| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 3.760205659470345, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.9944963455200195, | |
| "logits/rejected": -2.9324326515197754, | |
| "logps/chosen": -1.0260854959487915, | |
| "logps/rejected": -1.189737319946289, | |
| "loss": -0.0244, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0260854959487915, | |
| "rewards/margins": 0.1636517494916916, | |
| "rewards/rejected": -1.189737319946289, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 4.14164094224387, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -3.0354819297790527, | |
| "logits/rejected": -2.9778807163238525, | |
| "logps/chosen": -1.1017096042633057, | |
| "logps/rejected": -1.1320788860321045, | |
| "loss": -0.0273, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.1017096042633057, | |
| "rewards/margins": 0.030369168147444725, | |
| "rewards/rejected": -1.1320788860321045, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 4.393103002918313, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -3.0005249977111816, | |
| "logits/rejected": -2.995767116546631, | |
| "logps/chosen": -1.0664865970611572, | |
| "logps/rejected": -1.0592315196990967, | |
| "loss": -0.0311, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -1.0664865970611572, | |
| "rewards/margins": -0.007255014963448048, | |
| "rewards/rejected": -1.0592315196990967, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 4.089316161372408, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -3.015296697616577, | |
| "logits/rejected": -2.9769978523254395, | |
| "logps/chosen": -1.0041254758834839, | |
| "logps/rejected": -1.0410144329071045, | |
| "loss": -0.0327, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -1.0041254758834839, | |
| "rewards/margins": 0.03688892349600792, | |
| "rewards/rejected": -1.0410144329071045, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 4.216563165990877, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -3.0237412452697754, | |
| "logits/rejected": -2.973569631576538, | |
| "logps/chosen": -0.9541431665420532, | |
| "logps/rejected": -1.1942131519317627, | |
| "loss": -0.0413, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9541431665420532, | |
| "rewards/margins": 0.2400699108839035, | |
| "rewards/rejected": -1.1942131519317627, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 5.658050284469291, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -3.030153274536133, | |
| "logits/rejected": -2.9625256061553955, | |
| "logps/chosen": -0.8760479092597961, | |
| "logps/rejected": -0.9085603952407837, | |
| "loss": -0.0368, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.8760479092597961, | |
| "rewards/margins": 0.032512493431568146, | |
| "rewards/rejected": -0.9085603952407837, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 5.044506012240391, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -3.0141396522521973, | |
| "logits/rejected": -2.995772361755371, | |
| "logps/chosen": -0.8659467697143555, | |
| "logps/rejected": -1.1117627620697021, | |
| "loss": -0.0476, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8659467697143555, | |
| "rewards/margins": 0.24581606686115265, | |
| "rewards/rejected": -1.1117627620697021, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 4.594342778483098, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.9778103828430176, | |
| "logits/rejected": -2.9474072456359863, | |
| "logps/chosen": -0.8786975741386414, | |
| "logps/rejected": -0.9444503784179688, | |
| "loss": -0.0484, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.8786975741386414, | |
| "rewards/margins": 0.06575289368629456, | |
| "rewards/rejected": -0.9444503784179688, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 4.865607139601026, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -3.025959014892578, | |
| "logits/rejected": -3.019007444381714, | |
| "logps/chosen": -0.8325697183609009, | |
| "logps/rejected": -0.9673300981521606, | |
| "loss": -0.0615, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8325697183609009, | |
| "rewards/margins": 0.134760320186615, | |
| "rewards/rejected": -0.9673300981521606, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 5.424026197216979, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.8837554454803467, | |
| "logits/rejected": -2.9218649864196777, | |
| "logps/chosen": -0.9015443921089172, | |
| "logps/rejected": -1.0318125486373901, | |
| "loss": -0.0525, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9015443921089172, | |
| "rewards/margins": 0.13026830554008484, | |
| "rewards/rejected": -1.0318125486373901, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 7.570587115119674, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -3.0432026386260986, | |
| "logits/rejected": -3.038170576095581, | |
| "logps/chosen": -0.9203222393989563, | |
| "logps/rejected": -1.0882291793823242, | |
| "loss": -0.0483, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.9203222393989563, | |
