+      "loss": 1.608,
+      "step": 21442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4733240604400635,
+      "learning_rate": 4.0661549707602337e-07,
+      "loss": 1.4109,
+      "step": 21443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0422544479370117,
+      "learning_rate": 4.0570175438596496e-07,
+      "loss": 1.6727,
+      "step": 21444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2354533672332764,
+      "learning_rate": 4.0478801169590644e-07,
+      "loss": 1.5079,
+      "step": 21445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.71097731590271,
+      "learning_rate": 4.03874269005848e-07,
+      "loss": 1.3501,
+      "step": 21446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5684614181518555,
+      "learning_rate": 4.029605263157895e-07,
+      "loss": 1.1978,
+      "step": 21447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.411932945251465,
+      "learning_rate": 4.02046783625731e-07,
+      "loss": 1.3163,
+      "step": 21448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.205204963684082,
+      "learning_rate": 4.0113304093567257e-07,
+      "loss": 1.2961,
+      "step": 21449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2359237670898438,
+      "learning_rate": 4.0021929824561405e-07,
+      "loss": 1.4245,
+      "step": 21450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4450221061706543,
+      "learning_rate": 3.9930555555555553e-07,
+      "loss": 1.2156,
+      "step": 21451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5066537857055664,
+      "learning_rate": 3.983918128654971e-07,
+      "loss": 1.4436,
+      "step": 21452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.614686965942383,
+      "learning_rate": 3.974780701754386e-07,
+      "loss": 1.4559,
+      "step": 21453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.862922430038452,
+      "learning_rate": 3.965643274853802e-07,
+      "loss": 1.2356,
+      "step": 21454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.165320873260498,
+      "learning_rate": 3.9565058479532167e-07,
+      "loss": 1.229,
+      "step": 21455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0694103240966797,
+      "learning_rate": 3.9473684210526315e-07,
+      "loss": 1.8504,
+      "step": 21456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7930495738983154,
+      "learning_rate": 3.9382309941520473e-07,
+      "loss": 1.5664,
+      "step": 21457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.733171224594116,
+      "learning_rate": 3.929093567251462e-07,
+      "loss": 1.3907,
+      "step": 21458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4072113037109375,
+      "learning_rate": 3.9199561403508775e-07,
+      "loss": 1.4389,
+      "step": 21459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.456568479537964,
+      "learning_rate": 3.910818713450293e-07,
+      "loss": 1.1268,
+      "step": 21460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4227991104125977,
+      "learning_rate": 3.9016812865497076e-07,
+      "loss": 1.2418,
+      "step": 21461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7192065715789795,
+      "learning_rate": 3.8925438596491235e-07,
+      "loss": 1.4611,
+      "step": 21462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.58845853805542,
+      "learning_rate": 3.883406432748538e-07,
+      "loss": 1.2828,
+      "step": 21463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.530071496963501,
+      "learning_rate": 3.874269005847953e-07,
+      "loss": 1.4626,
+      "step": 21464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.310565710067749,
+      "learning_rate": 3.865131578947369e-07,
+      "loss": 1.6887,
+      "step": 21465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.856572389602661,
+      "learning_rate": 3.8559941520467837e-07,
+      "loss": 1.5781,
+      "step": 21466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2998385429382324,
+      "learning_rate": 3.846856725146199e-07,
+      "loss": 1.3162,
+      "step": 21467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3086626529693604,
+      "learning_rate": 3.8377192982456144e-07,
+      "loss": 1.3285,
+      "step": 21468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.350153923034668,
+      "learning_rate": 3.828581871345029e-07,
+      "loss": 1.0408,
+      "step": 21469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5469372272491455,
+      "learning_rate": 3.819444444444445e-07,
+      "loss": 1.2501,
+      "step": 21470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6200902462005615,
+      "learning_rate": 3.81030701754386e-07,
+      "loss": 1.5383,
+      "step": 21471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6859564781188965,
+      "learning_rate": 3.801169590643275e-07,
+      "loss": 1.3597,
+      "step": 21472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.608724355697632,
+      "learning_rate": 3.7920321637426905e-07,
+      "loss": 1.6063,
+      "step": 21473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.513902425765991,
+      "learning_rate": 3.7828947368421053e-07,
+      "loss": 1.6831,
+      "step": 21474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.084711790084839,
+      "learning_rate": 3.7737573099415207e-07,
+      "loss": 1.3641,
+      "step": 21475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.608180046081543,
+      "learning_rate": 3.764619883040936e-07,
+      "loss": 1.5525,
+      "step": 21476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6946325302124023,
+      "learning_rate": 3.7554824561403513e-07,
+      "loss": 1.6512,
+      "step": 21477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1995556354522705,
+      "learning_rate": 3.7463450292397667e-07,
+      "loss": 1.438,
+      "step": 21478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.672640800476074,
+      "learning_rate": 3.7372076023391815e-07,
+      "loss": 1.4992,
+      "step": 21479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0834567546844482,
+      "learning_rate": 3.728070175438597e-07,
+      "loss": 1.8412,
+      "step": 21480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4913885593414307,
+      "learning_rate": 3.718932748538012e-07,
+      "loss": 1.4045,
+      "step": 21481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.632122278213501,
+      "learning_rate": 3.709795321637427e-07,
+      "loss": 1.6801,
+      "step": 21482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.522487163543701,
+      "learning_rate": 3.700657894736842e-07,
+      "loss": 1.5505,
+      "step": 21483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.715156078338623,
+      "learning_rate": 3.6915204678362576e-07,
+      "loss": 1.4406,
+      "step": 21484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5625977516174316,
+      "learning_rate": 3.682383040935673e-07,
+      "loss": 1.9462,
+      "step": 21485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.649083137512207,
+      "learning_rate": 3.673245614035088e-07,
+      "loss": 1.6429,
+      "step": 21486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.808124303817749,
+      "learning_rate": 3.664108187134503e-07,
+      "loss": 1.4861,
+      "step": 21487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.219233989715576,
+      "learning_rate": 3.6549707602339184e-07,
+      "loss": 1.1067,
+      "step": 21488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.316075325012207,
+      "learning_rate": 3.6458333333333337e-07,
+      "loss": 1.0626,
+      "step": 21489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.489638328552246,
+      "learning_rate": 3.636695906432749e-07,
+      "loss": 1.6887,
+      "step": 21490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.041841506958008,
+      "learning_rate": 3.627558479532164e-07,
+      "loss": 1.0534,
+      "step": 21491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.508507251739502,
+      "learning_rate": 3.618421052631579e-07,
+      "loss": 1.294,
+      "step": 21492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8644986152648926,
+      "learning_rate": 3.6092836257309945e-07,
+      "loss": 1.8192,
+      "step": 21493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.945457935333252,
+      "learning_rate": 3.60014619883041e-07,
+      "loss": 1.7769,
+      "step": 21494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0776398181915283,
+      "learning_rate": 3.5910087719298247e-07,
+      "loss": 1.469,
+      "step": 21495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3763484954833984,
+      "learning_rate": 3.58187134502924e-07,
+      "loss": 1.4348,
+      "step": 21496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.156381845474243,
+      "learning_rate": 3.5727339181286553e-07,
+      "loss": 1.0087,
+      "step": 21497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.604731798171997,
+      "learning_rate": 3.5635964912280707e-07,
+      "loss": 1.5022,
