diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4662 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9969742813918305,
+  "eval_steps": 500,
+  "global_step": 660,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0030257186081694403,
+      "grad_norm": 3.279416072530215,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.6903,
+      "step": 1
+    },
+    {
+      "epoch": 0.006051437216338881,
+      "grad_norm": 3.4354880255370754,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.7324,
+      "step": 2
+    },
+    {
+      "epoch": 0.009077155824508321,
+      "grad_norm": 3.0335821702138324,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.681,
+      "step": 3
+    },
+    {
+      "epoch": 0.012102874432677761,
+      "grad_norm": 2.9058300089333513,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.682,
+      "step": 4
+    },
+    {
+      "epoch": 0.015128593040847202,
+      "grad_norm": 3.3226317470392033,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.7049,
+      "step": 5
+    },
+    {
+      "epoch": 0.018154311649016642,
+      "grad_norm": 2.8917491675682645,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.6956,
+      "step": 6
+    },
+    {
+      "epoch": 0.02118003025718608,
+      "grad_norm": 2.717064127454105,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.6636,
+      "step": 7
+    },
+    {
+      "epoch": 0.024205748865355523,
+      "grad_norm": 2.2974289388558082,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.6455,
+      "step": 8
+    },
+    {
+      "epoch": 0.02723146747352496,
+      "grad_norm": 1.7322016509488496,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.6389,
+      "step": 9
+    },
+    {
+      "epoch": 0.030257186081694403,
+      "grad_norm": 1.6482050561848078,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.6276,
+      "step": 10
+    },
+    {
+      "epoch": 0.03328290468986384,
+      "grad_norm": 1.6074281475915506,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.603,
+      "step": 11
+    },
+    {
+      "epoch": 0.036308623298033284,
+      "grad_norm": 1.804909371782916,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.6052,
+      "step": 12
+    },
+    {
+      "epoch": 0.039334341906202726,
+      "grad_norm": 2.68596984151035,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.6362,
+      "step": 13
+    },
+    {
+      "epoch": 0.04236006051437216,
+      "grad_norm": 2.3991535235837422,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.5733,
+      "step": 14
+    },
+    {
+      "epoch": 0.0453857791225416,
+      "grad_norm": 1.7325057015881502,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.5713,
+      "step": 15
+    },
+    {
+      "epoch": 0.048411497730711045,
+      "grad_norm": 1.4238244199081926,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.5725,
+      "step": 16
+    },
+    {
+      "epoch": 0.05143721633888049,
+      "grad_norm": 1.2983597251010743,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.5739,
+      "step": 17
+    },
+    {
+      "epoch": 0.05446293494704992,
+      "grad_norm": 1.4323223775446203,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.581,
+      "step": 18
+    },
+    {
+      "epoch": 0.057488653555219364,
+      "grad_norm": 1.3852154675139237,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.5548,
+      "step": 19
+    },
+    {
+      "epoch": 0.060514372163388806,
+      "grad_norm": 1.2173739194527462,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.5714,
+      "step": 20
+    },
+    {
+      "epoch": 0.06354009077155824,
+      "grad_norm": 1.0435400209206185,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.5402,
+      "step": 21
+    },
+    {
+      "epoch": 0.06656580937972768,
+      "grad_norm": 1.0810646516199343,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5495,
+      "step": 22
+    },
+    {
+      "epoch": 0.06959152798789713,
+      "grad_norm": 1.0534251418051463,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.518,
+      "step": 23
+    },
+    {
+      "epoch": 0.07261724659606657,
+      "grad_norm": 0.979792119867968,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.533,
+      "step": 24
+    },
+    {
+      "epoch": 0.07564296520423601,
+      "grad_norm": 0.9879599559513101,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.5086,
+      "step": 25
+    },
+    {
+      "epoch": 0.07866868381240545,
+      "grad_norm": 0.9368104748761843,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.5136,
+      "step": 26
+    },
+    {
+      "epoch": 0.08169440242057488,
+      "grad_norm": 0.8705532335325452,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.4903,
+      "step": 27
+    },
+    {
+      "epoch": 0.08472012102874432,
+      "grad_norm": 0.914971220283082,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.503,
+      "step": 28
+    },
+    {
+      "epoch": 0.08774583963691376,
+      "grad_norm": 0.9140222130355907,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.4865,
+      "step": 29
+    },
+    {
+      "epoch": 0.0907715582450832,
+      "grad_norm": 0.886687009935771,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5332,
+      "step": 30
+    },
+    {
+      "epoch": 0.09379727685325265,
+      "grad_norm": 0.9277062549564096,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.498,
+      "step": 31
+    },
+    {
+      "epoch": 0.09682299546142209,
+      "grad_norm": 0.9518977120369461,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.5006,
+      "step": 32
+    },
+    {
+      "epoch": 0.09984871406959153,
+      "grad_norm": 0.9491680571287733,
+      "learning_rate": 1e-05,
+      "loss": 0.4873,
+      "step": 33
+    },
+    {
+      "epoch": 0.10287443267776097,
+      "grad_norm": 1.0396002150936259,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 0.5052,
+      "step": 34
+    },
+    {
+      "epoch": 0.1059001512859304,
+      "grad_norm": 0.8244921614304973,
+      "learning_rate": 1.0606060606060606e-05,
+      "loss": 0.4964,
+      "step": 35
+    },
+    {
+      "epoch": 0.10892586989409984,
+      "grad_norm": 1.0037564938122436,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5291,
+      "step": 36
+    },
+    {
+      "epoch": 0.11195158850226929,
+      "grad_norm": 0.9363599716291893,
+      "learning_rate": 1.1212121212121212e-05,
+      "loss": 0.5047,
+      "step": 37
+    },
+    {
+      "epoch": 0.11497730711043873,
+      "grad_norm": 0.7880621795944337,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 0.4826,
+      "step": 38
+    },
+    {
+      "epoch": 0.11800302571860817,
+      "grad_norm": 0.9023305747408022,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.5241,
+      "step": 39
+    },
+    {
+      "epoch": 0.12102874432677761,
+      "grad_norm": 0.9322651831225484,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.4975,
+      "step": 40
+    },
+    {
+      "epoch": 0.12405446293494705,
+      "grad_norm": 0.8908401673319695,
+      "learning_rate": 1.2424242424242425e-05,
+      "loss": 0.4675,
+      "step": 41
+    },
+    {
+      "epoch": 0.12708018154311648,
+      "grad_norm": 0.9101421687689539,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.4804,
+      "step": 42
+    },
+    {
+      "epoch": 0.13010590015128592,
+      "grad_norm": 1.030504015259605,
+      "learning_rate": 1.3030303030303032e-05,
+      "loss": 0.5232,
+      "step": 43
+    },
+    {
+      "epoch": 0.13313161875945537,
+      "grad_norm": 0.8915436267838286,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.495,
+      "step": 44
+    },
+    {
+      "epoch": 0.1361573373676248,
+      "grad_norm": 0.9339712988837848,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.4966,
+      "step": 45
+    },
+    {
+      "epoch": 0.13918305597579425,
+      "grad_norm": 0.9341652929615809,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 0.4751,
+      "step": 46
+    },
+    {
+      "epoch": 0.1422087745839637,
+      "grad_norm": 0.877714151382283,
+      "learning_rate": 1.4242424242424245e-05,
+      "loss": 0.4896,
+      "step": 47
+    },
+    {
+      "epoch": 0.14523449319213314,
+      "grad_norm": 0.8965629924673825,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.4665,
+      "step": 48
+    },
+    {
+      "epoch": 0.14826021180030258,
+      "grad_norm": 0.8606928118905104,
+      "learning_rate": 1.484848484848485e-05,
+      "loss": 0.4987,
+      "step": 49
+    },
+    {
+      "epoch": 0.15128593040847202,
+      "grad_norm": 0.850026465956348,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.4763,
+      "step": 50
+    },
+    {
+      "epoch": 0.15431164901664146,
+      "grad_norm": 0.954016049593091,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.5029,
+      "step": 51
+    },
+    {
+      "epoch": 0.1573373676248109,
+      "grad_norm": 0.8237369784584443,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.4651,
+      "step": 52
+    },
+    {
+      "epoch": 0.16036308623298035,
+      "grad_norm": 0.8551153235539422,
+      "learning_rate": 1.606060606060606e-05,
+      "loss": 0.466,
+      "step": 53
+    },
+    {
+      "epoch": 0.16338880484114976,
+      "grad_norm": 0.8608128214412675,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.4755,
+      "step": 54
+    },
+    {
+      "epoch": 0.1664145234493192,
+      "grad_norm": 0.8599355391220561,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4692,
+      "step": 55
+    },
+    {
+      "epoch": 0.16944024205748864,
+      "grad_norm": 0.8729999516744185,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 0.4919,
+      "step": 56
+    },
+    {
+      "epoch": 0.17246596066565809,
+      "grad_norm": 0.9233443564305477,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.4851,
+      "step": 57
+    },
+    {
+      "epoch": 0.17549167927382753,
+      "grad_norm": 0.9161043710976144,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 0.5085,
+      "step": 58
+    },
+    {
+      "epoch": 0.17851739788199697,
+      "grad_norm": 0.8912945938265192,
+      "learning_rate": 1.787878787878788e-05,
+      "loss": 0.4983,
+      "step": 59
+    },
+    {
+      "epoch": 0.1815431164901664,
+      "grad_norm": 1.0039960758340145,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.4905,
+      "step": 60
+    },
+    {
+      "epoch": 0.18456883509833585,
+      "grad_norm": 0.884983198504567,
+      "learning_rate": 1.8484848484848487e-05,
+      "loss": 0.4857,
+      "step": 61
+    },
+    {
+      "epoch": 0.1875945537065053,
+      "grad_norm": 0.9163960370171632,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 0.4925,
+      "step": 62
+    },
+    {
+      "epoch": 0.19062027231467474,
+      "grad_norm": 1.053239722547164,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.476,
+      "step": 63
+    },
+    {
+      "epoch": 0.19364599092284418,
+      "grad_norm": 1.0075770547198781,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 0.4885,
+      "step": 64
+    },
+    {
+      "epoch": 0.19667170953101362,
+      "grad_norm": 0.8629638643842753,
+      "learning_rate": 1.96969696969697e-05,
+      "loss": 0.463,
+      "step": 65
+    },
+    {
+      "epoch": 0.19969742813918306,
+      "grad_norm": 0.8393337984185706,
+      "learning_rate": 2e-05,
+      "loss": 0.4472,
+      "step": 66
+    },
+    {
+      "epoch": 0.2027231467473525,
+      "grad_norm": 1.0030070719559565,
+      "learning_rate": 1.9999860139251737e-05,
+      "loss": 0.467,
+      "step": 67
+    },
+    {
+      "epoch": 0.20574886535552195,
+      "grad_norm": 0.8826407418425095,
+      "learning_rate": 1.9999440560919153e-05,
+      "loss": 0.4896,
+      "step": 68
+    },
+    {
+      "epoch": 0.2087745839636914,
+      "grad_norm": 0.9840225965572931,
+      "learning_rate": 1.9998741276738753e-05,
+      "loss": 0.4889,
+      "step": 69
+    },
+    {
+      "epoch": 0.2118003025718608,
+      "grad_norm": 0.9311823951856565,
+      "learning_rate": 1.999776230627102e-05,
+      "loss": 0.4627,
+      "step": 70
+    },
+    {
+      "epoch": 0.21482602118003025,
+      "grad_norm": 0.9628534677467433,
+      "learning_rate": 1.9996503676899863e-05,
+      "loss": 0.4707,
+      "step": 71
+    },
+    {
+      "epoch": 0.2178517397881997,
+      "grad_norm": 0.8931477138782017,
+      "learning_rate": 1.9994965423831853e-05,
+      "loss": 0.4497,
+      "step": 72
+    },
+    {
+      "epoch": 0.22087745839636913,
+      "grad_norm": 0.875257463004832,
+      "learning_rate": 1.9993147590095232e-05,
+      "loss": 0.4773,
+      "step": 73
+    },
+    {
+      "epoch": 0.22390317700453857,
+      "grad_norm": 0.8483917938637087,
+      "learning_rate": 1.999105022653872e-05,
+      "loss": 0.4632,
+      "step": 74
+    },
+    {
+      "epoch": 0.22692889561270801,
+      "grad_norm": 0.8193328296037632,
+      "learning_rate": 1.9988673391830082e-05,
+      "loss": 0.4757,
+      "step": 75
+    },
+    {
+      "epoch": 0.22995461422087746,
+      "grad_norm": 0.8024807890558652,
+      "learning_rate": 1.9986017152454497e-05,
+      "loss": 0.4798,
+      "step": 76
+    },
+    {
+      "epoch": 0.2329803328290469,
+      "grad_norm": 0.8305306343976248,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.4717,
+      "step": 77
+    },
+    {
+      "epoch": 0.23600605143721634,
+      "grad_norm": 0.8519003250452306,
+      "learning_rate": 1.9979866764718846e-05,
+      "loss": 0.4631,
+      "step": 78
+    },
+    {
+      "epoch": 0.23903177004538578,
+      "grad_norm": 0.8052741232146056,
+      "learning_rate": 1.997637278839835e-05,
+      "loss": 0.4633,
+      "step": 79
+    },
+    {
+      "epoch": 0.24205748865355523,
+      "grad_norm": 0.8171996216026898,
+      "learning_rate": 1.9972599751485225e-05,
+      "loss": 0.4598,
+      "step": 80
+    },
+    {
+      "epoch": 0.24508320726172467,
+      "grad_norm": 0.8121112977735965,
+      "learning_rate": 1.9968547759519426e-05,
+      "loss": 0.4845,
+      "step": 81
+    },
+    {
+      "epoch": 0.2481089258698941,
+      "grad_norm": 0.7616664515070739,
+      "learning_rate": 1.9964216925843876e-05,
+      "loss": 0.4818,
+      "step": 82
+    },
+    {
+      "epoch": 0.25113464447806355,
+      "grad_norm": 0.7817385414589161,
+      "learning_rate": 1.9959607371601303e-05,
+      "loss": 0.4571,
+      "step": 83
+    },
+    {
+      "epoch": 0.25416036308623297,
+      "grad_norm": 0.8312392599516752,
+      "learning_rate": 1.9954719225730847e-05,
+      "loss": 0.4616,
+      "step": 84
+    },
+    {
+      "epoch": 0.25718608169440244,
+      "grad_norm": 0.7985599777949173,