+      "step": 21498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.230449676513672,
+      "learning_rate": 3.5544590643274855e-07,
+      "loss": 1.2949,
+      "step": 21499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9580721855163574,
+      "learning_rate": 3.545321637426901e-07,
+      "loss": 1.5024,
+      "step": 21500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8252794742584229,
+      "learning_rate": 3.536184210526316e-07,
+      "loss": 1.1066,
+      "step": 21501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5611860752105713,
+      "learning_rate": 3.5270467836257315e-07,
+      "loss": 1.7095,
+      "step": 21502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9705168008804321,
+      "learning_rate": 3.517909356725147e-07,
+      "loss": 1.1744,
+      "step": 21503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8402488231658936,
+      "learning_rate": 3.5087719298245616e-07,
+      "loss": 1.6298,
+      "step": 21504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.715398073196411,
+      "learning_rate": 3.499634502923977e-07,
+      "loss": 1.4994,
+      "step": 21505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7265098094940186,
+      "learning_rate": 3.490497076023392e-07,
+      "loss": 1.6406,
+      "step": 21506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4895832538604736,
+      "learning_rate": 3.481359649122807e-07,
+      "loss": 1.4847,
+      "step": 21507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4097652435302734,
+      "learning_rate": 3.472222222222223e-07,
+      "loss": 1.6743,
+      "step": 21508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4485232830047607,
+      "learning_rate": 3.4630847953216377e-07,
+      "loss": 1.5051,
+      "step": 21509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.226696014404297,
+      "learning_rate": 3.453947368421053e-07,
+      "loss": 1.2468,
+      "step": 21510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7671988010406494,
+      "learning_rate": 3.4448099415204684e-07,
+      "loss": 1.7171,
+      "step": 21511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9140262603759766,
+      "learning_rate": 3.435672514619883e-07,
+      "loss": 1.6855,
+      "step": 21512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.377332925796509,
+      "learning_rate": 3.4265350877192985e-07,
+      "loss": 1.2705,
+      "step": 21513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.304612636566162,
+      "learning_rate": 3.417397660818714e-07,
+      "loss": 1.3536,
+      "step": 21514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.566892623901367,
+      "learning_rate": 3.4082602339181287e-07,
+      "loss": 1.5506,
+      "step": 21515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3831448554992676,
+      "learning_rate": 3.3991228070175445e-07,
+      "loss": 1.4178,
+      "step": 21516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1551029682159424,
+      "learning_rate": 3.3899853801169593e-07,
+      "loss": 1.3326,
+      "step": 21517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.426229476928711,
+      "learning_rate": 3.3808479532163747e-07,
+      "loss": 1.5246,
+      "step": 21518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.704036235809326,
+      "learning_rate": 3.37171052631579e-07,
+      "loss": 1.855,
+      "step": 21519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.420884370803833,
+      "learning_rate": 3.362573099415205e-07,
+      "loss": 1.3875,
+      "step": 21520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3703503608703613,
+      "learning_rate": 3.3534356725146207e-07,
+      "loss": 1.2768,
+      "step": 21521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0416529178619385,
+      "learning_rate": 3.3442982456140355e-07,
+      "loss": 1.2093,
+      "step": 21522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2721588611602783,
+      "learning_rate": 3.33516081871345e-07,
+      "loss": 1.3535,
+      "step": 21523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.660243034362793,
+      "learning_rate": 3.326023391812866e-07,
+      "loss": 1.6417,
+      "step": 21524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5419819355010986,
+      "learning_rate": 3.316885964912281e-07,
+      "loss": 1.2873,
+      "step": 21525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4907631874084473,
+      "learning_rate": 3.307748538011696e-07,
+      "loss": 1.4832,
+      "step": 21526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4055516719818115,
+      "learning_rate": 3.2986111111111116e-07,
+      "loss": 1.2394,
+      "step": 21527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.320819139480591,
+      "learning_rate": 3.2894736842105264e-07,
+      "loss": 1.4551,
+      "step": 21528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6426544189453125,
+      "learning_rate": 3.280336257309942e-07,
+      "loss": 1.7377,
+      "step": 21529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.385921001434326,
+      "learning_rate": 3.271198830409357e-07,
+      "loss": 1.2814,
+      "step": 21530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1955654621124268,
+      "learning_rate": 3.262061403508772e-07,
+      "loss": 1.1279,
+      "step": 21531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7468092441558838,
+      "learning_rate": 3.2529239766081877e-07,
+      "loss": 1.1204,
+      "step": 21532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4426352977752686,
+      "learning_rate": 3.2437865497076025e-07,
+      "loss": 1.1957,
+      "step": 21533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6317219734191895,
+      "learning_rate": 3.2346491228070184e-07,
+      "loss": 1.633,
+      "step": 21534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.758361339569092,
+      "learning_rate": 3.225511695906433e-07,
+      "loss": 1.4298,
+      "step": 21535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.478671073913574,
+      "learning_rate": 3.216374269005848e-07,
+      "loss": 1.3441,
+      "step": 21536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.797008752822876,
+      "learning_rate": 3.207236842105264e-07,
+      "loss": 1.4561,
+      "step": 21537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3420891761779785,
+      "learning_rate": 3.1980994152046787e-07,
+      "loss": 1.3968,
+      "step": 21538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5412073135375977,
+      "learning_rate": 3.1889619883040935e-07,
+      "loss": 1.3288,
+      "step": 21539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.989206075668335,
+      "learning_rate": 3.1798245614035093e-07,
+      "loss": 1.7008,
+      "step": 21540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6354286670684814,
+      "learning_rate": 3.170687134502924e-07,
+      "loss": 1.4587,
+      "step": 21541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5820438861846924,
+      "learning_rate": 3.16154970760234e-07,
+      "loss": 1.4504,
+      "step": 21542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5593271255493164,
+      "learning_rate": 3.152412280701755e-07,
+      "loss": 1.4127,
+      "step": 21543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6730329990386963,
+      "learning_rate": 3.1432748538011696e-07,
+      "loss": 1.6183,
+      "step": 21544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.605602979660034,
+      "learning_rate": 3.1341374269005855e-07,
+      "loss": 1.7573,
+      "step": 21545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5214269161224365,
+      "learning_rate": 3.125e-07,
+      "loss": 1.4065,
+      "step": 21546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2905027866363525,
+      "learning_rate": 3.1158625730994156e-07,
+      "loss": 1.3976,
+      "step": 21547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.52972149848938,
+      "learning_rate": 3.1067251461988304e-07,
+      "loss": 1.5622,
+      "step": 21548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4207072257995605,
+      "learning_rate": 3.0975877192982457e-07,
+      "loss": 1.432,
+      "step": 21549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.558323383331299,
+      "learning_rate": 3.088450292397661e-07,
+      "loss": 1.4171,
+      "step": 21550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.329927921295166,
+      "learning_rate": 3.0793128654970764e-07,
+      "loss": 1.3306,
+      "step": 21551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.42118501663208,
+      "learning_rate": 3.070175438596491e-07,
+      "loss": 1.568,
+      "step": 21552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2384703159332275,
+      "learning_rate": 3.0610380116959065e-07,
+      "loss": 1.2148,
+      "step": 21553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0438790321350098,
+      "learning_rate": 3.051900584795322e-07,
+      "loss": 1.3395,
+      "step": 21554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.43713641166687,
+      "learning_rate": 3.042763157894737e-07,
+      "loss": 1.5298,
+      "step": 21555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.054957866668701,
+      "learning_rate": 3.033625730994152e-07,
+      "loss": 1.017,
+      "step": 21556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.355942487716675,
+      "learning_rate": 3.0244883040935673e-07,
+      "loss": 1.3356,
+      "step": 21557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.402902126312256,
+      "learning_rate": 3.0153508771929827e-07,
+      "loss": 1.3847,
+      "step": 21558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2585768699645996,
+      "learning_rate": 3.006213450292398e-07,
+      "loss": 1.233,
+      "step": 21559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3668346405029297,
+      "learning_rate": 2.997076023391813e-07,
+      "loss": 1.2225,
+      "step": 21560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.551835775375366,
+      "learning_rate": 2.987938596491228e-07,
+      "loss": 1.3908,
+      "step": 21561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5131888389587402,
+      "learning_rate": 2.9788011695906435e-07,
+      "loss": 1.557,
+      "step": 21562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3254904747009277,
+      "learning_rate": 2.969663742690059e-07,
+      "loss": 1.2986,
+      "step": 21563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6066741943359375,
+      "learning_rate": 2.9605263157894736e-07,
+      "loss": 1.2267,
+      "step": 21564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6491899490356445,
+      "learning_rate": 2.951388888888889e-07,
+      "loss": 1.5942,
+      "step": 21565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0924954414367676,
+      "learning_rate": 2.942251461988304e-07,
+      "loss": 1.1755,
+      "step": 21566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4109342098236084,
+      "learning_rate": 2.9331140350877196e-07,
+      "loss": 1.2937,
+      "step": 21567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.177309989929199,
+      "learning_rate": 2.9239766081871344e-07,
+      "loss": 1.1276,
+      "step": 21568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7102341651916504,
+      "learning_rate": 2.9148391812865497e-07,
+      "loss": 1.4218,
+      "step": 21569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5988218784332275,
+      "learning_rate": 2.905701754385965e-07,
+      "loss": 1.3947,
+      "step": 21570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1347296237945557,
+      "learning_rate": 2.8965643274853804e-07,
+      "loss": 1.2876,
+      "step": 21571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.040102243423462,
+      "learning_rate": 2.887426900584795e-07,
+      "loss": 1.2903,
+      "step": 21572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9224605560302734,
+      "learning_rate": 2.8782894736842105e-07,
+      "loss": 1.6092,
+      "step": 21573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3383405208587646,
+      "learning_rate": 2.869152046783626e-07,
+      "loss": 1.135,
+      "step": 21574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4624710083007812,
+      "learning_rate": 2.860014619883041e-07,
+      "loss": 1.4046,
+      "step": 21575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8506299257278442,
+      "learning_rate": 2.850877192982456e-07,
+      "loss": 0.9787,
+      "step": 21576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4446542263031006,
+      "learning_rate": 2.8417397660818713e-07,
+      "loss": 1.2161,
+      "step": 21577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.379640579223633,
+      "learning_rate": 2.8326023391812867e-07,
+      "loss": 1.4591,
+      "step": 21578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5009090900421143,
+      "learning_rate": 2.823464912280702e-07,
+      "loss": 1.3487,
+      "step": 21579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.41569447517395,
+      "learning_rate": 2.814327485380117e-07,
+      "loss": 1.2246,
+      "step": 21580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5914559364318848,
+      "learning_rate": 2.805190058479532e-07,
+      "loss": 1.4302,
+      "step": 21581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1667978763580322,
+      "learning_rate": 2.7960526315789475e-07,
+      "loss": 1.5498,
+      "step": 21582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0076744556427,
+      "learning_rate": 2.786915204678363e-07,
+      "loss": 1.5492,
+      "step": 21583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8286991119384766,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 1.303,
+      "step": 21584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4181227684020996,
+      "learning_rate": 2.768640350877193e-07,
+      "loss": 1.3405,
+      "step": 21585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4550886154174805,
+      "learning_rate": 2.759502923976608e-07,
+      "loss": 1.463,
+      "step": 21586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.766855001449585,
+      "learning_rate": 2.7503654970760236e-07,
+      "loss": 1.7075,
+      "step": 21587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.725289821624756,
+      "learning_rate": 2.741228070175439e-07,
+      "loss": 1.4056,
+      "step": 21588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.876328945159912,
+      "learning_rate": 2.7320906432748537e-07,
+      "loss": 1.5582,
+      "step": 21589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3966407775878906,
+      "learning_rate": 2.722953216374269e-07,
+      "loss": 1.278,
+      "step": 21590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3402576446533203,
+      "learning_rate": 2.7138157894736844e-07,
+      "loss": 1.3895,
+      "step": 21591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4926676750183105,
+      "learning_rate": 2.7046783625730997e-07,
+      "loss": 1.3602,
+      "step": 21592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5467946529388428,
+      "learning_rate": 2.6955409356725145e-07,
+      "loss": 1.6764,
+      "step": 21593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5095415115356445,
+      "learning_rate": 2.68640350877193e-07,
+      "loss": 1.6615,
+      "step": 21594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6859824657440186,
+      "learning_rate": 2.677266081871345e-07,
+      "loss": 1.5974,
+      "step": 21595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.170522928237915,
+      "learning_rate": 2.6681286549707605e-07,
+      "loss": 1.3008,
+      "step": 21596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.781050682067871,
+      "learning_rate": 2.658991228070176e-07,
+      "loss": 1.46,
+      "step": 21597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4631590843200684,
+      "learning_rate": 2.6498538011695907e-07,
+      "loss": 1.4311,
+      "step": 21598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8558757305145264,
+      "learning_rate": 2.640716374269006e-07,
+      "loss": 1.6065,
+      "step": 21599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2173473834991455,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 1.08,
+      "step": 21600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1980626583099365,
+      "learning_rate": 2.6224415204678367e-07,
+      "loss": 1.2649,
+      "step": 21601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.921384811401367,
+      "learning_rate": 2.6133040935672515e-07,
+      "loss": 1.7006,
+      "step": 21602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3383865356445312,
+      "learning_rate": 2.604166666666667e-07,
+      "loss": 1.3631,
+      "step": 21603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.225032091140747,
+      "learning_rate": 2.595029239766082e-07,
+      "loss": 1.242,
+      "step": 21604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9471213817596436,
+      "learning_rate": 2.5858918128654975e-07,
+      "loss": 1.753,
+      "step": 21605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.907275676727295,
+      "learning_rate": 2.576754385964912e-07,
+      "loss": 1.7433,
+      "step": 21606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9024178981781006,
+      "learning_rate": 2.5676169590643276e-07,
+      "loss": 2.1588,
+      "step": 21607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1981606483459473,
+      "learning_rate": 2.558479532163743e-07,
+      "loss": 1.5162,
+      "step": 21608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5027270317077637,
+      "learning_rate": 2.549342105263158e-07,
+      "loss": 1.5697,
+      "step": 21609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5382637977600098,
+      "learning_rate": 2.5402046783625736e-07,
+      "loss": 1.7822,
+      "step": 21610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4542198181152344,