+      "learning_rate": 1.994955262496446e-05,
+      "loss": 0.4782,
+      "step": 85
+    },
+    {
+      "epoch": 0.26021180030257185,
+      "grad_norm": 0.819000625176893,
+      "learning_rate": 1.9944107713823068e-05,
+      "loss": 0.472,
+      "step": 86
+    },
+    {
+      "epoch": 0.2632375189107413,
+      "grad_norm": 0.838706937923567,
+      "learning_rate": 1.9938384644612542e-05,
+      "loss": 0.4638,
+      "step": 87
+    },
+    {
+      "epoch": 0.26626323751891073,
+      "grad_norm": 0.8690244087376964,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.4506,
+      "step": 88
+    },
+    {
+      "epoch": 0.2692889561270802,
+      "grad_norm": 0.8279225783754451,
+      "learning_rate": 1.9926104680106484e-05,
+      "loss": 0.4642,
+      "step": 89
+    },
+    {
+      "epoch": 0.2723146747352496,
+      "grad_norm": 0.8253744592024457,
+      "learning_rate": 1.9919548128307954e-05,
+      "loss": 0.4774,
+      "step": 90
+    },
+    {
+      "epoch": 0.2753403933434191,
+      "grad_norm": 0.8390698421487144,
+      "learning_rate": 1.9912714105424694e-05,
+      "loss": 0.4692,
+      "step": 91
+    },
+    {
+      "epoch": 0.2783661119515885,
+      "grad_norm": 0.8290910659199547,
+      "learning_rate": 1.990560280261901e-05,
+      "loss": 0.4843,
+      "step": 92
+    },
+    {
+      "epoch": 0.2813918305597579,
+      "grad_norm": 0.863223875838923,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4493,
+      "step": 93
+    },
+    {
+      "epoch": 0.2844175491679274,
+      "grad_norm": 0.7983874023585454,
+      "learning_rate": 1.9890549160664633e-05,
+      "loss": 0.4853,
+      "step": 94
+    },
+    {
+      "epoch": 0.2874432677760968,
+      "grad_norm": 0.8311689528143783,
+      "learning_rate": 1.9882607242598663e-05,
+      "loss": 0.4577,
+      "step": 95
+    },
+    {
+      "epoch": 0.29046898638426627,
+      "grad_norm": 0.8285992328360374,
+      "learning_rate": 1.9874388886763944e-05,
+      "loss": 0.4603,
+      "step": 96
+    },
+    {
+      "epoch": 0.2934947049924357,
+      "grad_norm": 0.8172256009225088,
+      "learning_rate": 1.9865894323045558e-05,
+      "loss": 0.4555,
+      "step": 97
+    },
+    {
+      "epoch": 0.29652042360060515,
+      "grad_norm": 0.8981349586255105,
+      "learning_rate": 1.9857123789054707e-05,
+      "loss": 0.5054,
+      "step": 98
+    },
+    {
+      "epoch": 0.29954614220877457,
+      "grad_norm": 0.8357528632510638,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.4759,
+      "step": 99
+    },
+    {
+      "epoch": 0.30257186081694404,
+      "grad_norm": 0.8170861345107651,
+      "learning_rate": 1.9838755799290993e-05,
+      "loss": 0.47,
+      "step": 100
+    },
+    {
+      "epoch": 0.30559757942511345,
+      "grad_norm": 0.8537677551356737,
+      "learning_rate": 1.9829158857310288e-05,
+      "loss": 0.4651,
+      "step": 101
+    },
+    {
+      "epoch": 0.3086232980332829,
+      "grad_norm": 0.8067961658771471,
+      "learning_rate": 1.9819286972627066e-05,
+      "loss": 0.4863,
+      "step": 102
+    },
+    {
+      "epoch": 0.31164901664145234,
+      "grad_norm": 0.8746668191129228,
+      "learning_rate": 1.9809140421379168e-05,
+      "loss": 0.4947,
+      "step": 103
+    },
+    {
+      "epoch": 0.3146747352496218,
+      "grad_norm": 0.7829420787843964,
+      "learning_rate": 1.979871948738743e-05,
+      "loss": 0.4514,
+      "step": 104
+    },
+    {
+      "epoch": 0.3177004538577912,
+      "grad_norm": 0.8190562772049138,
+      "learning_rate": 1.978802446214779e-05,
+      "loss": 0.461,
+      "step": 105
+    },
+    {
+      "epoch": 0.3207261724659607,
+      "grad_norm": 0.8326480526282966,
+      "learning_rate": 1.9777055644823087e-05,
+      "loss": 0.4665,
+      "step": 106
+    },
+    {
+      "epoch": 0.3237518910741301,
+      "grad_norm": 0.7526458143224174,
+      "learning_rate": 1.9765813342234726e-05,
+      "loss": 0.4679,
+      "step": 107
+    },
+    {
+      "epoch": 0.3267776096822995,
+      "grad_norm": 0.8250298041525463,
+      "learning_rate": 1.9754297868854075e-05,
+      "loss": 0.4954,
+      "step": 108
+    },
+    {
+      "epoch": 0.329803328290469,
+      "grad_norm": 0.7908264887994304,
+      "learning_rate": 1.9742509546793673e-05,
+      "loss": 0.442,
+      "step": 109
+    },
+    {
+      "epoch": 0.3328290468986384,
+      "grad_norm": 0.763960765945944,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.4531,
+      "step": 110
+    },
+    {
+      "epoch": 0.3358547655068079,
+      "grad_norm": 0.8588828558635526,
+      "learning_rate": 1.9718115683235418e-05,
+      "loss": 0.4709,
+      "step": 111
+    },
+    {
+      "epoch": 0.3388804841149773,
+      "grad_norm": 0.7804012627965901,
+      "learning_rate": 1.970551082408636e-05,
+      "loss": 0.4581,
+      "step": 112
+    },
+    {
+      "epoch": 0.34190620272314676,
+      "grad_norm": 0.7275303337967168,
+      "learning_rate": 1.969263448093608e-05,
+      "loss": 0.4476,
+      "step": 113
+    },
+    {
+      "epoch": 0.34493192133131617,
+      "grad_norm": 0.7890144686288182,
+      "learning_rate": 1.9679487013963566e-05,
+      "loss": 0.4715,
+      "step": 114
+    },
+    {
+      "epoch": 0.34795763993948564,
+      "grad_norm": 0.7741873261309357,
+      "learning_rate": 1.9666068790931733e-05,
+      "loss": 0.4664,
+      "step": 115
+    },
+    {
+      "epoch": 0.35098335854765506,
+      "grad_norm": 0.8022176325597185,
+      "learning_rate": 1.9652380187177128e-05,
+      "loss": 0.4626,
+      "step": 116
+    },
+    {
+      "epoch": 0.3540090771558245,
+      "grad_norm": 0.9010514465090704,
+      "learning_rate": 1.9638421585599422e-05,
+      "loss": 0.479,
+      "step": 117
+    },
+    {
+      "epoch": 0.35703479576399394,
+      "grad_norm": 0.788537606981892,
+      "learning_rate": 1.9624193376650708e-05,
+      "loss": 0.4442,
+      "step": 118
+    },
+    {
+      "epoch": 0.3600605143721634,
+      "grad_norm": 0.8236642820280867,
+      "learning_rate": 1.960969595832457e-05,
+      "loss": 0.4553,
+      "step": 119
+    },
+    {
+      "epoch": 0.3630862329803328,
+      "grad_norm": 0.8636914308030584,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4892,
+      "step": 120
+    },
+    {
+      "epoch": 0.3661119515885023,
+      "grad_norm": 0.788482165986484,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.4566,
+      "step": 121
+    },
+    {
+      "epoch": 0.3691376701966717,
+      "grad_norm": 0.7273112240056128,
+      "learning_rate": 1.956459253990476e-05,
+      "loss": 0.4642,
+      "step": 122
+    },
+    {
+      "epoch": 0.3721633888048411,
+      "grad_norm": 0.801964247055127,
+      "learning_rate": 1.9549022414440738e-05,
+      "loss": 0.4671,
+      "step": 123
+    },
+    {
+      "epoch": 0.3751891074130106,
+      "grad_norm": 0.7939036863515218,
+      "learning_rate": 1.9533185182292705e-05,
+      "loss": 0.4786,
+      "step": 124
+    },
+    {
+      "epoch": 0.37821482602118,
+      "grad_norm": 0.7412356711962693,
+      "learning_rate": 1.9517081286462082e-05,
+      "loss": 0.4844,
+      "step": 125
+    },
+    {
+      "epoch": 0.3812405446293495,
+      "grad_norm": 0.7597404715537383,
+      "learning_rate": 1.9500711177409456e-05,
+      "loss": 0.4841,
+      "step": 126
+    },
+    {
+      "epoch": 0.3842662632375189,
+      "grad_norm": 0.748118267973097,
+      "learning_rate": 1.9484075313041968e-05,
+      "loss": 0.4582,
+      "step": 127
+    },
+    {
+      "epoch": 0.38729198184568836,
+      "grad_norm": 0.7632977171677049,
+      "learning_rate": 1.9467174158700507e-05,
+      "loss": 0.4462,
+      "step": 128
+    },
+    {
+      "epoch": 0.3903177004538578,
+      "grad_norm": 0.7671250061725821,
+      "learning_rate": 1.9450008187146685e-05,
+      "loss": 0.4219,
+      "step": 129
+    },
+    {
+      "epoch": 0.39334341906202724,
+      "grad_norm": 0.7229960998520764,
+      "learning_rate": 1.9432577878549635e-05,
+      "loss": 0.451,
+      "step": 130
+    },
+    {
+      "epoch": 0.39636913767019666,
+      "grad_norm": 0.7745544039196061,
+      "learning_rate": 1.9414883720472557e-05,
+      "loss": 0.4704,
+      "step": 131
+    },
+    {
+      "epoch": 0.39939485627836613,
+      "grad_norm": 0.7163363154235218,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.4788,
+      "step": 132
+    },
+    {
+      "epoch": 0.40242057488653554,
+      "grad_norm": 0.7889586882198709,
+      "learning_rate": 1.937870584301945e-05,
+      "loss": 0.4766,
+      "step": 133
+    },
+    {
+      "epoch": 0.405446293494705,
+      "grad_norm": 0.7555355595597854,
+      "learning_rate": 1.9360223135616423e-05,
+      "loss": 0.471,
+      "step": 134
+    },
+    {
+      "epoch": 0.4084720121028744,
+      "grad_norm": 0.8526265019841974,
+      "learning_rate": 1.9341478602651068e-05,
+      "loss": 0.4707,
+      "step": 135
+    },
+    {
+      "epoch": 0.4114977307110439,
+      "grad_norm": 0.7603688428468797,
+      "learning_rate": 1.932247276844826e-05,
+      "loss": 0.4499,
+      "step": 136
+    },
+    {
+      "epoch": 0.4145234493192133,
+      "grad_norm": 0.7219755089330839,
+      "learning_rate": 1.9303206164642037e-05,
+      "loss": 0.4624,
+      "step": 137
+    },
+    {
+      "epoch": 0.4175491679273828,
+      "grad_norm": 0.8027829909575993,
+      "learning_rate": 1.9283679330160726e-05,
+      "loss": 0.4711,
+      "step": 138
+    },
+    {
+      "epoch": 0.4205748865355522,
+      "grad_norm": 0.8012196033768283,
+      "learning_rate": 1.9263892811211865e-05,
+      "loss": 0.4853,
+      "step": 139
+    },
+    {
+      "epoch": 0.4236006051437216,
+      "grad_norm": 0.7244741532505259,
+      "learning_rate": 1.9243847161266924e-05,
+      "loss": 0.4618,
+      "step": 140
+    },
+    {
+      "epoch": 0.4266263237518911,
+      "grad_norm": 0.8058867979616773,
+      "learning_rate": 1.9223542941045817e-05,
+      "loss": 0.4822,
+      "step": 141
+    },
+    {
+      "epoch": 0.4296520423600605,
+      "grad_norm": 0.7353699629577677,
+      "learning_rate": 1.920298071850123e-05,
+      "loss": 0.4671,
+      "step": 142
+    },
+    {
+      "epoch": 0.43267776096822996,
+      "grad_norm": 0.7619972611080683,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.453,
+      "step": 143
+    },
+    {
+      "epoch": 0.4357034795763994,
+      "grad_norm": 0.835751661426491,
+      "learning_rate": 1.9161084574320696e-05,
+      "loss": 0.455,
+      "step": 144
+    },
+    {
+      "epoch": 0.43872919818456885,
+      "grad_norm": 0.8235992602935628,
+      "learning_rate": 1.913975182460996e-05,
+      "loss": 0.4712,
+      "step": 145
+    },
+    {
+      "epoch": 0.44175491679273826,
+      "grad_norm": 0.7710504940357691,
+      "learning_rate": 1.9118163416393392e-05,
+      "loss": 0.4502,
+      "step": 146
+    },
+    {
+      "epoch": 0.44478063540090773,
+      "grad_norm": 0.789164619571882,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4601,
+      "step": 147
+    },
+    {
+      "epoch": 0.44780635400907715,
+      "grad_norm": 0.7804846045281089,
+      "learning_rate": 1.9074222047073945e-05,
+      "loss": 0.4829,
+      "step": 148
+    },
+    {
+      "epoch": 0.4508320726172466,
+      "grad_norm": 0.7488730502331767,
+      "learning_rate": 1.9051870315105626e-05,
+      "loss": 0.4798,
+      "step": 149
+    },
+    {
+      "epoch": 0.45385779122541603,
+      "grad_norm": 0.7577557036184778,
+      "learning_rate": 1.9029265382866216e-05,
+      "loss": 0.455,
+      "step": 150
+    },
+    {
+      "epoch": 0.4568835098335855,
+      "grad_norm": 0.8323906425833675,
+      "learning_rate": 1.9006407882664256e-05,
+      "loss": 0.4573,
+      "step": 151
+    },
+    {
+      "epoch": 0.4599092284417549,
+      "grad_norm": 0.7355121464891837,
+      "learning_rate": 1.8983298453873172e-05,
+      "loss": 0.4433,
+      "step": 152
+    },
+    {
+      "epoch": 0.4629349470499244,
+      "grad_norm": 0.7953485201434655,
+      "learning_rate": 1.895993774291336e-05,
+      "loss": 0.4493,
+      "step": 153
+    },
+    {
+      "epoch": 0.4659606656580938,
+      "grad_norm": 0.7597317920270753,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.435,
+      "step": 154
+    },
+    {
+      "epoch": 0.4689863842662632,
+      "grad_norm": 0.7868266495647096,
+      "learning_rate": 1.891246509529539e-05,
+      "loss": 0.4579,
+      "step": 155
+    },
+    {
+      "epoch": 0.4720121028744327,
+      "grad_norm": 0.8624718797927682,
+      "learning_rate": 1.8888354486549238e-05,
+      "loss": 0.4659,
+      "step": 156
+    },
+    {
+      "epoch": 0.4750378214826021,
+      "grad_norm": 0.7382694605135437,
+      "learning_rate": 1.886399525142122e-05,
+      "loss": 0.4561,
+      "step": 157
+    },
+    {
+      "epoch": 0.47806354009077157,
+      "grad_norm": 0.7778399686637706,
+      "learning_rate": 1.8839388071291506e-05,
+      "loss": 0.4579,
+      "step": 158
+    },
+    {
+      "epoch": 0.481089258698941,
+      "grad_norm": 0.7610693699342035,
+      "learning_rate": 1.881453363447582e-05,
+      "loss": 0.4755,
+      "step": 159
+    },
+    {
+      "epoch": 0.48411497730711045,
+      "grad_norm": 0.82539836353124,
+      "learning_rate": 1.8789432636206197e-05,
+      "loss": 0.4388,
+      "step": 160
+    },
+    {
+      "epoch": 0.48714069591527986,
+      "grad_norm": 0.8228496733835307,
+      "learning_rate": 1.8764085778611507e-05,
+      "loss": 0.4752,
+      "step": 161
+    },
+    {
+      "epoch": 0.49016641452344933,
+      "grad_norm": 0.7857900924336991,
+      "learning_rate": 1.873849377069785e-05,
+      "loss": 0.4236,
+      "step": 162
+    },
+    {
+      "epoch": 0.49319213313161875,
+      "grad_norm": 0.7931503240984769,
+      "learning_rate": 1.87126573283287e-05,
+      "loss": 0.4603,
+      "step": 163
+    },
+    {
+      "epoch": 0.4962178517397882,
+      "grad_norm": 0.813025896979347,
+      "learning_rate": 1.8686577174204887e-05,
+      "loss": 0.4678,
+      "step": 164
+    },
+    {
+      "epoch": 0.49924357034795763,
+      "grad_norm": 0.7914366513974338,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.4385,
+      "step": 165
+    },
+    {
+      "epoch": 0.5022692889561271,
+      "grad_norm": 0.7583142372697701,
+      "learning_rate": 1.863368865556191e-05,
+      "loss": 0.4429,
+      "step": 166
+    },
+    {
+      "epoch": 0.5052950075642966,
+      "grad_norm": 0.7563736532557938,
+      "learning_rate": 1.8606881770448305e-05,
+      "loss": 0.4492,
+      "step": 167
+    },
+    {