+      "learning_rate": 2.5310672514619884e-07,
+      "loss": 1.4263,
+      "step": 21611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3532228469848633,
+      "learning_rate": 2.5219298245614037e-07,
+      "loss": 1.4592,
+      "step": 21612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.460881471633911,
+      "learning_rate": 2.512792397660819e-07,
+      "loss": 1.3922,
+      "step": 21613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8488450050354004,
+      "learning_rate": 2.5036549707602344e-07,
+      "loss": 1.2602,
+      "step": 21614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.249922752380371,
+      "learning_rate": 2.494517543859649e-07,
+      "loss": 1.4088,
+      "step": 21615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.315544366836548,
+      "learning_rate": 2.4853801169590645e-07,
+      "loss": 1.5294,
+      "step": 21616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3469083309173584,
+      "learning_rate": 2.47624269005848e-07,
+      "loss": 1.4508,
+      "step": 21617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.550776958465576,
+      "learning_rate": 2.467105263157895e-07,
+      "loss": 1.5237,
+      "step": 21618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3183798789978027,
+      "learning_rate": 2.4579678362573105e-07,
+      "loss": 1.8493,
+      "step": 21619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2179675102233887,
+      "learning_rate": 2.4488304093567253e-07,
+      "loss": 1.2659,
+      "step": 21620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6199629306793213,
+      "learning_rate": 2.4396929824561407e-07,
+      "loss": 1.6752,
+      "step": 21621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.473742961883545,
+      "learning_rate": 2.430555555555556e-07,
+      "loss": 1.2523,
+      "step": 21622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9415910243988037,
+      "learning_rate": 2.4214181286549713e-07,
+      "loss": 1.6728,
+      "step": 21623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2685294151306152,
+      "learning_rate": 2.412280701754386e-07,
+      "loss": 1.1734,
+      "step": 21624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7558398246765137,
+      "learning_rate": 2.4031432748538015e-07,
+      "loss": 1.4824,
+      "step": 21625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.265681266784668,
+      "learning_rate": 2.394005847953217e-07,
+      "loss": 1.3633,
+      "step": 21626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2046611309051514,
+      "learning_rate": 2.384868421052632e-07,
+      "loss": 1.3681,
+      "step": 21627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4778599739074707,
+      "learning_rate": 2.375730994152047e-07,
+      "loss": 1.3892,
+      "step": 21628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.576784133911133,
+      "learning_rate": 2.3665935672514623e-07,
+      "loss": 1.5073,
+      "step": 21629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0278000831604004,
+      "learning_rate": 2.3574561403508773e-07,
+      "loss": 1.2227,
+      "step": 21630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3669581413269043,
+      "learning_rate": 2.3483187134502927e-07,
+      "loss": 1.3737,
+      "step": 21631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1753032207489014,
+      "learning_rate": 2.3391812865497077e-07,
+      "loss": 1.4401,
+      "step": 21632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7437314987182617,
+      "learning_rate": 2.330043859649123e-07,
+      "loss": 1.2595,
+      "step": 21633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.536522388458252,
+      "learning_rate": 2.320906432748538e-07,
+      "loss": 1.3669,
+      "step": 21634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1594560146331787,
+      "learning_rate": 2.3117690058479535e-07,
+      "loss": 1.0998,
+      "step": 21635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.879133939743042,
+      "learning_rate": 2.3026315789473685e-07,
+      "loss": 1.6525,
+      "step": 21636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.90643048286438,
+      "learning_rate": 2.2934941520467839e-07,
+      "loss": 1.8693,
+      "step": 21637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5942463874816895,
+      "learning_rate": 2.2843567251461992e-07,
+      "loss": 1.6211,
+      "step": 21638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3038172721862793,
+      "learning_rate": 2.2752192982456143e-07,
+      "loss": 1.5071,
+      "step": 21639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.658632278442383,
+      "learning_rate": 2.2660818713450293e-07,
+      "loss": 1.6844,
+      "step": 21640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9700658321380615,
+      "learning_rate": 2.2569444444444447e-07,
+      "loss": 1.6231,
+      "step": 21641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5957820415496826,
+      "learning_rate": 2.24780701754386e-07,
+      "loss": 1.3476,
+      "step": 21642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.497816324234009,
+      "learning_rate": 2.238669590643275e-07,
+      "loss": 1.5933,
+      "step": 21643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1153292655944824,
+      "learning_rate": 2.22953216374269e-07,
+      "loss": 1.2811,
+      "step": 21644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.91013503074646,
+      "learning_rate": 2.2203947368421055e-07,
+      "loss": 1.644,
+      "step": 21645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1227242946624756,
+      "learning_rate": 2.2112573099415208e-07,
+      "loss": 1.7392,
+      "step": 21646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3596861362457275,
+      "learning_rate": 2.2021198830409359e-07,
+      "loss": 1.4311,
+      "step": 21647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4831061363220215,
+      "learning_rate": 2.192982456140351e-07,
+      "loss": 1.3144,
+      "step": 21648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.534285068511963,
+      "learning_rate": 2.1838450292397663e-07,
+      "loss": 1.4907,
+      "step": 21649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.669559955596924,
+      "learning_rate": 2.1747076023391816e-07,
+      "loss": 1.2849,
+      "step": 21650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.622035503387451,
+      "learning_rate": 2.165570175438597e-07,
+      "loss": 1.619,
+      "step": 21651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7106564044952393,
+      "learning_rate": 2.1564327485380117e-07,
+      "loss": 1.4998,
+      "step": 21652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.687727212905884,
+      "learning_rate": 2.147295321637427e-07,
+      "loss": 1.6055,
+      "step": 21653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.486264705657959,
+      "learning_rate": 2.1381578947368424e-07,
+      "loss": 1.416,
+      "step": 21654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.442674398422241,
+      "learning_rate": 2.1290204678362577e-07,
+      "loss": 1.5762,
+      "step": 21655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.604210615158081,
+      "learning_rate": 2.1198830409356725e-07,
+      "loss": 1.2653,
+      "step": 21656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1552352905273438,
+      "learning_rate": 2.1107456140350879e-07,
+      "loss": 0.991,
+      "step": 21657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.632561206817627,
+      "learning_rate": 2.1016081871345032e-07,
+      "loss": 1.5228,
+      "step": 21658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.202277183532715,
+      "learning_rate": 2.0924707602339185e-07,
+      "loss": 1.0525,
+      "step": 21659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7175445556640625,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 1.4705,
+      "step": 21660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.402303695678711,
+      "learning_rate": 2.0741959064327487e-07,
+      "loss": 1.3711,
+      "step": 21661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.114659547805786,
+      "learning_rate": 2.065058479532164e-07,
+      "loss": 1.3113,
+      "step": 21662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6580491065979004,
+      "learning_rate": 2.0559210526315793e-07,
+      "loss": 1.5851,
+      "step": 21663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.504556894302368,
+      "learning_rate": 2.046783625730994e-07,
+      "loss": 1.5126,
+      "step": 21664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6513609886169434,
+      "learning_rate": 2.0376461988304095e-07,
+      "loss": 2.0189,
+      "step": 21665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.317274808883667,
+      "learning_rate": 2.0285087719298248e-07,
+      "loss": 1.8518,
+      "step": 21666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3872599601745605,