+      "epoch": 0.5083207261724659,
+      "grad_norm": 0.7288685989405779,
+      "learning_rate": 1.8579834132349773e-05,
+      "loss": 0.4436,
+      "step": 168
+    },
+    {
+      "epoch": 0.5113464447806354,
+      "grad_norm": 0.714299073646468,
+      "learning_rate": 1.8552546497846893e-05,
+      "loss": 0.4547,
+      "step": 169
+    },
+    {
+      "epoch": 0.5143721633888049,
+      "grad_norm": 0.787604701934352,
+      "learning_rate": 1.8525019630233463e-05,
+      "loss": 0.4778,
+      "step": 170
+    },
+    {
+      "epoch": 0.5173978819969742,
+      "grad_norm": 0.7446032561487276,
+      "learning_rate": 1.8497254299495147e-05,
+      "loss": 0.4392,
+      "step": 171
+    },
+    {
+      "epoch": 0.5204236006051437,
+      "grad_norm": 0.8676076037724432,
+      "learning_rate": 1.8469251282287925e-05,
+      "loss": 0.4555,
+      "step": 172
+    },
+    {
+      "epoch": 0.5234493192133132,
+      "grad_norm": 0.8199309565532229,
+      "learning_rate": 1.8441011361916387e-05,
+      "loss": 0.439,
+      "step": 173
+    },
+    {
+      "epoch": 0.5264750378214826,
+      "grad_norm": 0.8787954120755868,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4667,
+      "step": 174
+    },
+    {
+      "epoch": 0.529500756429652,
+      "grad_norm": 0.7862271259579355,
+      "learning_rate": 1.8383823978010077e-05,
+      "loss": 0.4643,
+      "step": 175
+    },
+    {
+      "epoch": 0.5325264750378215,
+      "grad_norm": 0.7537982206692815,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.426,
+      "step": 176
+    },
+    {
+      "epoch": 0.5355521936459909,
+      "grad_norm": 0.9069634996997274,
+      "learning_rate": 1.8325698546347714e-05,
+      "loss": 0.4757,
+      "step": 177
+    },
+    {
+      "epoch": 0.5385779122541604,
+      "grad_norm": 0.7332754254206979,
+      "learning_rate": 1.8296286090880362e-05,
+      "loss": 0.4663,
+      "step": 178
+    },
+    {
+      "epoch": 0.5416036308623298,
+      "grad_norm": 0.7896029677267946,
+      "learning_rate": 1.8266641570456915e-05,
+      "loss": 0.4498,
+      "step": 179
+    },
+    {
+      "epoch": 0.5446293494704992,
+      "grad_norm": 0.7030809709931444,
+      "learning_rate": 1.8236765814298328e-05,
+      "loss": 0.462,
+      "step": 180
+    },
+    {
+      "epoch": 0.5476550680786687,
+      "grad_norm": 0.8192255527574019,
+      "learning_rate": 1.820665965809373e-05,
+      "loss": 0.4533,
+      "step": 181
+    },
+    {
+      "epoch": 0.5506807866868382,
+      "grad_norm": 0.8086912792802334,
+      "learning_rate": 1.8176323943977034e-05,
+      "loss": 0.4536,
+      "step": 182
+    },
+    {
+      "epoch": 0.5537065052950075,
+      "grad_norm": 0.9646929897203245,
+      "learning_rate": 1.814575952050336e-05,
+      "loss": 0.4549,
+      "step": 183
+    },
+    {
+      "epoch": 0.556732223903177,
+      "grad_norm": 0.7893710805237983,
+      "learning_rate": 1.8114967242625342e-05,
+      "loss": 0.4406,
+      "step": 184
+    },
+    {
+      "epoch": 0.5597579425113465,
+      "grad_norm": 0.747405301682776,
+      "learning_rate": 1.808394797166919e-05,
+      "loss": 0.4531,
+      "step": 185
+    },
+    {
+      "epoch": 0.5627836611195158,
+      "grad_norm": 0.7689727269918262,
+      "learning_rate": 1.8052702575310588e-05,
+      "loss": 0.4156,
+      "step": 186
+    },
+    {
+      "epoch": 0.5658093797276853,
+      "grad_norm": 0.7531788137398425,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.4564,
+      "step": 187
+    },
+    {
+      "epoch": 0.5688350983358548,
+      "grad_norm": 0.7024466197581899,
+      "learning_rate": 1.7989536908690413e-05,
+      "loss": 0.4268,
+      "step": 188
+    },
+    {
+      "epoch": 0.5718608169440242,
+      "grad_norm": 0.8356315060244377,
+      "learning_rate": 1.7957618405308323e-05,
+      "loss": 0.4661,
+      "step": 189
+    },
+    {
+      "epoch": 0.5748865355521936,
+      "grad_norm": 0.7479426784232934,
+      "learning_rate": 1.792547731023332e-05,
+      "loss": 0.4388,
+      "step": 190
+    },
+    {
+      "epoch": 0.5779122541603631,
+      "grad_norm": 0.6997465257129355,
+      "learning_rate": 1.789311452252092e-05,
+      "loss": 0.4302,
+      "step": 191
+    },
+    {
+      "epoch": 0.5809379727685325,
+      "grad_norm": 0.7330574904453382,
+      "learning_rate": 1.7860530947427878e-05,
+      "loss": 0.4401,
+      "step": 192
+    },
+    {
+      "epoch": 0.583963691376702,
+      "grad_norm": 0.7200579183910508,
+      "learning_rate": 1.782772749638682e-05,
+      "loss": 0.4322,
+      "step": 193
+    },
+    {
+      "epoch": 0.5869894099848714,
+      "grad_norm": 0.7531714605191969,
+      "learning_rate": 1.779470508698079e-05,
+      "loss": 0.4326,
+      "step": 194
+    },
+    {
+      "epoch": 0.5900151285930408,
+      "grad_norm": 0.704412670653764,
+      "learning_rate": 1.776146464291757e-05,
+      "loss": 0.4462,
+      "step": 195
+    },
+    {
+      "epoch": 0.5930408472012103,
+      "grad_norm": 0.7614643543847789,
+      "learning_rate": 1.772800709400383e-05,
+      "loss": 0.4725,
+      "step": 196
+    },
+    {
+      "epoch": 0.5960665658093798,
+      "grad_norm": 0.7324045163381419,
+      "learning_rate": 1.7694333376119144e-05,
+      "loss": 0.4237,
+      "step": 197
+    },
+    {
+      "epoch": 0.5990922844175491,
+      "grad_norm": 0.7979902027529663,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.433,
+      "step": 198
+    },
+    {
+      "epoch": 0.6021180030257186,
+      "grad_norm": 0.6913582928480547,
+      "learning_rate": 1.762634120716238e-05,
+      "loss": 0.4372,
+      "step": 199
+    },
+    {
+      "epoch": 0.6051437216338881,
+      "grad_norm": 0.7252349239170433,
+      "learning_rate": 1.7592024657977432e-05,
+      "loss": 0.4631,
+      "step": 200
+    },
+    {
+      "epoch": 0.6081694402420574,
+      "grad_norm": 0.6712624258843044,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4512,
+      "step": 201
+    },
+    {
+      "epoch": 0.6111951588502269,
+      "grad_norm": 0.7006452839594959,
+      "learning_rate": 1.75227554297058e-05,
+      "loss": 0.4312,
+      "step": 202
+    },
+    {
+      "epoch": 0.6142208774583964,
+      "grad_norm": 0.6872991486250615,
+      "learning_rate": 1.7487804688228327e-05,
+      "loss": 0.4474,
+      "step": 203
+    },
+    {
+      "epoch": 0.6172465960665658,
+      "grad_norm": 0.700936854042298,
+      "learning_rate": 1.745264449675755e-05,
+      "loss": 0.4371,
+      "step": 204
+    },
+    {
+      "epoch": 0.6202723146747352,
+      "grad_norm": 0.7399346477963574,
+      "learning_rate": 1.7417275838799596e-05,
+      "loss": 0.4285,
+      "step": 205
+    },
+    {
+      "epoch": 0.6232980332829047,
+      "grad_norm": 0.7237193875173202,
+      "learning_rate": 1.7381699703691866e-05,
+      "loss": 0.4451,
+      "step": 206
+    },
+    {
+      "epoch": 0.6263237518910741,
+      "grad_norm": 0.6788729286708666,
+      "learning_rate": 1.734591708657533e-05,
+      "loss": 0.4262,
+      "step": 207
+    },
+    {
+      "epoch": 0.6293494704992436,
+      "grad_norm": 0.7869303974657352,
+      "learning_rate": 1.730992898836672e-05,
+      "loss": 0.4273,
+      "step": 208
+    },
+    {
+      "epoch": 0.632375189107413,
+      "grad_norm": 0.6985269939651845,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.4081,
+      "step": 209
+    },
+    {
+      "epoch": 0.6354009077155824,
+      "grad_norm": 0.7043355646685842,
+      "learning_rate": 1.72373403810507e-05,
+      "loss": 0.4688,
+      "step": 210
+    },
+    {
+      "epoch": 0.6384266263237519,
+      "grad_norm": 0.8112888174013457,
+      "learning_rate": 1.720074190240269e-05,
+      "loss": 0.445,
+      "step": 211
+    },
+    {
+      "epoch": 0.6414523449319214,
+      "grad_norm": 0.8119097064290122,
+      "learning_rate": 1.7163942003524574e-05,
+      "loss": 0.4668,
+      "step": 212
+    },
+    {
+      "epoch": 0.6444780635400907,
+      "grad_norm": 0.6793303011607158,
+      "learning_rate": 1.7126941713788633e-05,
+      "loss": 0.4579,
+      "step": 213
+    },
+    {
+      "epoch": 0.6475037821482602,
+      "grad_norm": 0.7283067581251398,
+      "learning_rate": 1.70897420681725e-05,
+      "loss": 0.4402,
+      "step": 214
+    },
+    {
+      "epoch": 0.6505295007564297,
+      "grad_norm": 0.7073994562538528,
+      "learning_rate": 1.7052344107230244e-05,
+      "loss": 0.4257,
+      "step": 215
+    },
+    {
+      "epoch": 0.653555219364599,
+      "grad_norm": 0.7265949920471793,
+      "learning_rate": 1.7014748877063212e-05,
+      "loss": 0.4375,
+      "step": 216
+    },
+    {
+      "epoch": 0.6565809379727685,
+      "grad_norm": 0.7698382482177225,
+      "learning_rate": 1.697695742929082e-05,
+      "loss": 0.4298,
+      "step": 217
+    },
+    {
+      "epoch": 0.659606656580938,
+      "grad_norm": 0.7496819774918994,
+      "learning_rate": 1.693897082102109e-05,
+      "loss": 0.4397,
+      "step": 218
+    },
+    {
+      "epoch": 0.6626323751891074,
+      "grad_norm": 0.7039449616359452,
+      "learning_rate": 1.6900790114821122e-05,
+      "loss": 0.432,
+      "step": 219
+    },
+    {
+      "epoch": 0.6656580937972768,
+      "grad_norm": 0.8323041379562198,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.4289,
+      "step": 220
+    },
+    {
+      "epoch": 0.6686838124054463,
+      "grad_norm": 0.6964371377304952,
+      "learning_rate": 1.682385068601563e-05,
+      "loss": 0.4272,
+      "step": 221
+    },
+    {
+      "epoch": 0.6717095310136157,
+      "grad_norm": 0.9037279845818564,
+      "learning_rate": 1.6785094115571323e-05,
+      "loss": 0.4173,
+      "step": 222
+    },
+    {
+      "epoch": 0.6747352496217852,
+      "grad_norm": 0.7240473776819164,
+      "learning_rate": 1.674614775145901e-05,
+      "loss": 0.4426,
+      "step": 223
+    },
+    {
+      "epoch": 0.6777609682299546,
+      "grad_norm": 0.7895012092481307,
+      "learning_rate": 1.670701268309221e-05,
+      "loss": 0.4424,
+      "step": 224
+    },
+    {
+      "epoch": 0.680786686838124,
+      "grad_norm": 0.7023859268429421,
+      "learning_rate": 1.666769000516292e-05,
+      "loss": 0.4271,
+      "step": 225
+    },
+    {
+      "epoch": 0.6838124054462935,
+      "grad_norm": 0.7710898133564811,
+      "learning_rate": 1.6628180817610963e-05,
+      "loss": 0.4383,
+      "step": 226
+    },
+    {
+      "epoch": 0.686838124054463,
+      "grad_norm": 0.7755971542693247,
+      "learning_rate": 1.658848622559325e-05,
+      "loss": 0.4159,
+      "step": 227
+    },
+    {
+      "epoch": 0.6898638426626323,
+      "grad_norm": 0.7454368496364951,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4264,
+      "step": 228
+    },
+    {
+      "epoch": 0.6928895612708018,
+      "grad_norm": 0.7465023342200379,
+      "learning_rate": 1.6508545274687936e-05,
+      "loss": 0.4714,
+      "step": 229
+    },
+    {
+      "epoch": 0.6959152798789713,
+      "grad_norm": 0.9340964111220721,
+      "learning_rate": 1.6468301151920576e-05,
+      "loss": 0.4498,
+      "step": 230
+    },
+    {
+      "epoch": 0.6989409984871406,
+      "grad_norm": 0.7229574222951994,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.451,
+      "step": 231
+    },
+    {
+      "epoch": 0.7019667170953101,
+      "grad_norm": 0.8229731529416733,
+      "learning_rate": 1.6387271240298082e-05,
+      "loss": 0.4488,
+      "step": 232
+    },
+    {
+      "epoch": 0.7049924357034796,
+      "grad_norm": 0.755061957480007,
+      "learning_rate": 1.6346487718023762e-05,
+      "loss": 0.4458,
+      "step": 233
+    },
+    {
+      "epoch": 0.708018154311649,
+      "grad_norm": 0.8486780857509837,
+      "learning_rate": 1.6305526670845225e-05,
+      "loss": 0.4375,
+      "step": 234
+    },
+    {
+      "epoch": 0.7110438729198184,
+      "grad_norm": 0.7646369011902477,
+      "learning_rate": 1.6264389244531015e-05,
+      "loss": 0.4272,
+      "step": 235
+    },
+    {
+      "epoch": 0.7140695915279879,
+      "grad_norm": 0.7792536412112867,
+      "learning_rate": 1.6223076589783368e-05,
+      "loss": 0.4505,
+      "step": 236
+    },
+    {
+      "epoch": 0.7170953101361573,
+      "grad_norm": 0.7819926086008847,
+      "learning_rate": 1.6181589862206053e-05,
+      "loss": 0.4557,
+      "step": 237
+    },
+    {
+      "epoch": 0.7201210287443268,
+      "grad_norm": 0.8504168375118846,
+      "learning_rate": 1.613993022227202e-05,
+      "loss": 0.4763,
+      "step": 238
+    },
+    {
+      "epoch": 0.7231467473524962,
+      "grad_norm": 0.6451316703623551,
+      "learning_rate": 1.6098098835290955e-05,
+      "loss": 0.4047,
+      "step": 239
+    },
+    {
+      "epoch": 0.7261724659606656,
+      "grad_norm": 0.8484610947300326,
+      "learning_rate": 1.6056096871376667e-05,
+      "loss": 0.4168,
+      "step": 240
+    },
+    {
+      "epoch": 0.7291981845688351,
+      "grad_norm": 0.6648807009441984,
+      "learning_rate": 1.6013925505414386e-05,
+      "loss": 0.457,
+      "step": 241
+    },
+    {
+      "epoch": 0.7322239031770046,
+      "grad_norm": 0.7132046576181856,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.4456,
+      "step": 242
+    },
+    {
+      "epoch": 0.735249621785174,
+      "grad_norm": 0.7333550214146264,
+      "learning_rate": 1.5929079290546408e-05,
+      "loss": 0.4303,
+      "step": 243
+    },
+    {
+      "epoch": 0.7382753403933434,
+      "grad_norm": 0.6644537177090747,
+      "learning_rate": 1.5886406814971728e-05,
+      "loss": 0.44,
+      "step": 244
+    },
+    {
+      "epoch": 0.7413010590015129,
+      "grad_norm": 0.6774992684759596,
+      "learning_rate": 1.584356968394471e-05,
+      "loss": 0.4292,
+      "step": 245
+    },
+    {
+      "epoch": 0.7443267776096822,
+      "grad_norm": 0.7592483585417094,
+      "learning_rate": 1.5800569095711983e-05,
+      "loss": 0.4705,
+      "step": 246
+    },
+    {
+      "epoch": 0.7473524962178517,
+      "grad_norm": 0.7469371609485048,
+      "learning_rate": 1.575740625309244e-05,
+      "loss": 0.4722,
+      "step": 247
+    },
+    {
+      "epoch": 0.7503782148260212,
+      "grad_norm": 0.6939456648248022,
+      "learning_rate": 1.5714082363443576e-05,
+      "loss": 0.4385,
+      "step": 248
+    },
+    {
+      "epoch": 0.7534039334341907,
+      "grad_norm": 0.7169779677163522,
+      "learning_rate": 1.5670598638627707e-05,
+      "loss": 0.437,
+      "step": 249
+    },
+    {
+      "epoch": 0.75642965204236,
+      "grad_norm": 0.6650315770512617,
+      "learning_rate": 1.5626956294978103e-05,
+      "loss": 0.4316,
+      "step": 250
+    },
+    {
+      "epoch": 0.7594553706505295,
+      "grad_norm": 0.6549057774150726,
+      "learning_rate": 1.5583156553264923e-05,
+      "loss": 0.4215,
+      "step": 251
+    },
+    {
+      "epoch": 0.762481089258699,
+      "grad_norm": 0.6832351631016924,
+      "learning_rate": 1.5539200638661106e-05,
+      "loss": 0.4319,
+      "step": 252
+    },
+    {
+      "epoch": 0.7655068078668684,
+      "grad_norm": 0.6602455283406224,
+      "learning_rate": 1.5495089780708062e-05,
+      "loss": 0.4599,
+      "step": 253
+    },
+    {
+      "epoch": 0.7685325264750378,
+      "grad_norm": 0.7833708568531678,
+      "learning_rate": 1.5450825213281317e-05,
+      "loss": 0.4327,
+      "step": 254
+    },
+    {
+      "epoch": 0.7715582450832073,
+      "grad_norm": 0.6904164868039643,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4401,
+      "step": 255
+    },
+    {
+      "epoch": 0.7745839636913767,
+      "grad_norm": 0.6683376808573165,
+      "learning_rate": 1.5361839906972095e-05,
+      "loss": 0.4134,
+      "step": 256
+    },
+    {
+      "epoch": 0.7776096822995462,
+      "grad_norm": 0.6964663755101195,
+      "learning_rate": 1.531712165719992e-05,
+      "loss": 0.4275,
+      "step": 257
+    },
+    {
+      "epoch": 0.7806354009077155,
+      "grad_norm": 0.6767851981213515,
+      "learning_rate": 1.5272254676105026e-05,
+      "loss": 0.4157,
+      "step": 258
+    },
+    {
+      "epoch": 0.783661119515885,
+      "grad_norm": 0.7372064106369701,
+      "learning_rate": 1.5227240218713326e-05,
+      "loss": 0.4196,
+      "step": 259
+    },
+    {
+      "epoch": 0.7866868381240545,
+      "grad_norm": 0.6876658464929606,
+      "learning_rate": 1.5182079544175957e-05,
+      "loss": 0.4431,
+      "step": 260
+    },
+    {
+      "epoch": 0.789712556732224,
+      "grad_norm": 0.7293057657272566,
+      "learning_rate": 1.5136773915734067e-05,
+      "loss": 0.4394,
+      "step": 261
+    },
+    {
+      "epoch": 0.7927382753403933,
+      "grad_norm": 0.6433606180863638,
+      "learning_rate": 1.5091324600683472e-05,
+      "loss": 0.4274,
+      "step": 262
+    },
+    {
+      "epoch": 0.7957639939485628,
+      "grad_norm": 0.7832243375268928,
+      "learning_rate": 1.5045732870339213e-05,
+      "loss": 0.403,
+      "step": 263
+    },
+    {
+      "epoch": 0.7987897125567323,
+      "grad_norm": 0.7837140811330554,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.434,
+      "step": 264
+    },
+    {
+      "epoch": 0.8018154311649016,
+      "grad_norm": 0.8065783805006891,
+      "learning_rate": 1.4954127268912525e-05,
+      "loss": 0.4329,
+      "step": 265
+    },
+    {
+      "epoch": 0.8048411497730711,
+      "grad_norm": 0.8854444547179848,
+      "learning_rate": 1.4908115960235683e-05,
+      "loss": 0.4468,
+      "step": 266
+    },
+    {
+      "epoch": 0.8078668683812406,
+      "grad_norm": 0.7052159845296222,
+      "learning_rate": 1.4861967361004687e-05,
+      "loss": 0.4328,
+      "step": 267
+    },
+    {
+      "epoch": 0.81089258698941,
+      "grad_norm": 0.8590011359424397,
+      "learning_rate": 1.4815682762095065e-05,
+      "loss": 0.4215,
+      "step": 268
+    },
+    {
+      "epoch": 0.8139183055975794,
+      "grad_norm": 1.0391907242005918,
+      "learning_rate": 1.476926345818654e-05,
+      "loss": 0.437,
+      "step": 269
+    },
+    {
+      "epoch": 0.8169440242057489,
+      "grad_norm": 0.732112114588027,
+      "learning_rate": 1.472271074772683e-05,
+      "loss": 0.4409,
+      "step": 270
+    },
+    {
+      "epoch": 0.8199697428139183,
+      "grad_norm": 0.9632854409246182,
+      "learning_rate": 1.4676025932895315e-05,
+      "loss": 0.4205,
+      "step": 271
+    },
+    {
+      "epoch": 0.8229954614220878,
+      "grad_norm": 0.8191601072499728,
+      "learning_rate": 1.4629210319566626e-05,
+      "loss": 0.4061,
+      "step": 272
+    },
+    {
+      "epoch": 0.8260211800302572,
+      "grad_norm": 0.8097019439533432,
+      "learning_rate": 1.4582265217274105e-05,
+      "loss": 0.4491,
+      "step": 273
+    },
+    {
+      "epoch": 0.8290468986384266,
+      "grad_norm": 1.014790235820325,
+      "learning_rate": 1.4535191939173179e-05,
+      "loss": 0.437,
+      "step": 274
+    },
+    {
+      "epoch": 0.8320726172465961,
+      "grad_norm": 0.7701803327681169,
+      "learning_rate": 1.4487991802004625e-05,
+      "loss": 0.4165,
+      "step": 275
+    },
+    {
+      "epoch": 0.8350983358547656,
+      "grad_norm": 0.7937839855681287,
+      "learning_rate": 1.4440666126057743e-05,
+      "loss": 0.4327,
+      "step": 276
+    },
+    {
+      "epoch": 0.8381240544629349,
+      "grad_norm": 0.949372760087925,
+      "learning_rate": 1.4393216235133427e-05,
+      "loss": 0.413,
+      "step": 277
+    },
+    {
+      "epoch": 0.8411497730711044,
+      "grad_norm": 0.7125435443836725,
+      "learning_rate": 1.4345643456507126e-05,
+      "loss": 0.4229,
+      "step": 278
+    },
+    {
+      "epoch": 0.8441754916792739,
+      "grad_norm": 0.7321249704174541,
+      "learning_rate": 1.4297949120891718e-05,
+      "loss": 0.4201,
+      "step": 279
+    },
+    {
+      "epoch": 0.8472012102874432,
+      "grad_norm": 0.7653630943456802,
+      "learning_rate": 1.4250134562400301e-05,
+      "loss": 0.4368,
+      "step": 280
+    },
+    {
+      "epoch": 0.8502269288956127,
+      "grad_norm": 0.8932077521523663,
+      "learning_rate": 1.4202201118508863e-05,
+      "loss": 0.4304,
+      "step": 281
+    },
+    {
+      "epoch": 0.8532526475037822,
+      "grad_norm": 0.7109215784650246,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4393,
+      "step": 282
+    },
+    {
+      "epoch": 0.8562783661119516,
+      "grad_norm": 0.893893571443574,
+      "learning_rate": 1.4105982941019751e-05,
+      "loss": 0.4589,
+      "step": 283
+    },
+    {
+      "epoch": 0.859304084720121,
+      "grad_norm": 0.9140197721621929,
+      "learning_rate": 1.405770089885134e-05,
+      "loss": 0.4292,
+      "step": 284
+    },
+    {
+      "epoch": 0.8623298033282905,
+      "grad_norm": 0.6883441319916367,
+      "learning_rate": 1.4009305354066138e-05,
+      "loss": 0.4463,
+      "step": 285
+    },
+    {
+      "epoch": 0.8653555219364599,
+      "grad_norm": 0.9064908510058771,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.4105,
+      "step": 286
+    },
+    {
+      "epoch": 0.8683812405446294,
+      "grad_norm": 1.024419540771557,
+      "learning_rate": 1.39121791746921e-05,
+      "loss": 0.4327,
+      "step": 287
+    },
+    {
+      "epoch": 0.8714069591527988,
+      "grad_norm": 0.6783916491342301,
+      "learning_rate": 1.3863451256931286e-05,
+      "loss": 0.4133,
+      "step": 288
+    },
+    {
+      "epoch": 0.8744326777609682,
+      "grad_norm": 0.8056925286593789,
+      "learning_rate": 1.381461527013374e-05,
+      "loss": 0.4179,
+      "step": 289
+    },
+    {
+      "epoch": 0.8774583963691377,
+      "grad_norm": 0.9507113711299003,
+      "learning_rate": 1.3765672580346986e-05,
+      "loss": 0.4548,
+      "step": 290
+    },
+    {
+      "epoch": 0.8804841149773072,
+      "grad_norm": 0.6850554340324797,
+      "learning_rate": 1.3716624556603275e-05,
+      "loss": 0.4309,
+      "step": 291
+    },
+    {
+      "epoch": 0.8835098335854765,
+      "grad_norm": 0.739044015483832,
+      "learning_rate": 1.3667472570881264e-05,
+      "loss": 0.43,
+      "step": 292
+    },
+    {
+      "epoch": 0.886535552193646,
+      "grad_norm": 0.8704823212754892,
+      "learning_rate": 1.361821799806765e-05,
+      "loss": 0.4355,
+      "step": 293
+    },
+    {
+      "epoch": 0.8895612708018155,
+      "grad_norm": 0.6775255179236076,
+      "learning_rate": 1.356886221591872e-05,
+      "loss": 0.4163,
+      "step": 294
+    },
+    {
+      "epoch": 0.8925869894099848,
+      "grad_norm": 0.6773001946649285,
+      "learning_rate": 1.3519406605021797e-05,
+      "loss": 0.4072,
+      "step": 295
+    },
+    {
+      "epoch": 0.8956127080181543,
+      "grad_norm": 0.7539343578167521,
+      "learning_rate": 1.3469852548756626e-05,
+      "loss": 0.4095,
+      "step": 296
+    },
+    {
+      "epoch": 0.8986384266263238,
+      "grad_norm": 0.8705998547340024,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.444,
+      "step": 297
+    },
+    {
+      "epoch": 0.9016641452344932,
+      "grad_norm": 0.7549497968943366,
+      "learning_rate": 1.3370454647370418e-05,
+      "loss": 0.4359,
+      "step": 298
+    },
+    {
+      "epoch": 0.9046898638426626,
+      "grad_norm": 0.7929687829781253,
+      "learning_rate": 1.3320613582622354e-05,
+      "loss": 0.4269,
+      "step": 299
+    },
+    {
+      "epoch": 0.9077155824508321,
+      "grad_norm": 0.802151550709066,
+      "learning_rate": 1.3270679633174219e-05,
+      "loss": 0.4365,
+      "step": 300
+    },
+    {
+      "epoch": 0.9107413010590015,
+      "grad_norm": 0.8814222010379341,
+      "learning_rate": 1.3220654195785917e-05,
+      "loss": 0.4422,
+      "step": 301
+    },
+    {
+      "epoch": 0.913767019667171,
+      "grad_norm": 0.7187453287090642,
+      "learning_rate": 1.3170538669776469e-05,
+      "loss": 0.4509,
+      "step": 302
+    },
+    {
+      "epoch": 0.9167927382753404,
+      "grad_norm": 0.7771910008351685,
+      "learning_rate": 1.3120334456984871e-05,
+      "loss": 0.4348,
+      "step": 303
+    },
+    {
+      "epoch": 0.9198184568835098,
+      "grad_norm": 0.8949713661413075,
+      "learning_rate": 1.3070042961730878e-05,
+      "loss": 0.4378,
+      "step": 304
+    },
+    {
+      "epoch": 0.9228441754916793,
+      "grad_norm": 0.6353705395444028,
+      "learning_rate": 1.3019665590775717e-05,
+      "loss": 0.3939,
+      "step": 305
+    },
+    {
+      "epoch": 0.9258698940998488,
+      "grad_norm": 0.8128777025871372,
+      "learning_rate": 1.296920375328275e-05,
+      "loss": 0.4471,
+      "step": 306
+    },
+    {
+      "epoch": 0.9288956127080181,
+      "grad_norm": 0.8744993932668976,
+      "learning_rate": 1.2918658860778046e-05,
+      "loss": 0.4322,
+      "step": 307
+    },
+    {
+      "epoch": 0.9319213313161876,
+      "grad_norm": 0.6904901525332652,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.4267,
+      "step": 308
+    },
+    {
+      "epoch": 0.9349470499243571,
+      "grad_norm": 0.709445884988274,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.4459,
+      "step": 309
+    },
+    {
+      "epoch": 0.9379727685325264,
+      "grad_norm": 0.8137042866966601,
+      "learning_rate": 1.2766540003065272e-05,
+      "loss": 0.4098,
+      "step": 310
+    },
+    {
+      "epoch": 0.9409984871406959,
+      "grad_norm": 0.6961338803344621,
+      "learning_rate": 1.2715677051645259e-05,
+      "loss": 0.428,
+      "step": 311
+    },
+    {
+      "epoch": 0.9440242057488654,
+      "grad_norm": 0.6753747864465509,
+      "learning_rate": 1.266473813690035e-05,
+      "loss": 0.433,
+      "step": 312
+    },
+    {
+      "epoch": 0.9470499243570348,
+      "grad_norm": 0.6944475467370473,
+      "learning_rate": 1.2613724683701491e-05,
+      "loss": 0.4192,
+      "step": 313
+    },
+    {
+      "epoch": 0.9500756429652042,
+      "grad_norm": 0.766769014909466,
+      "learning_rate": 1.2562638119004627e-05,
+      "loss": 0.4322,
+      "step": 314
+    },
+    {
+      "epoch": 0.9531013615733737,
+      "grad_norm": 0.6573325916988719,
+      "learning_rate": 1.2511479871810792e-05,
+      "loss": 0.4341,
+      "step": 315
+    },
+    {
+      "epoch": 0.9561270801815431,
+      "grad_norm": 0.712135465765564,
+      "learning_rate": 1.2460251373126136e-05,
+      "loss": 0.3973,
+      "step": 316
+    },
+    {
+      "epoch": 0.9591527987897126,
+      "grad_norm": 0.7376765494220082,
+      "learning_rate": 1.2408954055921884e-05,
+      "loss": 0.4106,
+      "step": 317
+    },
+    {
+      "epoch": 0.962178517397882,
+      "grad_norm": 0.6627556537154219,
+      "learning_rate": 1.2357589355094275e-05,
+      "loss": 0.435,
+      "step": 318
+    },
+    {
+      "epoch": 0.9652042360060514,
+      "grad_norm": 0.7270641209512483,
+      "learning_rate": 1.2306158707424402e-05,
+      "loss": 0.425,
+      "step": 319
+    },
+    {
+      "epoch": 0.9682299546142209,
+      "grad_norm": 0.6608607428518309,
+      "learning_rate": 1.2254663551538047e-05,
+      "loss": 0.3945,
+      "step": 320
+    },
+    {
+      "epoch": 0.9712556732223904,
+      "grad_norm": 0.6893548462856828,
+      "learning_rate": 1.2203105327865407e-05,
+      "loss": 0.4417,
+      "step": 321
+    },
+    {
+      "epoch": 0.9742813918305597,
+      "grad_norm": 0.682790511119452,
+      "learning_rate": 1.215148547860084e-05,
+      "loss": 0.4348,
+      "step": 322
+    },
+    {
+      "epoch": 0.9773071104387292,
+      "grad_norm": 0.673411672450261,
+      "learning_rate": 1.2099805447662485e-05,
+      "loss": 0.4036,
+      "step": 323
+    },
+    {
+      "epoch": 0.9803328290468987,
+      "grad_norm": 0.6770378250756861,
+      "learning_rate": 1.2048066680651908e-05,
+      "loss": 0.4338,
+      "step": 324
+    },
+    {
+      "epoch": 0.983358547655068,
+      "grad_norm": 0.6670604940572964,
+      "learning_rate": 1.1996270624813642e-05,
+      "loss": 0.4095,
+      "step": 325
+    },
+    {
+      "epoch": 0.9863842662632375,
+      "grad_norm": 0.6586760049553148,
+      "learning_rate": 1.194441872899471e-05,
+      "loss": 0.4135,
+      "step": 326
+    },
+    {
+      "epoch": 0.989409984871407,
+      "grad_norm": 0.6482055539349019,
+      "learning_rate": 1.1892512443604103e-05,
+      "loss": 0.4162,
+      "step": 327
+    },
+    {
+      "epoch": 0.9924357034795764,
+      "grad_norm": 0.6611420671664457,
+      "learning_rate": 1.1840553220572204e-05,
+      "loss": 0.427,
+      "step": 328
+    },
+    {
+      "epoch": 0.9954614220877458,
+      "grad_norm": 0.6970024457453018,
+      "learning_rate": 1.1788542513310178e-05,
+      "loss": 0.4579,
+      "step": 329
+    },
+    {
+      "epoch": 0.9984871406959153,
+      "grad_norm": 0.6583387250155186,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.4319,
+      "step": 330
+    },
+    {
+      "epoch": 1.0015128593040847,
+      "grad_norm": 0.7168019773228154,
+      "learning_rate": 1.1684372466900306e-05,
+      "loss": 0.3562,
+      "step": 331
+    },
+    {
+      "epoch": 1.0045385779122542,
+      "grad_norm": 0.7662804434604034,
+      "learning_rate": 1.1632216041612595e-05,
+      "loss": 0.3337,
+      "step": 332
+    },
+    {