+      "learning_rate": 2.01937134502924e-07,
+      "loss": 1.2785,
+      "step": 21667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.33233904838562,
+      "learning_rate": 2.010233918128655e-07,
+      "loss": 1.6103,
+      "step": 21668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.060314416885376,
+      "learning_rate": 2.0010964912280703e-07,
+      "loss": 1.1377,
+      "step": 21669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6913504600524902,
+      "learning_rate": 1.9919590643274856e-07,
+      "loss": 1.2984,
+      "step": 21670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6251509189605713,
+      "learning_rate": 1.982821637426901e-07,
+      "loss": 1.6207,
+      "step": 21671
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4747161865234375,
+      "learning_rate": 1.9736842105263157e-07,
+      "loss": 1.3966,
+      "step": 21672
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1849701404571533,
+      "learning_rate": 1.964546783625731e-07,
+      "loss": 1.0692,
+      "step": 21673
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0736584663391113,
+      "learning_rate": 1.9554093567251464e-07,
+      "loss": 1.6175,
+      "step": 21674
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8928993940353394,
+      "learning_rate": 1.9462719298245617e-07,
+      "loss": 0.9738,
+      "step": 21675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.519763469696045,
+      "learning_rate": 1.9371345029239765e-07,
+      "loss": 1.6074,
+      "step": 21676
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9740116596221924,
+      "learning_rate": 1.9279970760233919e-07,
+      "loss": 1.5454,
+      "step": 21677
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.427485227584839,
+      "learning_rate": 1.9188596491228072e-07,
+      "loss": 1.4458,
+      "step": 21678
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4841690063476562,
+      "learning_rate": 1.9097222222222225e-07,
+      "loss": 1.4023,
+      "step": 21679
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8112343549728394,
+      "learning_rate": 1.9005847953216376e-07,
+      "loss": 1.2373,
+      "step": 21680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.623400926589966,
+      "learning_rate": 1.8914473684210527e-07,
+      "loss": 1.2303,
+      "step": 21681
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1178765296936035,
+      "learning_rate": 1.882309941520468e-07,
+      "loss": 1.1537,
+      "step": 21682
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5394675731658936,
+      "learning_rate": 1.8731725146198833e-07,
+      "loss": 1.4926,
+      "step": 21683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5408711433410645,
+      "learning_rate": 1.8640350877192984e-07,
+      "loss": 1.1547,
+      "step": 21684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3168246746063232,
+      "learning_rate": 1.8548976608187135e-07,
+      "loss": 1.3846,
+      "step": 21685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5555238723754883,
+      "learning_rate": 1.8457602339181288e-07,
+      "loss": 1.4777,
+      "step": 21686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.562896251678467,
+      "learning_rate": 1.836622807017544e-07,
+      "loss": 1.4469,
+      "step": 21687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.579273223876953,
+      "learning_rate": 1.8274853801169592e-07,
+      "loss": 1.9824,
+      "step": 21688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.60686993598938,
+      "learning_rate": 1.8183479532163745e-07,
+      "loss": 1.4015,
+      "step": 21689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.735466957092285,
+      "learning_rate": 1.8092105263157896e-07,
+      "loss": 1.5018,
+      "step": 21690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7296998500823975,
+      "learning_rate": 1.800073099415205e-07,
+      "loss": 1.2855,
+      "step": 21691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4257144927978516,
+      "learning_rate": 1.79093567251462e-07,
+      "loss": 1.5046,
+      "step": 21692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6312408447265625,
+      "learning_rate": 1.7817982456140353e-07,
+      "loss": 1.8122,
+      "step": 21693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4655072689056396,
+      "learning_rate": 1.7726608187134504e-07,
+      "loss": 1.544,
+      "step": 21694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1827852725982666,
+      "learning_rate": 1.7635233918128657e-07,
+      "loss": 1.4054,
+      "step": 21695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.646902084350586,
+      "learning_rate": 1.7543859649122808e-07,
+      "loss": 1.3071,
+      "step": 21696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.724013566970825,
+      "learning_rate": 1.745248538011696e-07,
+      "loss": 1.5639,
+      "step": 21697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6095335483551025,
+      "learning_rate": 1.7361111111111115e-07,
+      "loss": 1.4935,
+      "step": 21698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8222198486328125,
+      "learning_rate": 1.7269736842105265e-07,
+      "loss": 1.5968,
+      "step": 21699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5539393424987793,
+      "learning_rate": 1.7178362573099416e-07,
+      "loss": 1.4394,
+      "step": 21700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.350374937057495,
+      "learning_rate": 1.708698830409357e-07,
+      "loss": 1.3732,
+      "step": 21701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1970224380493164,
+      "learning_rate": 1.6995614035087723e-07,
+      "loss": 1.2243,
+      "step": 21702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2568252086639404,
+      "learning_rate": 1.6904239766081873e-07,
+      "loss": 1.27,
+      "step": 21703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0917789936065674,
+      "learning_rate": 1.6812865497076024e-07,
+      "loss": 1.2331,
+      "step": 21704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9083282947540283,
+      "learning_rate": 1.6721491228070177e-07,
+      "loss": 1.6727,
+      "step": 21705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.730363607406616,
+      "learning_rate": 1.663011695906433e-07,
+      "loss": 1.3747,
+      "step": 21706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.777621269226074,
+      "learning_rate": 1.653874269005848e-07,
+      "loss": 1.3967,
+      "step": 21707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.252739667892456,
+      "learning_rate": 1.6447368421052632e-07,
+      "loss": 1.4173,
+      "step": 21708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.583880662918091,
+      "learning_rate": 1.6355994152046785e-07,
+      "loss": 1.4639,
+      "step": 21709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3604788780212402,
+      "learning_rate": 1.6264619883040939e-07,
+      "loss": 1.8367,
+      "step": 21710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.012045383453369,
+      "learning_rate": 1.6173245614035092e-07,
+      "loss": 1.4434,
+      "step": 21711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7572386264801025,
+      "learning_rate": 1.608187134502924e-07,
+      "loss": 1.3426,
+      "step": 21712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.349689245223999,
+      "learning_rate": 1.5990497076023393e-07,
+      "loss": 1.5299,
+      "step": 21713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6923577785491943,
+      "learning_rate": 1.5899122807017547e-07,
+      "loss": 1.5354,
+      "step": 21714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5874929428100586,
+      "learning_rate": 1.58077485380117e-07,
+      "loss": 1.5785,
+      "step": 21715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.029313325881958,
+      "learning_rate": 1.5716374269005848e-07,
+      "loss": 1.6709,
+      "step": 21716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1049649715423584,
+      "learning_rate": 1.5625e-07,
+      "loss": 1.4348,
+      "step": 21717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1837241649627686,
+      "learning_rate": 1.5533625730994152e-07,
+      "loss": 1.2256,
+      "step": 21718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7027976512908936,
+      "learning_rate": 1.5442251461988305e-07,
+      "loss": 1.6184,
+      "step": 21719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5921528339385986,
+      "learning_rate": 1.5350877192982456e-07,
+      "loss": 1.1888,
+      "step": 21720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3243281841278076,
+      "learning_rate": 1.525950292397661e-07,
+      "loss": 1.3652,
+      "step": 21721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5255181789398193,
+      "learning_rate": 1.516812865497076e-07,
+      "loss": 1.4568,
+      "step": 21722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5468204021453857,