+      "epoch": 1.0075642965204237,
+      "grad_norm": 0.6905128336379659,
+      "learning_rate": 1.15800139597335e-05,
+      "loss": 0.3108,
+      "step": 333
+    },
+    {
+      "epoch": 1.0105900151285931,
+      "grad_norm": 0.7211673315651629,
+      "learning_rate": 1.1527767681467472e-05,
+      "loss": 0.3422,
+      "step": 334
+    },
+    {
+      "epoch": 1.0136157337367624,
+      "grad_norm": 0.7488652863989967,
+      "learning_rate": 1.1475478668255223e-05,
+      "loss": 0.3229,
+      "step": 335
+    },
+    {
+      "epoch": 1.0166414523449319,
+      "grad_norm": 0.763446946050623,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3099,
+      "step": 336
+    },
+    {
+      "epoch": 1.0196671709531013,
+      "grad_norm": 0.8324769463165991,
+      "learning_rate": 1.1370778288690947e-05,
+      "loss": 0.3162,
+      "step": 337
+    },
+    {
+      "epoch": 1.0226928895612708,
+      "grad_norm": 0.7515394783698328,
+      "learning_rate": 1.1318369851033604e-05,
+      "loss": 0.311,
+      "step": 338
+    },
+    {
+      "epoch": 1.0257186081694403,
+      "grad_norm": 0.7106739298402036,
+      "learning_rate": 1.1265924535737494e-05,
+      "loss": 0.3127,
+      "step": 339
+    },
+    {
+      "epoch": 1.0287443267776097,
+      "grad_norm": 0.7389162586572955,
+      "learning_rate": 1.121344380981082e-05,
+      "loss": 0.3214,
+      "step": 340
+    },
+    {
+      "epoch": 1.0317700453857792,
+      "grad_norm": 0.7403040091843337,
+      "learning_rate": 1.1160929141252303e-05,
+      "loss": 0.3156,
+      "step": 341
+    },
+    {
+      "epoch": 1.0347957639939485,
+      "grad_norm": 0.7209058777027049,
+      "learning_rate": 1.1108381999010111e-05,
+      "loss": 0.3052,
+      "step": 342
+    },
+    {
+      "epoch": 1.037821482602118,
+      "grad_norm": 0.7286285716196962,
+      "learning_rate": 1.1055803852940772e-05,
+      "loss": 0.299,
+      "step": 343
+    },
+    {
+      "epoch": 1.0408472012102874,
+      "grad_norm": 0.7416253546807615,
+      "learning_rate": 1.1003196173768051e-05,
+      "loss": 0.2963,
+      "step": 344
+    },
+    {
+      "epoch": 1.0438729198184569,
+      "grad_norm": 0.8050504301380201,
+      "learning_rate": 1.0950560433041825e-05,
+      "loss": 0.3162,
+      "step": 345
+    },
+    {
+      "epoch": 1.0468986384266263,
+      "grad_norm": 0.7515953381957162,
+      "learning_rate": 1.0897898103096917e-05,
+      "loss": 0.3081,
+      "step": 346
+    },
+    {
+      "epoch": 1.0499243570347958,
+      "grad_norm": 0.7259952697107869,
+      "learning_rate": 1.0845210657011893e-05,
+      "loss": 0.3077,
+      "step": 347
+    },
+    {
+      "epoch": 1.0529500756429653,
+      "grad_norm": 0.8105680711734808,
+      "learning_rate": 1.0792499568567885e-05,
+      "loss": 0.3135,
+      "step": 348
+    },
+    {
+      "epoch": 1.0559757942511347,
+      "grad_norm": 0.7940723097397853,
+      "learning_rate": 1.0739766312207344e-05,
+      "loss": 0.3071,
+      "step": 349
+    },
+    {
+      "epoch": 1.059001512859304,
+      "grad_norm": 0.7634182429028121,
+      "learning_rate": 1.068701236299281e-05,
+      "loss": 0.3107,
+      "step": 350
+    },
+    {
+      "epoch": 1.0620272314674735,
+      "grad_norm": 0.7626382643821981,
+      "learning_rate": 1.0634239196565646e-05,
+      "loss": 0.3068,
+      "step": 351
+    },
+    {
+      "epoch": 1.065052950075643,
+      "grad_norm": 0.7883515633392242,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.3205,
+      "step": 352
+    },
+    {
+      "epoch": 1.0680786686838124,
+      "grad_norm": 0.704497421703472,
+      "learning_rate": 1.0528641117285315e-05,
+      "loss": 0.3185,
+      "step": 353
+    },
+    {
+      "epoch": 1.0711043872919819,
+      "grad_norm": 0.7162111361600043,
+      "learning_rate": 1.0475819158237426e-05,
+      "loss": 0.2918,
+      "step": 354
+    },
+    {
+      "epoch": 1.0741301059001513,
+      "grad_norm": 0.7527031367097157,
+      "learning_rate": 1.0422983889504831e-05,
+      "loss": 0.3313,
+      "step": 355
+    },
+    {
+      "epoch": 1.0771558245083208,
+      "grad_norm": 0.6936015930843374,
+      "learning_rate": 1.0370136789003582e-05,
+      "loss": 0.3135,
+      "step": 356
+    },
+    {
+      "epoch": 1.08018154311649,
+      "grad_norm": 0.6556445787843884,
+      "learning_rate": 1.031727933498068e-05,
+      "loss": 0.2893,
+      "step": 357
+    },
+    {
+      "epoch": 1.0832072617246595,
+      "grad_norm": 0.7018955546982815,
+      "learning_rate": 1.0264413005972736e-05,
+      "loss": 0.3095,
+      "step": 358
+    },
+    {
+      "epoch": 1.086232980332829,
+      "grad_norm": 0.7401844553675462,
+      "learning_rate": 1.0211539280764617e-05,
+      "loss": 0.3287,
+      "step": 359
+    },
+    {
+      "epoch": 1.0892586989409985,
+      "grad_norm": 0.7518351125230497,
+      "learning_rate": 1.015865963834808e-05,
+      "loss": 0.3043,
+      "step": 360
+    },
+    {
+      "epoch": 1.092284417549168,
+      "grad_norm": 0.7218851930033655,
+      "learning_rate": 1.0105775557880398e-05,
+      "loss": 0.3148,
+      "step": 361
+    },
+    {
+      "epoch": 1.0953101361573374,
+      "grad_norm": 0.6938071583513528,
+      "learning_rate": 1.0052888518642978e-05,
+      "loss": 0.3326,
+      "step": 362
+    },
+    {
+      "epoch": 1.0983358547655069,
+      "grad_norm": 0.6831258945539711,
+      "learning_rate": 1e-05,
+      "loss": 0.309,
+      "step": 363
+    },
+    {
+      "epoch": 1.1013615733736764,
+      "grad_norm": 0.7063476273426411,
+      "learning_rate": 9.947111481357023e-06,
+      "loss": 0.3132,
+      "step": 364
+    },
+    {
+      "epoch": 1.1043872919818456,
+      "grad_norm": 0.7332385135057867,
+      "learning_rate": 9.894224442119606e-06,
+      "loss": 0.3009,
+      "step": 365
+    },
+    {
+      "epoch": 1.107413010590015,
+      "grad_norm": 0.674238594500562,
+      "learning_rate": 9.841340361651921e-06,
+      "loss": 0.2965,
+      "step": 366
+    },
+    {
+      "epoch": 1.1104387291981845,
+      "grad_norm": 0.711649910336382,
+      "learning_rate": 9.788460719235386e-06,
+      "loss": 0.3118,
+      "step": 367
+    },
+    {
+      "epoch": 1.113464447806354,
+      "grad_norm": 0.7573253264443083,
+      "learning_rate": 9.735586994027267e-06,
+      "loss": 0.3112,
+      "step": 368
+    },
+    {
+      "epoch": 1.1164901664145235,
+      "grad_norm": 0.717620552908331,
+      "learning_rate": 9.682720665019325e-06,
+      "loss": 0.3171,
+      "step": 369
+    },
+    {
+      "epoch": 1.119515885022693,
+      "grad_norm": 0.6723501898055355,
+      "learning_rate": 9.62986321099642e-06,
+      "loss": 0.2882,
+      "step": 370
+    },
+    {
+      "epoch": 1.1225416036308624,
+      "grad_norm": 0.7525618951489431,
+      "learning_rate": 9.57701611049517e-06,
+      "loss": 0.3127,
+      "step": 371
+    },
+    {
+      "epoch": 1.1255673222390317,
+      "grad_norm": 0.7284792348274622,
+      "learning_rate": 9.524180841762577e-06,
+      "loss": 0.3165,
+      "step": 372
+    },
+    {
+      "epoch": 1.1285930408472011,
+      "grad_norm": 0.7188640108257898,
+      "learning_rate": 9.471358882714687e-06,
+      "loss": 0.3032,
+      "step": 373
+    },
+    {
+      "epoch": 1.1316187594553706,
+      "grad_norm": 0.732541521963472,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.3138,
+      "step": 374
+    },
+    {
+      "epoch": 1.13464447806354,
+      "grad_norm": 0.7289731190690716,
+      "learning_rate": 9.365760803434356e-06,
+      "loss": 0.3172,
+      "step": 375
+    },
+    {
+      "epoch": 1.1376701966717095,
+      "grad_norm": 0.7347795092318653,
+      "learning_rate": 9.312987637007191e-06,
+      "loss": 0.3197,
+      "step": 376
+    },
+    {
+      "epoch": 1.140695915279879,
+      "grad_norm": 0.6885798786216215,
+      "learning_rate": 9.260233687792657e-06,
+      "loss": 0.3035,
+      "step": 377
+    },
+    {
+      "epoch": 1.1437216338880485,
+      "grad_norm": 0.7073134196208323,
+      "learning_rate": 9.207500431432115e-06,
+      "loss": 0.3196,
+      "step": 378
+    },
+    {
+      "epoch": 1.146747352496218,
+      "grad_norm": 0.7060165283699794,
+      "learning_rate": 9.154789342988108e-06,
+      "loss": 0.3102,
+      "step": 379
+    },
+    {
+      "epoch": 1.1497730711043872,
+      "grad_norm": 0.7300034524326682,
+      "learning_rate": 9.102101896903084e-06,
+      "loss": 0.3626,
+      "step": 380
+    },
+    {
+      "epoch": 1.1527987897125567,
+      "grad_norm": 0.6892576533248072,
+      "learning_rate": 9.049439566958176e-06,
+      "loss": 0.3148,
+      "step": 381
+    },
+    {
+      "epoch": 1.1558245083207261,
+      "grad_norm": 0.6869429930265033,
+      "learning_rate": 8.99680382623195e-06,
+      "loss": 0.3084,
+      "step": 382
+    },
+    {
+      "epoch": 1.1588502269288956,
+      "grad_norm": 0.7182008430337521,
+      "learning_rate": 8.944196147059233e-06,
+      "loss": 0.2886,
+      "step": 383
+    },
+    {
+      "epoch": 1.161875945537065,
+      "grad_norm": 0.730237686737673,
+      "learning_rate": 8.89161800098989e-06,
+      "loss": 0.3092,
+      "step": 384
+    },
+    {
+      "epoch": 1.1649016641452345,
+      "grad_norm": 0.7752816408135582,
+      "learning_rate": 8.839070858747697e-06,
+      "loss": 0.3161,
+      "step": 385
+    },
+    {
+      "epoch": 1.167927382753404,
+      "grad_norm": 0.7262319367217432,
+      "learning_rate": 8.786556190189183e-06,
+      "loss": 0.3011,
+      "step": 386
+    },
+    {
+      "epoch": 1.1709531013615733,
+      "grad_norm": 0.6875798864729167,
+      "learning_rate": 8.734075464262507e-06,
+      "loss": 0.3082,
+      "step": 387
+    },
+    {
+      "epoch": 1.1739788199697427,
+      "grad_norm": 0.7325411516180098,
+      "learning_rate": 8.681630148966397e-06,
+      "loss": 0.304,
+      "step": 388
+    },
+    {
+      "epoch": 1.1770045385779122,
+      "grad_norm": 0.6772330240109783,
+      "learning_rate": 8.629221711309056e-06,
+      "loss": 0.3157,
+      "step": 389
+    },
+    {
+      "epoch": 1.1800302571860817,
+      "grad_norm": 0.7308650516464686,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2995,
+      "step": 390
+    },
+    {
+      "epoch": 1.1830559757942511,
+      "grad_norm": 0.6708378364498481,
+      "learning_rate": 8.52452133174478e-06,
+      "loss": 0.2876,
+      "step": 391
+    },
+    {
+      "epoch": 1.1860816944024206,
+      "grad_norm": 0.6812060830260644,
+      "learning_rate": 8.472232318532531e-06,
+      "loss": 0.3038,
+      "step": 392
+    },
+    {
+      "epoch": 1.18910741301059,
+      "grad_norm": 0.6769369678923325,
+      "learning_rate": 8.419986040266502e-06,
+      "loss": 0.3008,
+      "step": 393
+    },
+    {
+      "epoch": 1.1921331316187596,
+      "grad_norm": 0.7285040905447597,
+      "learning_rate": 8.367783958387407e-06,
+      "loss": 0.3061,
+      "step": 394
+    },
+    {
+      "epoch": 1.1951588502269288,
+      "grad_norm": 0.7064458783443105,
+      "learning_rate": 8.315627533099697e-06,
+      "loss": 0.3036,
+      "step": 395
+    },
+    {
+      "epoch": 1.1981845688350983,
+      "grad_norm": 0.7130035283971766,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.3119,
+      "step": 396
+    },
+    {
+      "epoch": 1.2012102874432677,
+      "grad_norm": 0.6915866867692465,
+      "learning_rate": 8.211457486689829e-06,
+      "loss": 0.2964,
+      "step": 397
+    },
+    {
+      "epoch": 1.2042360060514372,
+      "grad_norm": 0.707387990967338,
+      "learning_rate": 8.159446779427798e-06,
+      "loss": 0.3233,
+      "step": 398
+    },
+    {
+      "epoch": 1.2072617246596067,
+      "grad_norm": 0.7470959448248157,
+      "learning_rate": 8.107487556395902e-06,
+      "loss": 0.2899,
+      "step": 399
+    },
+    {
+      "epoch": 1.2102874432677762,
+      "grad_norm": 0.7323034204801262,
+      "learning_rate": 8.055581271005292e-06,
+      "loss": 0.3152,
+      "step": 400
+    },
+    {
+      "epoch": 1.2133131618759456,
+      "grad_norm": 0.7573079815444309,
+      "learning_rate": 8.00372937518636e-06,
+      "loss": 0.3094,
+      "step": 401
+    },
+    {
+      "epoch": 1.2163388804841149,
+      "grad_norm": 0.6740289232186525,
+      "learning_rate": 7.951933319348095e-06,
+      "loss": 0.3049,
+      "step": 402
+    },
+    {
+      "epoch": 1.2193645990922843,
+      "grad_norm": 0.6667973403523348,
+      "learning_rate": 7.900194552337516e-06,
+      "loss": 0.2995,
+      "step": 403
+    },
+    {
+      "epoch": 1.2223903177004538,
+      "grad_norm": 0.7095441875687234,
+      "learning_rate": 7.848514521399167e-06,
+      "loss": 0.3193,
+      "step": 404
+    },
+    {
+      "epoch": 1.2254160363086233,
+      "grad_norm": 0.7013875324479552,
+      "learning_rate": 7.796894672134594e-06,
+      "loss": 0.3113,
+      "step": 405
+    },
+    {
+      "epoch": 1.2284417549167927,
+      "grad_norm": 0.6713844588443191,
+      "learning_rate": 7.745336448461958e-06,
+      "loss": 0.2965,
+      "step": 406
+    },
+    {
+      "epoch": 1.2314674735249622,
+      "grad_norm": 0.6413960014638571,
+      "learning_rate": 7.6938412925756e-06,
+      "loss": 0.2916,
+      "step": 407
+    },
+    {
+      "epoch": 1.2344931921331317,
+      "grad_norm": 0.7143028648656433,
+      "learning_rate": 7.642410644905726e-06,
+      "loss": 0.3051,
+      "step": 408
+    },
+    {
+      "epoch": 1.2375189107413012,
+      "grad_norm": 0.6792905205151105,
+      "learning_rate": 7.591045944078119e-06,
+      "loss": 0.2888,
+      "step": 409
+    },
+    {
+      "epoch": 1.2405446293494704,
+      "grad_norm": 0.7171505140672608,
+      "learning_rate": 7.539748626873866e-06,
+      "loss": 0.307,
+      "step": 410
+    },
+    {
+      "epoch": 1.2435703479576399,
+      "grad_norm": 0.745668079331383,
+      "learning_rate": 7.488520128189209e-06,
+      "loss": 0.3238,
+      "step": 411
+    },
+    {
+      "epoch": 1.2465960665658093,
+      "grad_norm": 0.6786713074342549,
+      "learning_rate": 7.4373618809953755e-06,
+      "loss": 0.2992,
+      "step": 412
+    },
+    {
+      "epoch": 1.2496217851739788,
+      "grad_norm": 0.7635813715230593,
+      "learning_rate": 7.386275316298513e-06,
+      "loss": 0.314,
+      "step": 413
+    },
+    {
+      "epoch": 1.2526475037821483,
+      "grad_norm": 0.6897637931373736,
+      "learning_rate": 7.335261863099652e-06,
+      "loss": 0.2891,
+      "step": 414
+    },
+    {
+      "epoch": 1.2556732223903178,
+      "grad_norm": 0.6953752819170786,
+      "learning_rate": 7.2843229483547405e-06,
+      "loss": 0.3026,
+      "step": 415
+    },
+    {
+      "epoch": 1.258698940998487,
+      "grad_norm": 0.6837631858746336,
+      "learning_rate": 7.233459996934731e-06,
+      "loss": 0.316,
+      "step": 416
+    },
+    {
+      "epoch": 1.2617246596066565,
+      "grad_norm": 0.6758963373408814,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3094,
+      "step": 417
+    },
+    {
+      "epoch": 1.264750378214826,
+      "grad_norm": 0.7211745146327605,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.3054,
+      "step": 418
+    },
+    {
+      "epoch": 1.2677760968229954,
+      "grad_norm": 0.7444336950031109,
+      "learning_rate": 7.081341139221955e-06,
+      "loss": 0.3149,
+      "step": 419
+    },
+    {
+      "epoch": 1.2708018154311649,
+      "grad_norm": 0.6898796584855196,
+      "learning_rate": 7.0307962467172555e-06,
+      "loss": 0.3096,
+      "step": 420
+    },
+    {
+      "epoch": 1.2738275340393344,
+      "grad_norm": 0.7455079623940967,
+      "learning_rate": 6.9803344092242855e-06,
+      "loss": 0.3035,
+      "step": 421
+    },
+    {
+      "epoch": 1.2768532526475038,
+      "grad_norm": 0.7229002646094703,
+      "learning_rate": 6.929957038269123e-06,
+      "loss": 0.3166,
+      "step": 422
+    },
+    {
+      "epoch": 1.2798789712556733,
+      "grad_norm": 0.6853928365986821,
+      "learning_rate": 6.87966554301513e-06,
+      "loss": 0.3124,
+      "step": 423
+    },
+    {
+      "epoch": 1.2829046898638428,
+      "grad_norm": 0.6620207458493395,
+      "learning_rate": 6.8294613302235325e-06,
+      "loss": 0.282,
+      "step": 424
+    },
+    {
+      "epoch": 1.2859304084720122,
+      "grad_norm": 0.6882226045008781,
+      "learning_rate": 6.779345804214088e-06,
+      "loss": 0.3097,
+      "step": 425
+    },
+    {
+      "epoch": 1.2889561270801815,
+      "grad_norm": 0.6886426093714056,
+      "learning_rate": 6.729320366825785e-06,
+      "loss": 0.3079,
+      "step": 426
+    },
+    {
+      "epoch": 1.291981845688351,
+      "grad_norm": 0.7050980415736454,
+      "learning_rate": 6.679386417377649e-06,
+      "loss": 0.2935,
+      "step": 427
+    },
+    {
+      "epoch": 1.2950075642965204,
+      "grad_norm": 0.7280983968358292,
+      "learning_rate": 6.629545352629583e-06,
+      "loss": 0.3199,
+      "step": 428
+    },
+    {
+      "epoch": 1.29803328290469,
+      "grad_norm": 0.6642605445323858,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.2887,
+      "step": 429
+    },
+    {
+      "epoch": 1.3010590015128594,
+      "grad_norm": 0.7149416136096906,
+      "learning_rate": 6.530147451243377e-06,
+      "loss": 0.3184,
+      "step": 430
+    },
+    {
+      "epoch": 1.3040847201210286,
+      "grad_norm": 0.7309184023681224,
+      "learning_rate": 6.480593394978208e-06,
+      "loss": 0.3303,
+      "step": 431
+    },
+    {
+      "epoch": 1.307110438729198,
+      "grad_norm": 0.7448704861010347,
+      "learning_rate": 6.431137784081283e-06,
+      "loss": 0.3049,
+      "step": 432
+    },
+    {
+      "epoch": 1.3101361573373675,
+      "grad_norm": 0.7136023803474852,
+      "learning_rate": 6.381782001932352e-06,
+      "loss": 0.307,
+      "step": 433
+    },
+    {
+      "epoch": 1.313161875945537,
+      "grad_norm": 0.7269127648930268,
+      "learning_rate": 6.33252742911874e-06,
+      "loss": 0.3119,
+      "step": 434
+    },
+    {
+      "epoch": 1.3161875945537065,
+      "grad_norm": 0.7506944802071415,
+      "learning_rate": 6.283375443396726e-06,
+      "loss": 0.3249,
+      "step": 435
+    },
+    {
+      "epoch": 1.319213313161876,
+      "grad_norm": 0.7134971842136345,
+      "learning_rate": 6.234327419653013e-06,
+      "loss": 0.3141,
+      "step": 436
+    },
+    {
+      "epoch": 1.3222390317700454,
+      "grad_norm": 0.7133633118564093,
+      "learning_rate": 6.185384729866264e-06,
+      "loss": 0.3184,
+      "step": 437
+    },
+    {
+      "epoch": 1.325264750378215,
+      "grad_norm": 0.7514788307954579,
+      "learning_rate": 6.136548743068713e-06,
+      "loss": 0.3236,
+      "step": 438
+    },
+    {
+      "epoch": 1.3282904689863844,
+      "grad_norm": 0.7322103762270158,
+      "learning_rate": 6.087820825307904e-06,
+      "loss": 0.3055,
+      "step": 439
+    },
+    {
+      "epoch": 1.3313161875945538,
+      "grad_norm": 0.6725687532550337,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.2988,
+      "step": 440
+    },
+    {
+      "epoch": 1.334341906202723,
+      "grad_norm": 0.7216431481814384,
+      "learning_rate": 5.990694645933866e-06,
+      "loss": 0.2957,
+      "step": 441
+    },
+    {
+      "epoch": 1.3373676248108926,
+      "grad_norm": 0.7341704283656727,
+      "learning_rate": 5.9422991011486635e-06,
+      "loss": 0.2958,
+      "step": 442
+    },
+    {
+      "epoch": 1.340393343419062,
+      "grad_norm": 0.6739480377713454,
+      "learning_rate": 5.894017058980249e-06,
+      "loss": 0.293,
+      "step": 443
+    },
+    {
+      "epoch": 1.3434190620272315,
+      "grad_norm": 0.7194950497651736,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3069,
+      "step": 444
+    },
+    {
+      "epoch": 1.346444780635401,
+      "grad_norm": 0.687277145559174,
+      "learning_rate": 5.797798881491138e-06,
+      "loss": 0.305,
+      "step": 445
+    },
+    {
+      "epoch": 1.3494704992435702,
+      "grad_norm": 0.7512903555776772,
+      "learning_rate": 5.749865437599703e-06,
+      "loss": 0.3123,
+      "step": 446
+    },
+    {
+      "epoch": 1.3524962178517397,
+      "grad_norm": 0.699682874033537,
+      "learning_rate": 5.702050879108284e-06,
+      "loss": 0.3022,
+      "step": 447
+    },
+    {
+      "epoch": 1.3555219364599091,
+      "grad_norm": 0.699370503592375,
+      "learning_rate": 5.654356543492883e-06,
+      "loss": 0.3065,
+      "step": 448
+    },
+    {
+      "epoch": 1.3585476550680786,
+      "grad_norm": 0.7581754930383585,
+      "learning_rate": 5.606783764866576e-06,
+      "loss": 0.2879,
+      "step": 449
+    },
+    {
+      "epoch": 1.361573373676248,
+      "grad_norm": 0.6796477317693127,
+      "learning_rate": 5.559333873942259e-06,
+      "loss": 0.2854,
+      "step": 450
+    },
+    {
+      "epoch": 1.3645990922844176,
+      "grad_norm": 0.6780800765077403,
+      "learning_rate": 5.512008197995379e-06,
+      "loss": 0.3078,
+      "step": 451
+    },
+    {
+      "epoch": 1.367624810892587,
+      "grad_norm": 0.6945926711618069,
+      "learning_rate": 5.464808060826825e-06,
+      "loss": 0.2986,
+      "step": 452
+    },
+    {
+      "epoch": 1.3706505295007565,
+      "grad_norm": 0.6784539879994838,
+      "learning_rate": 5.417734782725896e-06,
+      "loss": 0.2924,
+      "step": 453
+    },
+    {
+      "epoch": 1.373676248108926,
+      "grad_norm": 0.6956680421202921,
+      "learning_rate": 5.370789680433376e-06,
+      "loss": 0.2848,
+      "step": 454
+    },
+    {
+      "epoch": 1.3767019667170954,
+      "grad_norm": 0.7362656238170324,
+      "learning_rate": 5.323974067104687e-06,
+      "loss": 0.2957,
+      "step": 455
+    },
+    {
+      "epoch": 1.3797276853252647,
+      "grad_norm": 0.6820386956286771,
+      "learning_rate": 5.277289252273175e-06,
+      "loss": 0.3027,
+      "step": 456
+    },
+    {
+      "epoch": 1.3827534039334342,
+      "grad_norm": 0.7157526518091163,
+      "learning_rate": 5.230736541813463e-06,
+      "loss": 0.3002,
+      "step": 457
+    },
+    {
+      "epoch": 1.3857791225416036,
+      "grad_norm": 0.7219280237013429,
+      "learning_rate": 5.184317237904939e-06,
+      "loss": 0.2999,
+      "step": 458
+    },
+    {
+      "epoch": 1.388804841149773,
+      "grad_norm": 0.7228462823822086,
+      "learning_rate": 5.138032638995315e-06,
+      "loss": 0.3012,
+      "step": 459
+    },
+    {
+      "epoch": 1.3918305597579426,
+      "grad_norm": 0.708229956602603,
+      "learning_rate": 5.091884039764321e-06,
+      "loss": 0.2913,
+      "step": 460
+    },
+    {
+      "epoch": 1.394856278366112,
+      "grad_norm": 0.6536882605641012,
+      "learning_rate": 5.045872731087479e-06,
+      "loss": 0.2912,
+      "step": 461
+    },
+    {
+      "epoch": 1.3978819969742813,
+      "grad_norm": 0.7027038941889816,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.3126,
+      "step": 462
+    },
+    {
+      "epoch": 1.4009077155824508,
+      "grad_norm": 0.7194364766519474,
+      "learning_rate": 4.954267129660789e-06,
+      "loss": 0.2991,
+      "step": 463
+    },
+    {
+      "epoch": 1.4039334341906202,
+      "grad_norm": 0.717452372337296,
+      "learning_rate": 4.908675399316534e-06,
+      "loss": 0.3167,
+      "step": 464
+    },
+    {
+      "epoch": 1.4069591527987897,
+      "grad_norm": 0.776437178061153,
+      "learning_rate": 4.863226084265939e-06,
+      "loss": 0.3051,
+      "step": 465
+    },
+    {
+      "epoch": 1.4099848714069592,
+      "grad_norm": 0.6730332019359635,
+      "learning_rate": 4.817920455824045e-06,
+      "loss": 0.3025,
+      "step": 466
+    },
+    {
+      "epoch": 1.4130105900151286,
+      "grad_norm": 0.6989533397063674,
+      "learning_rate": 4.772759781286679e-06,
+      "loss": 0.3084,
+      "step": 467
+    },
+    {
+      "epoch": 1.416036308623298,
+      "grad_norm": 0.6672240278517017,
+      "learning_rate": 4.727745323894976e-06,
+      "loss": 0.2858,
+      "step": 468
+    },
+    {
+      "epoch": 1.4190620272314676,
+      "grad_norm": 0.6413415101632947,
+      "learning_rate": 4.682878342800087e-06,
+      "loss": 0.2878,
+      "step": 469
+    },
+    {
+      "epoch": 1.422087745839637,
+      "grad_norm": 0.7245343570795441,
+      "learning_rate": 4.638160093027908e-06,
+      "loss": 0.3066,
+      "step": 470
+    },
+    {
+      "epoch": 1.4251134644478063,
+      "grad_norm": 0.6653202320821319,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.3017,
+      "step": 471
+    },
+    {
+      "epoch": 1.4281391830559758,
+      "grad_norm": 0.6636515497997716,
+      "learning_rate": 4.549174786718684e-06,
+      "loss": 0.2953,
+      "step": 472
+    },
+    {
+      "epoch": 1.4311649016641452,
+      "grad_norm": 0.6614799829829542,
+      "learning_rate": 4.504910219291941e-06,
+      "loss": 0.2942,
+      "step": 473
+    },
+    {
+      "epoch": 1.4341906202723147,
+      "grad_norm": 0.7126882961288521,
+      "learning_rate": 4.460799361338898e-06,
+      "loss": 0.3081,
+      "step": 474
+    },
+    {
+      "epoch": 1.4372163388804842,
+      "grad_norm": 0.7194264868045459,
+      "learning_rate": 4.416843446735077e-06,
+      "loss": 0.2965,
+      "step": 475
+    },
+    {
+      "epoch": 1.4402420574886536,
+      "grad_norm": 0.6810520910741196,
+      "learning_rate": 4.373043705021899e-06,
+      "loss": 0.2909,
+      "step": 476
+    },
+    {
+      "epoch": 1.4432677760968229,
+      "grad_norm": 0.6476127656397964,
+      "learning_rate": 4.3294013613722944e-06,
+      "loss": 0.2878,
+      "step": 477
+    },
+    {
+      "epoch": 1.4462934947049924,
+      "grad_norm": 0.7223531163934885,
+      "learning_rate": 4.2859176365564294e-06,
+      "loss": 0.302,
+      "step": 478
+    },
+    {
+      "epoch": 1.4493192133131618,
+      "grad_norm": 0.657268668890294,
+      "learning_rate": 4.2425937469075626e-06,
+      "loss": 0.2959,
+      "step": 479
+    },
+    {
+      "epoch": 1.4523449319213313,
+      "grad_norm": 0.7098452358713787,
+      "learning_rate": 4.19943090428802e-06,
+      "loss": 0.3129,
+      "step": 480
+    },
+    {
+      "epoch": 1.4553706505295008,
+      "grad_norm": 0.7207250178403358,
+      "learning_rate": 4.1564303160552935e-06,
+      "loss": 0.3079,
+      "step": 481
+    },
+    {
+      "epoch": 1.4583963691376702,
+      "grad_norm": 0.7134077336507284,
+      "learning_rate": 4.113593185028273e-06,
+      "loss": 0.3045,
+      "step": 482
+    },
+    {
+      "epoch": 1.4614220877458397,
+      "grad_norm": 0.6798541474551784,
+      "learning_rate": 4.070920709453597e-06,
+      "loss": 0.2994,
+      "step": 483
+    },
+    {
+      "epoch": 1.4644478063540092,
+      "grad_norm": 0.6666042903056767,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.3025,
+      "step": 484
+    },
+    {
+      "epoch": 1.4674735249621786,
+      "grad_norm": 0.667440872368227,
+      "learning_rate": 3.986074494585619e-06,
+      "loss": 0.2984,
+      "step": 485
+    },
+    {
+      "epoch": 1.470499243570348,
+      "grad_norm": 0.7052722259276454,
+      "learning_rate": 3.943903128623336e-06,
+      "loss": 0.3049,
+      "step": 486
+    },
+    {
+      "epoch": 1.4735249621785174,
+      "grad_norm": 0.7108983396528757,
+      "learning_rate": 3.9019011647090465e-06,
+      "loss": 0.3046,
+      "step": 487
+    },
+    {
+      "epoch": 1.4765506807866868,
+      "grad_norm": 0.7214873885253321,
+      "learning_rate": 3.860069777727983e-06,
+      "loss": 0.2997,
+      "step": 488
+    },
+    {
+      "epoch": 1.4795763993948563,
+      "grad_norm": 0.6642868729573265,
+      "learning_rate": 3.818410137793947e-06,
+      "loss": 0.2878,
+      "step": 489
+    },
+    {
+      "epoch": 1.4826021180030258,
+      "grad_norm": 0.6686843597233881,
+      "learning_rate": 3.7769234102166365e-06,
+      "loss": 0.302,
+      "step": 490
+    },
+    {
+      "epoch": 1.4856278366111952,
+      "grad_norm": 0.679585499293679,
+      "learning_rate": 3.735610755468988e-06,
+      "loss": 0.3009,
+      "step": 491
+    },
+    {
+      "epoch": 1.4886535552193645,
+      "grad_norm": 0.7210810089762842,
+      "learning_rate": 3.6944733291547784e-06,
+      "loss": 0.3048,
+      "step": 492
+    },
+    {
+      "epoch": 1.491679273827534,
+      "grad_norm": 0.686206014696815,
+      "learning_rate": 3.653512281976238e-06,
+      "loss": 0.2889,
+      "step": 493
+    },
+    {
+      "epoch": 1.4947049924357034,
+      "grad_norm": 0.7105964690398467,
+      "learning_rate": 3.612728759701919e-06,
+      "loss": 0.3017,
+      "step": 494
+    },
+    {
+      "epoch": 1.497730711043873,
+      "grad_norm": 0.7139226331111,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 0.3046,
+      "step": 495
+    },
+    {
+      "epoch": 1.5007564296520424,
+      "grad_norm": 0.678477299277824,
+      "learning_rate": 3.5316988480794255e-06,
+      "loss": 0.2887,
+      "step": 496
+    },
+    {
+      "epoch": 1.5037821482602118,
+      "grad_norm": 0.7259703631854253,
+      "learning_rate": 3.4914547253120655e-06,
+      "loss": 0.3051,
+      "step": 497
+    },
+    {
+      "epoch": 1.5068078668683813,