+      "learning_rate": 1.5076754385964913e-07,
+      "loss": 1.5268,
+      "step": 21723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.467778444290161,
+      "learning_rate": 1.4985380116959064e-07,
+      "loss": 1.4025,
+      "step": 21724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0466532707214355,
+      "learning_rate": 1.4894005847953217e-07,
+      "loss": 1.4032,
+      "step": 21725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6050479412078857,
+      "learning_rate": 1.4802631578947368e-07,
+      "loss": 1.6582,
+      "step": 21726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7807772159576416,
+      "learning_rate": 1.471125730994152e-07,
+      "loss": 1.3272,
+      "step": 21727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8733887672424316,
+      "learning_rate": 1.4619883040935672e-07,
+      "loss": 1.5547,
+      "step": 21728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3583970069885254,
+      "learning_rate": 1.4528508771929825e-07,
+      "loss": 1.2866,
+      "step": 21729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4887235164642334,
+      "learning_rate": 1.4437134502923976e-07,
+      "loss": 1.7652,
+      "step": 21730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7067034244537354,
+      "learning_rate": 1.434576023391813e-07,
+      "loss": 1.6318,
+      "step": 21731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4593913555145264,
+      "learning_rate": 1.425438596491228e-07,
+      "loss": 2.0611,
+      "step": 21732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4230175018310547,
+      "learning_rate": 1.4163011695906433e-07,
+      "loss": 1.6906,
+      "step": 21733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6509642601013184,
+      "learning_rate": 1.4071637426900584e-07,
+      "loss": 1.273,
+      "step": 21734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.409830331802368,
+      "learning_rate": 1.3980263157894737e-07,
+      "loss": 1.2389,
+      "step": 21735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5293354988098145,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 1.3355,
+      "step": 21736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.163126230239868,
+      "learning_rate": 1.379751461988304e-07,
+      "loss": 1.8913,
+      "step": 21737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.868504285812378,
+      "learning_rate": 1.3706140350877195e-07,
+      "loss": 1.5232,
+      "step": 21738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.065842628479004,
+      "learning_rate": 1.3614766081871345e-07,
+      "loss": 1.3962,
+      "step": 21739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1902029514312744,
+      "learning_rate": 1.3523391812865499e-07,
+      "loss": 1.1444,
+      "step": 21740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3186893463134766,
+      "learning_rate": 1.343201754385965e-07,
+      "loss": 1.4159,
+      "step": 21741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9910417795181274,
+      "learning_rate": 1.3340643274853803e-07,
+      "loss": 1.0132,
+      "step": 21742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4129223823547363,
+      "learning_rate": 1.3249269005847953e-07,
+      "loss": 1.3146,
+      "step": 21743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4383203983306885,
+      "learning_rate": 1.3157894736842107e-07,
+      "loss": 1.5919,
+      "step": 21744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.437377691268921,
+      "learning_rate": 1.3066520467836257e-07,
+      "loss": 1.2493,
+      "step": 21745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.760256290435791,
+      "learning_rate": 1.297514619883041e-07,
+      "loss": 1.4871,
+      "step": 21746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5865347385406494,
+      "learning_rate": 1.288377192982456e-07,
+      "loss": 1.5105,
+      "step": 21747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.115097761154175,
+      "learning_rate": 1.2792397660818715e-07,
+      "loss": 1.4194,
+      "step": 21748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1342573165893555,
+      "learning_rate": 1.2701023391812868e-07,
+      "loss": 1.8928,
+      "step": 21749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.098813056945801,
+      "learning_rate": 1.2609649122807019e-07,
+      "loss": 1.8284,
+      "step": 21750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.52069091796875,
+      "learning_rate": 1.2518274853801172e-07,
+      "loss": 1.6735,
+      "step": 21751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.439969062805176,
+      "learning_rate": 1.2426900584795323e-07,
+      "loss": 1.407,
+      "step": 21752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.681612730026245,
+      "learning_rate": 1.2335526315789476e-07,
+      "loss": 1.6069,
+      "step": 21753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3779025077819824,
+      "learning_rate": 1.2244152046783627e-07,
+      "loss": 1.5002,
+      "step": 21754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4974465370178223,
+      "learning_rate": 1.215277777777778e-07,
+      "loss": 1.6381,
+      "step": 21755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2377097606658936,
+      "learning_rate": 1.206140350877193e-07,
+      "loss": 1.3026,
+      "step": 21756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.553929090499878,
+      "learning_rate": 1.1970029239766084e-07,
+      "loss": 1.4958,
+      "step": 21757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2016704082489014,
+      "learning_rate": 1.1878654970760235e-07,
+      "loss": 1.1546,
+      "step": 21758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.519390106201172,
+      "learning_rate": 1.1787280701754387e-07,
+      "loss": 1.5318,
+      "step": 21759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4388694763183594,
+      "learning_rate": 1.1695906432748539e-07,
+      "loss": 1.7403,
+      "step": 21760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0725183486938477,
+      "learning_rate": 1.160453216374269e-07,
+      "loss": 1.1931,
+      "step": 21761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7569122314453125,
+      "learning_rate": 1.1513157894736843e-07,
+      "loss": 1.0872,
+      "step": 21762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6148059368133545,
+      "learning_rate": 1.1421783625730996e-07,
+      "loss": 1.2829,
+      "step": 21763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.393139123916626,
+      "learning_rate": 1.1330409356725147e-07,
+      "loss": 1.2777,
+      "step": 21764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.683936357498169,
+      "learning_rate": 1.12390350877193e-07,
+      "loss": 1.686,
+      "step": 21765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0935699939727783,
+      "learning_rate": 1.114766081871345e-07,
+      "loss": 1.2999,
+      "step": 21766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9223549365997314,
+      "learning_rate": 1.1056286549707604e-07,
+      "loss": 1.4194,
+      "step": 21767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4835140705108643,
+      "learning_rate": 1.0964912280701755e-07,
+      "loss": 1.4463,
+      "step": 21768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4865102767944336,
+      "learning_rate": 1.0873538011695908e-07,
+      "loss": 1.497,
+      "step": 21769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.221510171890259,
+      "learning_rate": 1.0782163742690059e-07,
+      "loss": 1.3863,
+      "step": 21770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9305685758590698,
+      "learning_rate": 1.0690789473684212e-07,
+      "loss": 1.3225,
+      "step": 21771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1341686248779297,
+      "learning_rate": 1.0599415204678363e-07,
+      "loss": 1.3172,
+      "step": 21772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.784371852874756,
+      "learning_rate": 1.0508040935672516e-07,
+      "loss": 1.8317,
+      "step": 21773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6400325298309326,
+      "learning_rate": 1.0416666666666667e-07,
+      "loss": 1.3345,
+      "step": 21774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2900588512420654,
+      "learning_rate": 1.032529239766082e-07,
+      "loss": 1.3515,
+      "step": 21775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.674191474914551,
+      "learning_rate": 1.023391812865497e-07,
+      "loss": 1.8373,
+      "step": 21776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.393681287765503,
+      "learning_rate": 1.0142543859649124e-07,
+      "loss": 1.2506,
+      "step": 21777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4576590061187744,
+      "learning_rate": 1.0051169590643275e-07,
+      "loss": 1.2943,
+      "step": 21778
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.37038516998291,