+      "grad_norm": 0.7416977334466437,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2965,
+      "step": 498
+    },
+    {
+      "epoch": 1.5098335854765508,
+      "grad_norm": 0.7133172735787201,
+      "learning_rate": 3.4115137744067516e-06,
+      "loss": 0.3025,
+      "step": 499
+    },
+    {
+      "epoch": 1.5128593040847202,
+      "grad_norm": 0.6851243116576939,
+      "learning_rate": 3.37181918238904e-06,
+      "loss": 0.303,
+      "step": 500
+    },
+    {
+      "epoch": 1.5158850226928897,
+      "grad_norm": 0.7023546247777874,
+      "learning_rate": 3.3323099948370853e-06,
+      "loss": 0.3134,
+      "step": 501
+    },
+    {
+      "epoch": 1.518910741301059,
+      "grad_norm": 0.7231771893466172,
+      "learning_rate": 3.292987316907792e-06,
+      "loss": 0.3049,
+      "step": 502
+    },
+    {
+      "epoch": 1.5219364599092284,
+      "grad_norm": 0.6755258382222047,
+      "learning_rate": 3.253852248540994e-06,
+      "loss": 0.2751,
+      "step": 503
+    },
+    {
+      "epoch": 1.524962178517398,
+      "grad_norm": 0.7155729504825645,
+      "learning_rate": 3.2149058844286796e-06,
+      "loss": 0.3112,
+      "step": 504
+    },
+    {
+      "epoch": 1.5279878971255674,
+      "grad_norm": 0.6692383318107048,
+      "learning_rate": 3.1761493139843734e-06,
+      "loss": 0.2927,
+      "step": 505
+    },
+    {
+      "epoch": 1.5310136157337366,
+      "grad_norm": 0.6598621061814046,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 0.28,
+      "step": 506
+    },
+    {
+      "epoch": 1.534039334341906,
+      "grad_norm": 0.6658364388415778,
+      "learning_rate": 3.099209885178882e-06,
+      "loss": 0.3009,
+      "step": 507
+    },
+    {
+      "epoch": 1.5370650529500756,
+      "grad_norm": 0.6768493302353455,
+      "learning_rate": 3.0610291789789094e-06,
+      "loss": 0.3048,
+      "step": 508
+    },
+    {
+      "epoch": 1.540090771558245,
+      "grad_norm": 0.687277559695711,
+      "learning_rate": 3.023042570709185e-06,
+      "loss": 0.2961,
+      "step": 509
+    },
+    {
+      "epoch": 1.5431164901664145,
+      "grad_norm": 0.7004001597195383,
+      "learning_rate": 2.9852511229367862e-06,
+      "loss": 0.2934,
+      "step": 510
+    },
+    {
+      "epoch": 1.546142208774584,
+      "grad_norm": 0.7354478813355237,
+      "learning_rate": 2.9476558927697605e-06,
+      "loss": 0.3097,
+      "step": 511
+    },
+    {
+      "epoch": 1.5491679273827534,
+      "grad_norm": 0.6607228219986576,
+      "learning_rate": 2.9102579318274994e-06,
+      "loss": 0.2985,
+      "step": 512
+    },
+    {
+      "epoch": 1.552193645990923,
+      "grad_norm": 0.7099534876004286,
+      "learning_rate": 2.8730582862113743e-06,
+      "loss": 0.3022,
+      "step": 513
+    },
+    {
+      "epoch": 1.5552193645990924,
+      "grad_norm": 0.69178683447553,
+      "learning_rate": 2.8360579964754277e-06,
+      "loss": 0.286,
+      "step": 514
+    },
+    {
+      "epoch": 1.5582450832072618,
+      "grad_norm": 0.76072103323838,
+      "learning_rate": 2.7992580975973136e-06,
+      "loss": 0.3178,
+      "step": 515
+    },
+    {
+      "epoch": 1.5612708018154313,
+      "grad_norm": 0.7088596188311644,
+      "learning_rate": 2.7626596189492983e-06,
+      "loss": 0.3073,
+      "step": 516
+    },
+    {
+      "epoch": 1.5642965204236006,
+      "grad_norm": 0.7279166827979408,
+      "learning_rate": 2.726263584269513e-06,
+      "loss": 0.3089,
+      "step": 517
+    },
+    {
+      "epoch": 1.56732223903177,
+      "grad_norm": 0.6724705575368527,
+      "learning_rate": 2.690071011633284e-06,
+      "loss": 0.2764,
+      "step": 518
+    },
+    {
+      "epoch": 1.5703479576399395,
+      "grad_norm": 0.7038374349290187,
+      "learning_rate": 2.6540829134246683e-06,
+      "loss": 0.2931,
+      "step": 519
+    },
+    {
+      "epoch": 1.573373676248109,
+      "grad_norm": 0.7105589359089628,
+      "learning_rate": 2.618300296308135e-06,
+      "loss": 0.2971,
+      "step": 520
+    },
+    {
+      "epoch": 1.5763993948562782,
+      "grad_norm": 0.7143570383757667,
+      "learning_rate": 2.582724161200405e-06,
+      "loss": 0.3063,
+      "step": 521
+    },
+    {
+      "epoch": 1.5794251134644477,
+      "grad_norm": 0.6805400380222137,
+      "learning_rate": 2.5473555032424534e-06,
+      "loss": 0.3007,
+      "step": 522
+    },
+    {
+      "epoch": 1.5824508320726172,
+      "grad_norm": 0.6889102771293127,
+      "learning_rate": 2.5121953117716744e-06,
+      "loss": 0.2955,
+      "step": 523
+    },
+    {
+      "epoch": 1.5854765506807866,
+      "grad_norm": 0.7115107193824134,
+      "learning_rate": 2.477244570294206e-06,
+      "loss": 0.2976,
+      "step": 524
+    },
+    {
+      "epoch": 1.588502269288956,
+      "grad_norm": 0.675253717312688,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.3029,
+      "step": 525
+    },
+    {
+      "epoch": 1.5915279878971256,
+      "grad_norm": 0.6688545405458325,
+      "learning_rate": 2.4079753420225694e-06,
+      "loss": 0.2912,
+      "step": 526
+    },
+    {
+      "epoch": 1.594553706505295,
+      "grad_norm": 0.6384115804314029,
+      "learning_rate": 2.3736587928376197e-06,
+      "loss": 0.297,
+      "step": 527
+    },
+    {
+      "epoch": 1.5975794251134645,
+      "grad_norm": 0.6972151993737078,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.2879,
+      "step": 528
+    },
+    {
+      "epoch": 1.600605143721634,
+      "grad_norm": 0.6934332066677149,
+      "learning_rate": 2.305666623880858e-06,
+      "loss": 0.2999,
+      "step": 529
+    },
+    {
+      "epoch": 1.6036308623298035,
+      "grad_norm": 0.6690675535067985,
+      "learning_rate": 2.27199290599617e-06,
+      "loss": 0.308,
+      "step": 530
+    },
+    {
+      "epoch": 1.606656580937973,
+      "grad_norm": 0.7369392497254498,
+      "learning_rate": 2.2385353570824308e-06,
+      "loss": 0.3119,
+      "step": 531
+    },
+    {
+      "epoch": 1.6096822995461422,
+      "grad_norm": 0.6989728605333634,
+      "learning_rate": 2.2052949130192136e-06,
+      "loss": 0.3091,
+      "step": 532
+    },
+    {
+      "epoch": 1.6127080181543116,
+      "grad_norm": 0.6666213750300519,
+      "learning_rate": 2.172272503613183e-06,
+      "loss": 0.2939,
+      "step": 533
+    },
+    {
+      "epoch": 1.615733736762481,
+      "grad_norm": 0.6680449560770696,
+      "learning_rate": 2.1394690525721275e-06,
+      "loss": 0.3093,
+      "step": 534
+    },
+    {
+      "epoch": 1.6187594553706506,
+      "grad_norm": 0.7007293018560314,
+      "learning_rate": 2.1068854774790783e-06,
+      "loss": 0.2977,
+      "step": 535
+    },
+    {
+      "epoch": 1.6217851739788198,
+      "grad_norm": 0.6404718319657102,
+      "learning_rate": 2.0745226897666858e-06,
+      "loss": 0.2981,
+      "step": 536
+    },
+    {
+      "epoch": 1.6248108925869893,
+      "grad_norm": 0.6828797751309723,
+      "learning_rate": 2.0423815946916783e-06,
+      "loss": 0.3003,
+      "step": 537
+    },
+    {
+      "epoch": 1.6278366111951588,
+      "grad_norm": 0.6689535921616235,
+      "learning_rate": 2.010463091309587e-06,
+      "loss": 0.2974,
+      "step": 538
+    },
+    {
+      "epoch": 1.6308623298033282,
+      "grad_norm": 0.6768424952773845,
+      "learning_rate": 1.9787680724495617e-06,
+      "loss": 0.2772,
+      "step": 539
+    },
+    {
+      "epoch": 1.6338880484114977,
+      "grad_norm": 0.64868460177231,
+      "learning_rate": 1.947297424689414e-06,
+      "loss": 0.2967,
+      "step": 540
+    },
+    {
+      "epoch": 1.6369137670196672,
+      "grad_norm": 0.6865566551379899,
+      "learning_rate": 1.9160520283308115e-06,
+      "loss": 0.3119,
+      "step": 541
+    },
+    {
+      "epoch": 1.6399394856278366,
+      "grad_norm": 0.7382793665950428,
+      "learning_rate": 1.8850327573746584e-06,
+      "loss": 0.3137,
+      "step": 542
+    },
+    {
+      "epoch": 1.6429652042360061,
+      "grad_norm": 0.6551497049623433,
+      "learning_rate": 1.854240479496643e-06,
+      "loss": 0.2898,
+      "step": 543
+    },
+    {
+      "epoch": 1.6459909228441756,
+      "grad_norm": 0.7112966406117509,
+      "learning_rate": 1.8236760560229715e-06,
+      "loss": 0.3098,
+      "step": 544
+    },
+    {
+      "epoch": 1.649016641452345,
+      "grad_norm": 0.6370498003558511,
+      "learning_rate": 1.7933403419062689e-06,
+      "loss": 0.2912,
+      "step": 545
+    },
+    {
+      "epoch": 1.6520423600605145,
+      "grad_norm": 0.6879325966454132,
+      "learning_rate": 1.7632341857016733e-06,
+      "loss": 0.2938,
+      "step": 546
+    },
+    {
+      "epoch": 1.6550680786686838,
+      "grad_norm": 0.7197130717353506,
+      "learning_rate": 1.7333584295430894e-06,
+      "loss": 0.3256,
+      "step": 547
+    },
+    {
+      "epoch": 1.6580937972768532,
+      "grad_norm": 0.6760028004615091,
+      "learning_rate": 1.7037139091196396e-06,
+      "loss": 0.3036,
+      "step": 548
+    },
+    {
+      "epoch": 1.6611195158850227,
+      "grad_norm": 0.6598159703681746,
+      "learning_rate": 1.6743014536522872e-06,
+      "loss": 0.3055,
+      "step": 549
+    },
+    {
+      "epoch": 1.6641452344931922,
+      "grad_norm": 0.7088482006326358,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 0.2997,
+      "step": 550
+    },
+    {
+      "epoch": 1.6671709531013614,
+      "grad_norm": 0.6882352095167241,
+      "learning_rate": 1.616176021989926e-06,
+      "loss": 0.2943,
+      "step": 551
+    },
+    {
+      "epoch": 1.670196671709531,
+      "grad_norm": 0.6461609507412539,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2861,
+      "step": 552
+    },
+    {
+      "epoch": 1.6732223903177004,
+      "grad_norm": 0.6935881220973035,
+      "learning_rate": 1.558988638083616e-06,
+      "loss": 0.3006,
+      "step": 553
+    },
+    {
+      "epoch": 1.6762481089258698,
+      "grad_norm": 0.6538862951693276,
+      "learning_rate": 1.5307487177120773e-06,
+      "loss": 0.2925,
+      "step": 554
+    },
+    {
+      "epoch": 1.6792738275340393,
+      "grad_norm": 0.7589708624044317,
+      "learning_rate": 1.5027457005048573e-06,
+      "loss": 0.3123,
+      "step": 555
+    },
+    {
+      "epoch": 1.6822995461422088,
+      "grad_norm": 0.6831541192496898,
+      "learning_rate": 1.4749803697665366e-06,
+      "loss": 0.2941,
+      "step": 556
+    },
+    {
+      "epoch": 1.6853252647503782,
+      "grad_norm": 0.6309343581635191,
+      "learning_rate": 1.4474535021531099e-06,
+      "loss": 0.2816,
+      "step": 557
+    },
+    {
+      "epoch": 1.6883509833585477,
+      "grad_norm": 0.6686527457604755,
+      "learning_rate": 1.4201658676502294e-06,
+      "loss": 0.2882,
+      "step": 558
+    },
+    {
+      "epoch": 1.6913767019667172,
+      "grad_norm": 0.6690445301873396,
+      "learning_rate": 1.3931182295516965e-06,
+      "loss": 0.301,
+      "step": 559
+    },
+    {
+      "epoch": 1.6944024205748867,
+      "grad_norm": 0.6956697274989876,
+      "learning_rate": 1.3663113444380905e-06,
+      "loss": 0.3025,
+      "step": 560
+    },
+    {
+      "epoch": 1.6974281391830561,
+      "grad_norm": 0.694791352371976,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.3115,
+      "step": 561
+    },
+    {
+      "epoch": 1.7004538577912254,
+      "grad_norm": 0.6764261372283936,
+      "learning_rate": 1.3134228257951142e-06,
+      "loss": 0.2928,
+      "step": 562
+    },
+    {
+      "epoch": 1.7034795763993948,
+      "grad_norm": 0.7029347092813593,
+      "learning_rate": 1.2873426716713012e-06,
+      "loss": 0.3015,
+      "step": 563
+    },
+    {
+      "epoch": 1.7065052950075643,
+      "grad_norm": 0.643842413682639,
+      "learning_rate": 1.2615062293021508e-06,
+      "loss": 0.2876,
+      "step": 564
+    },
+    {
+      "epoch": 1.7095310136157338,
+      "grad_norm": 0.6579806462038966,
+      "learning_rate": 1.2359142213884933e-06,
+      "loss": 0.289,
+      "step": 565
+    },
+    {
+      "epoch": 1.712556732223903,
+      "grad_norm": 0.6513867473150848,
+      "learning_rate": 1.2105673637938054e-06,
+      "loss": 0.2835,
+      "step": 566
+    },
+    {
+      "epoch": 1.7155824508320725,
+      "grad_norm": 0.6786555553103795,
+      "learning_rate": 1.1854663655241804e-06,
+      "loss": 0.2966,
+      "step": 567
+    },
+    {
+      "epoch": 1.718608169440242,
+      "grad_norm": 0.698261419673545,
+      "learning_rate": 1.1606119287084982e-06,
+      "loss": 0.3153,
+      "step": 568
+    },
+    {
+      "epoch": 1.7216338880484114,
+      "grad_norm": 0.6629022486571401,
+      "learning_rate": 1.136004748578785e-06,
+      "loss": 0.2931,
+      "step": 569
+    },
+    {
+      "epoch": 1.724659606656581,
+      "grad_norm": 0.6667951796661243,
+      "learning_rate": 1.1116455134507665e-06,
+      "loss": 0.2929,
+      "step": 570
+    },
+    {
+      "epoch": 1.7276853252647504,
+      "grad_norm": 0.675500503187363,
+      "learning_rate": 1.0875349047046113e-06,
+      "loss": 0.284,
+      "step": 571
+    },
+    {
+      "epoch": 1.7307110438729199,
+      "grad_norm": 0.664134191056973,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 0.2866,
+      "step": 572
+    },
+    {
+      "epoch": 1.7337367624810893,
+      "grad_norm": 0.7050883839557025,
+      "learning_rate": 1.0400622570866426e-06,
+      "loss": 0.2922,
+      "step": 573
+    },
+    {
+      "epoch": 1.7367624810892588,
+      "grad_norm": 0.6688146517568023,
+      "learning_rate": 1.0167015461268303e-06,
+      "loss": 0.292,
+      "step": 574
+    },
+    {
+      "epoch": 1.7397881996974283,
+      "grad_norm": 0.6326947922482544,
+      "learning_rate": 9.935921173357444e-07,
+      "loss": 0.279,
+      "step": 575
+    },
+    {
+      "epoch": 1.7428139183055977,
+      "grad_norm": 0.6965184231322301,
+      "learning_rate": 9.707346171337895e-07,
+      "loss": 0.2908,
+      "step": 576
+    },
+    {
+      "epoch": 1.745839636913767,
+      "grad_norm": 0.6477879083180327,
+      "learning_rate": 9.481296848943744e-07,
+      "loss": 0.2834,
+      "step": 577
+    },
+    {
+      "epoch": 1.7488653555219364,
+      "grad_norm": 0.6663148015934732,
+      "learning_rate": 9.257779529260558e-07,
+      "loss": 0.2942,
+      "step": 578
+    },
+    {
+      "epoch": 1.751891074130106,
+      "grad_norm": 0.6635539576873594,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.279,
+      "step": 579
+    },
+    {
+      "epoch": 1.7549167927382754,
+      "grad_norm": 0.649088532474129,
+      "learning_rate": 8.818365836066101e-07,
+      "loss": 0.2801,
+      "step": 580