+      "learning_rate": 9.959795321637428e-08,
+      "loss": 1.4524,
+      "step": 21779
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7629446983337402,
+      "learning_rate": 9.868421052631579e-08,
+      "loss": 1.5592,
+      "step": 21780
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.419520378112793,
+      "learning_rate": 9.777046783625732e-08,
+      "loss": 1.4374,
+      "step": 21781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8345837593078613,
+      "learning_rate": 9.685672514619883e-08,
+      "loss": 1.6621,
+      "step": 21782
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9128830432891846,
+      "learning_rate": 9.594298245614036e-08,
+      "loss": 1.8323,
+      "step": 21783
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7270631790161133,
+      "learning_rate": 9.502923976608188e-08,
+      "loss": 1.8315,
+      "step": 21784
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0040359497070312,
+      "learning_rate": 9.41154970760234e-08,
+      "loss": 1.0084,
+      "step": 21785
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6767547130584717,
+      "learning_rate": 9.320175438596492e-08,
+      "loss": 1.8525,
+      "step": 21786
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3877511024475098,
+      "learning_rate": 9.228801169590644e-08,
+      "loss": 1.5119,
+      "step": 21787
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6279661655426025,
+      "learning_rate": 9.137426900584796e-08,
+      "loss": 1.6307,
+      "step": 21788
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.685792922973633,
+      "learning_rate": 9.046052631578948e-08,
+      "loss": 1.6067,
+      "step": 21789
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5921545028686523,
+      "learning_rate": 8.9546783625731e-08,
+      "loss": 1.6389,
+      "step": 21790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.519810914993286,
+      "learning_rate": 8.863304093567252e-08,
+      "loss": 1.6043,
+      "step": 21791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9688389301300049,
+      "learning_rate": 8.771929824561404e-08,
+      "loss": 1.0163,
+      "step": 21792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5417065620422363,
+      "learning_rate": 8.680555555555557e-08,
+      "loss": 1.5774,
+      "step": 21793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4508895874023438,
+      "learning_rate": 8.589181286549708e-08,
+      "loss": 1.4256,
+      "step": 21794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6855525970458984,
+      "learning_rate": 8.497807017543861e-08,
+      "loss": 1.3369,
+      "step": 21795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9905962944030762,
+      "learning_rate": 8.406432748538012e-08,
+      "loss": 1.1125,
+      "step": 21796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5282139778137207,
+      "learning_rate": 8.315058479532165e-08,
+      "loss": 1.5741,
+      "step": 21797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7030816078186035,
+      "learning_rate": 8.223684210526316e-08,
+      "loss": 1.5381,
+      "step": 21798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5657453536987305,
+      "learning_rate": 8.132309941520469e-08,
+      "loss": 1.1921,
+      "step": 21799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.597447633743286,
+      "learning_rate": 8.04093567251462e-08,
+      "loss": 1.7205,
+      "step": 21800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9662163257598877,
+      "learning_rate": 7.949561403508773e-08,
+      "loss": 1.8065,
+      "step": 21801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2945330142974854,
+      "learning_rate": 7.858187134502924e-08,
+      "loss": 1.2869,
+      "step": 21802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.298844814300537,
+      "learning_rate": 7.766812865497076e-08,
+      "loss": 0.9473,
+      "step": 21803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2202186584472656,
+      "learning_rate": 7.675438596491228e-08,
+      "loss": 1.287,
+      "step": 21804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.426694631576538,
+      "learning_rate": 7.58406432748538e-08,
+      "loss": 1.4952,
+      "step": 21805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.589184045791626,
+      "learning_rate": 7.492690058479532e-08,
+      "loss": 1.3831,
+      "step": 21806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.739058256149292,
+      "learning_rate": 7.401315789473684e-08,
+      "loss": 1.5957,
+      "step": 21807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2490158081054688,
+      "learning_rate": 7.309941520467836e-08,
+      "loss": 1.3545,
+      "step": 21808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.674926996231079,
+      "learning_rate": 7.218567251461988e-08,
+      "loss": 1.8179,
+      "step": 21809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3392717838287354,
+      "learning_rate": 7.12719298245614e-08,
+      "loss": 1.5126,
+      "step": 21810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.397620916366577,
+      "learning_rate": 7.035818713450292e-08,
+      "loss": 1.2272,
+      "step": 21811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2898294925689697,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 1.3976,
+      "step": 21812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.678316593170166,
+      "learning_rate": 6.853070175438597e-08,
+      "loss": 1.3886,
+      "step": 21813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.642162561416626,
+      "learning_rate": 6.761695906432749e-08,
+      "loss": 1.6045,
+      "step": 21814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0421602725982666,
+      "learning_rate": 6.670321637426901e-08,
+      "loss": 1.5831,
+      "step": 21815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.316189765930176,
+      "learning_rate": 6.578947368421053e-08,
+      "loss": 1.3954,
+      "step": 21816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9402239322662354,
+      "learning_rate": 6.487573099415205e-08,
+      "loss": 1.4391,
+      "step": 21817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.06777286529541,
+      "learning_rate": 6.396198830409357e-08,
+      "loss": 1.0949,
+      "step": 21818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.530350923538208,
+      "learning_rate": 6.304824561403509e-08,
+      "loss": 1.2822,
+      "step": 21819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.577709913253784,
+      "learning_rate": 6.213450292397661e-08,
+      "loss": 1.4954,
+      "step": 21820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.664989471435547,
+      "learning_rate": 6.122076023391813e-08,
+      "loss": 1.4058,
+      "step": 21821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.912096619606018,
+      "learning_rate": 6.030701754385965e-08,
+      "loss": 1.1482,
+      "step": 21822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.613640785217285,
+      "learning_rate": 5.939327485380117e-08,
+      "loss": 1.8453,
+      "step": 21823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6404218673706055,
+      "learning_rate": 5.847953216374269e-08,
+      "loss": 1.647,
+      "step": 21824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5289595127105713,
+      "learning_rate": 5.756578947368421e-08,
+      "loss": 1.3738,
+      "step": 21825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.104945659637451,
+      "learning_rate": 5.665204678362573e-08,
+      "loss": 1.2913,
+      "step": 21826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6786303520202637,
+      "learning_rate": 5.573830409356725e-08,
+      "loss": 1.3959,
+      "step": 21827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4499142169952393,
+      "learning_rate": 5.482456140350877e-08,
+      "loss": 1.6116,
+      "step": 21828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.048004150390625,
+      "learning_rate": 5.391081871345029e-08,
+      "loss": 1.7095,
+      "step": 21829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.875838041305542,
+      "learning_rate": 5.299707602339181e-08,
+      "loss": 1.8729,
+      "step": 21830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.842928647994995,
+      "learning_rate": 5.208333333333333e-08,
+      "loss": 1.7574,
+      "step": 21831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.83215594291687,
+      "learning_rate": 5.116959064327485e-08,
+      "loss": 1.3136,
+      "step": 21832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1407699584960938,
+      "learning_rate": 5.025584795321637e-08,
+      "loss": 1.2048,
+      "step": 21833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2906007766723633,
+      "learning_rate": 4.934210526315789e-08,
+      "loss": 1.3322,
+      "step": 21834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9683785438537598,
+      "learning_rate": 4.842836257309941e-08,