+    },
+    {
+      "epoch": 1.7579425113464446,
+      "grad_norm": 0.6584667039859262,
+      "learning_rate": 8.602481753900427e-07,
+      "loss": 0.2962,
+      "step": 581
+    },
+    {
+      "epoch": 1.760968229954614,
+      "grad_norm": 0.6694188935688425,
+      "learning_rate": 8.389154256793042e-07,
+      "loss": 0.2838,
+      "step": 582
+    },
+    {
+      "epoch": 1.7639939485627836,
+      "grad_norm": 0.721208324353601,
+      "learning_rate": 8.178389311972612e-07,
+      "loss": 0.3337,
+      "step": 583
+    },
+    {
+      "epoch": 1.767019667170953,
+      "grad_norm": 0.6633804467464602,
+      "learning_rate": 7.970192814987676e-07,
+      "loss": 0.2939,
+      "step": 584
+    },
+    {
+      "epoch": 1.7700453857791225,
+      "grad_norm": 0.6604004001698957,
+      "learning_rate": 7.764570589541876e-07,
+      "loss": 0.279,
+      "step": 585
+    },
+    {
+      "epoch": 1.773071104387292,
+      "grad_norm": 0.6598181341159414,
+      "learning_rate": 7.561528387330797e-07,
+      "loss": 0.2732,
+      "step": 586
+    },
+    {
+      "epoch": 1.7760968229954615,
+      "grad_norm": 0.6959930985616869,
+      "learning_rate": 7.361071887881376e-07,
+      "loss": 0.3021,
+      "step": 587
+    },
+    {
+      "epoch": 1.779122541603631,
+      "grad_norm": 0.6776447122636321,
+      "learning_rate": 7.163206698392744e-07,
+      "loss": 0.2862,
+      "step": 588
+    },
+    {
+      "epoch": 1.7821482602118004,
+      "grad_norm": 0.7070190566774118,
+      "learning_rate": 6.96793835357964e-07,
+      "loss": 0.2987,
+      "step": 589
+    },
+    {
+      "epoch": 1.7851739788199699,
+      "grad_norm": 0.6683547636129819,
+      "learning_rate": 6.775272315517423e-07,
+      "loss": 0.2823,
+      "step": 590
+    },
+    {
+      "epoch": 1.7881996974281393,
+      "grad_norm": 0.711790455074334,
+      "learning_rate": 6.585213973489335e-07,
+      "loss": 0.309,
+      "step": 591
+    },
+    {
+      "epoch": 1.7912254160363086,
+      "grad_norm": 0.6914836654776516,
+      "learning_rate": 6.397768643835755e-07,
+      "loss": 0.3052,
+      "step": 592
+    },
+    {
+      "epoch": 1.794251134644478,
+      "grad_norm": 0.7318896627223466,
+      "learning_rate": 6.212941569805508e-07,
+      "loss": 0.3049,
+      "step": 593
+    },
+    {
+      "epoch": 1.7972768532526475,
+      "grad_norm": 0.7077354070195336,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.3113,
+      "step": 594
+    },
+    {
+      "epoch": 1.800302571860817,
+      "grad_norm": 0.6880261143441702,
+      "learning_rate": 5.851162795274445e-07,
+      "loss": 0.2819,
+      "step": 595
+    },
+    {
+      "epoch": 1.8033282904689862,
+      "grad_norm": 0.6606169598516135,
+      "learning_rate": 5.674221214503639e-07,
+      "loss": 0.281,
+      "step": 596
+    },
+    {
+      "epoch": 1.8063540090771557,
+      "grad_norm": 0.6810754761438064,
+      "learning_rate": 5.499918128533155e-07,
+      "loss": 0.2752,
+      "step": 597
+    },
+    {
+      "epoch": 1.8093797276853252,
+      "grad_norm": 0.7050164222045213,
+      "learning_rate": 5.328258412994958e-07,
+      "loss": 0.307,
+      "step": 598
+    },
+    {
+      "epoch": 1.8124054462934946,
+      "grad_norm": 0.6834282673388616,
+      "learning_rate": 5.159246869580348e-07,
+      "loss": 0.2938,
+      "step": 599
+    },
+    {
+      "epoch": 1.8154311649016641,
+      "grad_norm": 0.7234512089625713,
+      "learning_rate": 4.992888225905467e-07,
+      "loss": 0.2997,
+      "step": 600
+    },
+    {
+      "epoch": 1.8184568835098336,
+      "grad_norm": 0.6889922711816542,
+      "learning_rate": 4.829187135379221e-07,
+      "loss": 0.3035,
+      "step": 601
+    },
+    {
+      "epoch": 1.821482602118003,
+      "grad_norm": 0.6713983710589325,
+      "learning_rate": 4.6681481770729844e-07,
+      "loss": 0.3002,
+      "step": 602
+    },
+    {
+      "epoch": 1.8245083207261725,
+      "grad_norm": 0.7306813304904591,
+      "learning_rate": 4.509775855592613e-07,
+      "loss": 0.3063,
+      "step": 603
+    },
+    {
+      "epoch": 1.827534039334342,
+      "grad_norm": 0.6721185445534387,
+      "learning_rate": 4.354074600952407e-07,
+      "loss": 0.2864,
+      "step": 604
+    },
+    {
+      "epoch": 1.8305597579425115,
+      "grad_norm": 0.6977332727090203,
+      "learning_rate": 4.2010487684511105e-07,
+      "loss": 0.2984,
+      "step": 605
+    },
+    {
+      "epoch": 1.833585476550681,
+      "grad_norm": 0.6546814804518726,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2691,
+      "step": 606
+    },
+    {
+      "epoch": 1.8366111951588502,
+      "grad_norm": 0.6718890894666122,
+      "learning_rate": 3.9030404167542777e-07,
+      "loss": 0.2924,
+      "step": 607
+    },
+    {
+      "epoch": 1.8396369137670197,
+      "grad_norm": 0.6799521567060918,
+      "learning_rate": 3.7580662334929517e-07,
+      "loss": 0.3003,
+      "step": 608
+    },
+    {
+      "epoch": 1.8426626323751891,
+      "grad_norm": 0.6982278583754727,
+      "learning_rate": 3.615784144005796e-07,
+      "loss": 0.3033,
+      "step": 609
+    },
+    {
+      "epoch": 1.8456883509833586,
+      "grad_norm": 0.6621381601368673,
+      "learning_rate": 3.476198128228736e-07,
+      "loss": 0.2944,
+      "step": 610
+    },
+    {
+      "epoch": 1.8487140695915278,
+      "grad_norm": 0.6933881670979741,
+      "learning_rate": 3.339312090682689e-07,
+      "loss": 0.2949,
+      "step": 611
+    },
+    {
+      "epoch": 1.8517397881996973,
+      "grad_norm": 0.6745856415252383,
+      "learning_rate": 3.2051298603643754e-07,
+      "loss": 0.2948,
+      "step": 612
+    },
+    {
+      "epoch": 1.8547655068078668,
+      "grad_norm": 0.703464191331348,
+      "learning_rate": 3.0736551906392354e-07,
+      "loss": 0.2973,
+      "step": 613
+    },
+    {
+      "epoch": 1.8577912254160363,
+      "grad_norm": 0.6792667389383877,
+      "learning_rate": 2.9448917591363923e-07,
+      "loss": 0.2914,
+      "step": 614
+    },
+    {
+      "epoch": 1.8608169440242057,
+      "grad_norm": 0.7009468850147937,
+      "learning_rate": 2.818843167645835e-07,
+      "loss": 0.2903,
+      "step": 615
+    },
+    {
+      "epoch": 1.8638426626323752,
+      "grad_norm": 0.6756363990496296,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 0.296,
+      "step": 616
+    },
+    {
+      "epoch": 1.8668683812405447,
+      "grad_norm": 0.6469410223883402,
+      "learning_rate": 2.5749045320632824e-07,
+      "loss": 0.2877,
+      "step": 617
+    },
+    {
+      "epoch": 1.8698940998487141,
+      "grad_norm": 0.6605017330026083,
+      "learning_rate": 2.4570213114592957e-07,
+      "loss": 0.2803,
+      "step": 618
+    },
+    {
+      "epoch": 1.8729198184568836,
+      "grad_norm": 0.6775799240187781,
+      "learning_rate": 2.3418665776527738e-07,
+      "loss": 0.2852,
+      "step": 619
+    },
+    {
+      "epoch": 1.875945537065053,
+      "grad_norm": 0.7285636653055645,
+      "learning_rate": 2.2294435517691504e-07,
+      "loss": 0.3096,
+      "step": 620
+    },
+    {
+      "epoch": 1.8789712556732225,
+      "grad_norm": 0.64010731451208,
+      "learning_rate": 2.119755378522137e-07,
+      "loss": 0.2788,
+      "step": 621
+    },
+    {
+      "epoch": 1.8819969742813918,
+      "grad_norm": 0.6488258765901107,
+      "learning_rate": 2.0128051261257165e-07,
+      "loss": 0.2779,
+      "step": 622
+    },
+    {
+      "epoch": 1.8850226928895613,
+      "grad_norm": 0.6399795016535205,
+      "learning_rate": 1.908595786208367e-07,
+      "loss": 0.2931,
+      "step": 623
+    },
+    {
+      "epoch": 1.8880484114977307,
+      "grad_norm": 0.6862209718955662,
+      "learning_rate": 1.8071302737293294e-07,
+      "loss": 0.2996,
+      "step": 624
+    },
+    {
+      "epoch": 1.8910741301059002,
+      "grad_norm": 0.6865084003052813,
+      "learning_rate": 1.7084114268971275e-07,
+      "loss": 0.2962,
+      "step": 625
+    },
+    {
+      "epoch": 1.8940998487140694,
+      "grad_norm": 0.6622551725178941,
+      "learning_rate": 1.612442007090076e-07,
+      "loss": 0.2711,
+      "step": 626
+    },
+    {
+      "epoch": 1.897125567322239,
+      "grad_norm": 0.67160587565472,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 0.2999,
+      "step": 627
+    },
+    {
+      "epoch": 1.9001512859304084,
+      "grad_norm": 0.6425136291535823,
+      "learning_rate": 1.4287621094529524e-07,
+      "loss": 0.2735,
+      "step": 628
+    },
+    {
+      "epoch": 1.9031770045385779,
+      "grad_norm": 0.7649165574009369,
+      "learning_rate": 1.3410567695444576e-07,
+      "loss": 0.329,
+      "step": 629
+    },
+    {
+      "epoch": 1.9062027231467473,
+      "grad_norm": 0.6769293863101926,
+      "learning_rate": 1.2561111323605714e-07,
+      "loss": 0.2865,
+      "step": 630
+    },
+    {
+      "epoch": 1.9092284417549168,
+      "grad_norm": 0.6866277812563215,
+      "learning_rate": 1.1739275740134004e-07,
+      "loss": 0.2971,
+      "step": 631
+    },
+    {
+      "epoch": 1.9122541603630863,
+      "grad_norm": 0.6966317379840327,
+      "learning_rate": 1.0945083933537104e-07,
+      "loss": 0.3083,
+      "step": 632
+    },
+    {
+      "epoch": 1.9152798789712557,
+      "grad_norm": 0.659587582703165,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2678,
+      "step": 633
+    },
+    {
+      "epoch": 1.9183055975794252,
+      "grad_norm": 0.687184022160123,
+      "learning_rate": 9.439719738099318e-08,
+      "loss": 0.2883,
+      "step": 634
+    },
+    {
+      "epoch": 1.9213313161875947,
+      "grad_norm": 0.6931304999400624,
+      "learning_rate": 8.728589457530857e-08,
+      "loss": 0.278,
+      "step": 635
+    },
+    {
+      "epoch": 1.9243570347957641,
+      "grad_norm": 0.6829598072833263,
+      "learning_rate": 8.04518716920466e-08,
+      "loss": 0.2949,
+      "step": 636
+    },
+    {
+      "epoch": 1.9273827534039334,
+      "grad_norm": 0.6414807219673527,
+      "learning_rate": 7.389531989351773e-08,
+      "loss": 0.2723,
+      "step": 637
+    },
+    {
+      "epoch": 1.9304084720121029,
+      "grad_norm": 0.6920820176757975,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 0.287,
+      "step": 638
+    },
+    {
+      "epoch": 1.9334341906202723,
+      "grad_norm": 0.6538262787085444,
+      "learning_rate": 6.161535538745877e-08,
+      "loss": 0.3063,
+      "step": 639
+    },
+    {
+      "epoch": 1.9364599092284418,
+      "grad_norm": 0.6961134953617497,
+      "learning_rate": 5.5892286176932875e-08,
+      "loss": 0.2985,
+      "step": 640
+    },
+    {
+      "epoch": 1.939485627836611,
+      "grad_norm": 0.6826715979777553,
+      "learning_rate": 5.044737503554165e-08,
+      "loss": 0.294,
+      "step": 641
+    },
+    {
+      "epoch": 1.9425113464447805,
+      "grad_norm": 0.6797567867081026,
+      "learning_rate": 4.528077426915412e-08,
+      "loss": 0.2934,
+      "step": 642
+    },
+    {
+      "epoch": 1.94553706505295,
+      "grad_norm": 0.6941501589612522,
+      "learning_rate": 4.0392628398699954e-08,
+      "loss": 0.3041,
+      "step": 643
+    },
+    {
+      "epoch": 1.9485627836611195,
+      "grad_norm": 0.632517183934029,
+      "learning_rate": 3.578307415612714e-08,
+      "loss": 0.2816,
+      "step": 644
+    },
+    {
+      "epoch": 1.951588502269289,
+      "grad_norm": 0.6883001806517491,
+      "learning_rate": 3.1452240480577265e-08,
+      "loss": 0.2771,
+      "step": 645
+    },
+    {
+      "epoch": 1.9546142208774584,
+      "grad_norm": 0.6722584559218414,
+      "learning_rate": 2.7400248514776184e-08,
+      "loss": 0.2955,
+      "step": 646
+    },
+    {
+      "epoch": 1.9576399394856279,
+      "grad_norm": 0.7161685769228949,
+      "learning_rate": 2.3627211601651157e-08,
+      "loss": 0.2956,
+      "step": 647
+    },
+    {
+      "epoch": 1.9606656580937973,
+      "grad_norm": 0.719307783822232,
+      "learning_rate": 2.013323528115674e-08,
+      "loss": 0.301,
+      "step": 648
+    },
+    {
+      "epoch": 1.9636913767019668,
+      "grad_norm": 0.6899617049527405,
+      "learning_rate": 1.6918417287318245e-08,
+      "loss": 0.2893,
+      "step": 649
+    },
+    {
+      "epoch": 1.9667170953101363,
+      "grad_norm": 0.7197753698743058,
+      "learning_rate": 1.3982847545507271e-08,
+      "loss": 0.2931,
+      "step": 650
+    },
+    {
+      "epoch": 1.9697428139183057,
+      "grad_norm": 0.6747037390045129,
+      "learning_rate": 1.1326608169920373e-08,
+      "loss": 0.2875,
+      "step": 651
+    },
+    {
+      "epoch": 1.972768532526475,
+      "grad_norm": 0.6922072213222901,
+      "learning_rate": 8.949773461282008e-09,
+      "loss": 0.2919,
+      "step": 652
+    },
+    {
+      "epoch": 1.9757942511346445,
+      "grad_norm": 0.6699845681439842,
+      "learning_rate": 6.8524099047695415e-09,
+      "loss": 0.2932,
+      "step": 653
+    },
+    {
+      "epoch": 1.978819969742814,
+      "grad_norm": 0.6698231305421234,
+      "learning_rate": 5.034576168149175e-09,
+      "loss": 0.2889,
+      "step": 654
+    },
+    {
+      "epoch": 1.9818456883509834,
+      "grad_norm": 0.6741930301945138,
+      "learning_rate": 3.4963231001383657e-09,
+      "loss": 0.2883,
+      "step": 655
+    },
+    {
+      "epoch": 1.9848714069591527,
+      "grad_norm": 0.6275395755719684,
+      "learning_rate": 2.237693728981416e-09,
+      "loss": 0.2852,
+      "step": 656
+    },
+    {
+      "epoch": 1.9878971255673221,
+      "grad_norm": 0.7055407733593824,
+      "learning_rate": 1.2587232612493172e-09,
+      "loss": 0.2919,
+      "step": 657
+    },
+    {
+      "epoch": 1.9909228441754916,
+      "grad_norm": 0.6847420748901661,
+      "learning_rate": 5.594390808494332e-10,
+      "loss": 0.307,
+      "step": 658
+    },
+    {
+      "epoch": 1.993948562783661,
+      "grad_norm": 0.640220364775614,
+      "learning_rate": 1.3986074826388697e-10,
+      "loss": 0.2731,
+      "step": 659
+    },
+    {
+      "epoch": 1.9969742813918305,
+      "grad_norm": 0.6867834548536167,
+      "learning_rate": 0.0,
+      "loss": 0.3053,
+      "step": 660
+    },
+    {
+      "epoch": 1.9969742813918305,
+      "step": 660,
+      "total_flos": 1.3381716445180723e+17,
+      "train_loss": 0.3828992527091142,
+      "train_runtime": 1786.8451,
+      "train_samples_per_second": 47.317,
+      "train_steps_per_second": 0.369
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 660,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3381716445180723e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}