+      "loss": 1.639,
+      "step": 21835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.653717279434204,
+      "learning_rate": 4.751461988304094e-08,
+      "loss": 1.8341,
+      "step": 21836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5230655670166016,
+      "learning_rate": 4.660087719298246e-08,
+      "loss": 1.5521,
+      "step": 21837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7422900199890137,
+      "learning_rate": 4.568713450292398e-08,
+      "loss": 1.4928,
+      "step": 21838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.887742519378662,
+      "learning_rate": 4.47733918128655e-08,
+      "loss": 1.7504,
+      "step": 21839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5193068981170654,
+      "learning_rate": 4.385964912280702e-08,
+      "loss": 1.1452,
+      "step": 21840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.670344829559326,
+      "learning_rate": 4.294590643274854e-08,
+      "loss": 1.2988,
+      "step": 21841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1920905113220215,
+      "learning_rate": 4.203216374269006e-08,
+      "loss": 1.3271,
+      "step": 21842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.805253028869629,
+      "learning_rate": 4.111842105263158e-08,
+      "loss": 1.7585,
+      "step": 21843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2325620651245117,
+      "learning_rate": 4.02046783625731e-08,
+      "loss": 1.2859,
+      "step": 21844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1596434116363525,
+      "learning_rate": 3.929093567251462e-08,
+      "loss": 1.161,
+      "step": 21845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.275374412536621,
+      "learning_rate": 3.837719298245614e-08,
+      "loss": 1.3141,
+      "step": 21846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.311910390853882,
+      "learning_rate": 3.746345029239766e-08,
+      "loss": 1.5112,
+      "step": 21847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7115187644958496,
+      "learning_rate": 3.654970760233918e-08,
+      "loss": 1.3751,
+      "step": 21848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4977049827575684,
+      "learning_rate": 3.56359649122807e-08,
+      "loss": 1.486,
+      "step": 21849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4782142639160156,
+      "learning_rate": 3.472222222222222e-08,
+      "loss": 1.7397,
+      "step": 21850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8656158447265625,
+      "learning_rate": 3.3808479532163747e-08,
+      "loss": 1.2658,
+      "step": 21851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.511246919631958,
+      "learning_rate": 3.2894736842105267e-08,
+      "loss": 1.433,
+      "step": 21852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6296727657318115,
+      "learning_rate": 3.1980994152046787e-08,
+      "loss": 1.2021,
+      "step": 21853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5933313369750977,
+      "learning_rate": 3.1067251461988307e-08,
+      "loss": 1.61,
+      "step": 21854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2787556648254395,
+      "learning_rate": 3.0153508771929827e-08,
+      "loss": 1.2393,
+      "step": 21855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4682693481445312,
+      "learning_rate": 2.9239766081871347e-08,
+      "loss": 1.4787,
+      "step": 21856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4773268699645996,
+      "learning_rate": 2.8326023391812867e-08,
+      "loss": 1.4329,
+      "step": 21857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4947092533111572,
+      "learning_rate": 2.7412280701754387e-08,
+      "loss": 1.3137,
+      "step": 21858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9353127479553223,
+      "learning_rate": 2.6498538011695907e-08,
+      "loss": 1.642,
+      "step": 21859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.085235595703125,
+      "learning_rate": 2.5584795321637427e-08,
+      "loss": 1.2553,
+      "step": 21860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6792521476745605,
+      "learning_rate": 2.4671052631578947e-08,
+      "loss": 1.7061,
+      "step": 21861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1173787117004395,
+      "learning_rate": 2.375730994152047e-08,
+      "loss": 1.8181,
+      "step": 21862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.085582971572876,
+      "learning_rate": 2.284356725146199e-08,
+      "loss": 1.2284,
+      "step": 21863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8990585803985596,
+      "learning_rate": 2.192982456140351e-08,
+      "loss": 1.7281,
+      "step": 21864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4107978343963623,
+      "learning_rate": 2.101608187134503e-08,
+      "loss": 1.4729,
+      "step": 21865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.353691339492798,
+      "learning_rate": 2.010233918128655e-08,
+      "loss": 1.3904,
+      "step": 21866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2081172466278076,
+      "learning_rate": 1.918859649122807e-08,
+      "loss": 1.2042,
+      "step": 21867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6521031856536865,
+      "learning_rate": 1.827485380116959e-08,
+      "loss": 1.3547,
+      "step": 21868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6400458812713623,
+      "learning_rate": 1.736111111111111e-08,
+      "loss": 1.5789,
+      "step": 21869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.513774871826172,
+      "learning_rate": 1.6447368421052633e-08,
+      "loss": 1.3413,
+      "step": 21870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.641343832015991,
+      "learning_rate": 1.5533625730994153e-08,
+      "loss": 1.4529,
+      "step": 21871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7285726070404053,
+      "learning_rate": 1.4619883040935673e-08,
+      "loss": 1.5572,
+      "step": 21872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5474977493286133,
+      "learning_rate": 1.3706140350877193e-08,
+      "loss": 1.2058,
+      "step": 21873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0482118129730225,
+      "learning_rate": 1.2792397660818713e-08,
+      "loss": 1.6546,
+      "step": 21874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.216183662414551,
+      "learning_rate": 1.1878654970760235e-08,
+      "loss": 1.1094,
+      "step": 21875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5800554752349854,
+      "learning_rate": 1.0964912280701755e-08,
+      "loss": 1.4376,
+      "step": 21876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.137537717819214,
+      "learning_rate": 1.0051169590643275e-08,
+      "loss": 1.2465,
+      "step": 21877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.861475944519043,
+      "learning_rate": 9.137426900584795e-09,
+      "loss": 1.5568,
+      "step": 21878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7229716777801514,
+      "learning_rate": 8.223684210526317e-09,
+      "loss": 1.4156,
+      "step": 21879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2448902130126953,
+      "learning_rate": 7.309941520467837e-09,
+      "loss": 1.198,
+      "step": 21880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.555177927017212,
+      "learning_rate": 6.396198830409357e-09,
+      "loss": 1.3884,
+      "step": 21881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.449700117111206,
+      "learning_rate": 5.4824561403508775e-09,
+      "loss": 1.4606,
+      "step": 21882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1086323261260986,
+      "learning_rate": 4.5687134502923975e-09,
+      "loss": 1.3659,
+      "step": 21883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.711329698562622,
+      "learning_rate": 3.6549707602339183e-09,
+      "loss": 1.4554,
+      "step": 21884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.437237024307251,
+      "learning_rate": 2.7412280701754387e-09,
+      "loss": 1.3348,
+      "step": 21885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1009199619293213,
+      "learning_rate": 1.8274853801169592e-09,
+      "loss": 1.2925,
+      "step": 21886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7382454872131348,
+      "learning_rate": 9.137426900584796e-10,
+      "loss": 1.4197,
+      "step": 21887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 22.333215713500977,
+      "learning_rate": 0.0,
+      "loss": 3.6765,
+      "step": 21888
+    },
+    {
+      "epoch": 1.0,
+      "step": 21888,
+      "total_flos": 6.124065594653491e+16,
+      "train_loss": 1.662468353180182,
+      "train_runtime": 7285.6469,
+      "train_samples_per_second": 96.132,
+      "train_steps_per_second": 3.004
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 21888,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 6.124065594653491e+16,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}