| "rewards/margins": 0.1679070144891739, | |
| "rewards/rejected": -1.0882291793823242, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 8.483337894809958, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.9748623371124268, | |
| "logits/rejected": -2.9038889408111572, | |
| "logps/chosen": -0.9680806398391724, | |
| "logps/rejected": -1.1494948863983154, | |
| "loss": -0.0682, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9680806398391724, | |
| "rewards/margins": 0.18141421675682068, | |
| "rewards/rejected": -1.1494948863983154, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 6.597482381521931, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.9652066230773926, | |
| "logits/rejected": -2.988760232925415, | |
| "logps/chosen": -0.9523311853408813, | |
| "logps/rejected": -1.1883153915405273, | |
| "loss": -0.0605, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9523311853408813, | |
| "rewards/margins": 0.23598428070545197, | |
| "rewards/rejected": -1.1883153915405273, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 7.227660542967968, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -3.0324559211730957, | |
| "logits/rejected": -2.9957118034362793, | |
| "logps/chosen": -1.03068208694458, | |
| "logps/rejected": -1.2021524906158447, | |
| "loss": -0.0806, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.03068208694458, | |
| "rewards/margins": 0.1714704930782318, | |
| "rewards/rejected": -1.2021524906158447, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 8.630957672516889, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.99220609664917, | |
| "logits/rejected": -2.9557840824127197, | |
| "logps/chosen": -1.0696258544921875, | |
| "logps/rejected": -1.317363977432251, | |
| "loss": -0.0631, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0696258544921875, | |
| "rewards/margins": 0.2477382868528366, | |
| "rewards/rejected": -1.317363977432251, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 6.399531931157647, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.994518280029297, | |
| "logits/rejected": -2.961357831954956, | |
| "logps/chosen": -1.0338717699050903, | |
| "logps/rejected": -1.4787652492523193, | |
| "loss": -0.0783, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0338717699050903, | |
| "rewards/margins": 0.44489365816116333, | |
| "rewards/rejected": -1.4787652492523193, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 7.358699467281965, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.9917774200439453, | |
| "logits/rejected": -2.9551658630371094, | |
| "logps/chosen": -0.9378916025161743, | |
| "logps/rejected": -1.334387183189392, | |
| "loss": -0.0833, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9378916025161743, | |
| "rewards/margins": 0.396495521068573, | |
| "rewards/rejected": -1.334387183189392, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 5.13038386551715, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.9607720375061035, | |
| "logits/rejected": -2.9549806118011475, | |
| "logps/chosen": -1.0484223365783691, | |
| "logps/rejected": -1.3985503911972046, | |
| "loss": -0.0862, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0484223365783691, | |
| "rewards/margins": 0.3501282036304474, | |
| "rewards/rejected": -1.3985503911972046, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 6.228634016037295, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -3.039447784423828, | |
| "logits/rejected": -3.0229854583740234, | |
| "logps/chosen": -1.0390279293060303, | |
| "logps/rejected": -1.2659882307052612, | |
| "loss": -0.0748, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0390279293060303, | |
| "rewards/margins": 0.22696030139923096, | |
| "rewards/rejected": -1.2659882307052612, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 13.46485741296935, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -3.0565896034240723, | |
| "logits/rejected": -3.0682952404022217, | |
| "logps/chosen": -1.0358636379241943, | |
| "logps/rejected": -1.3846577405929565, | |
| "loss": -0.0825, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.0358636379241943, | |
| "rewards/margins": 0.34879425168037415, | |
| "rewards/rejected": -1.3846577405929565, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 13.308269094275541, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -3.013173818588257, | |
| "logits/rejected": -3.020575761795044, | |
| "logps/chosen": -1.2395133972167969, | |
| "logps/rejected": -1.6027743816375732, | |
| "loss": -0.1058, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2395133972167969, | |
| "rewards/margins": 0.363260954618454, | |
| "rewards/rejected": -1.6027743816375732, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 14.80943809386291, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.907844066619873, | |
| "logits/rejected": -2.9418551921844482, | |
| "logps/chosen": -1.0689195394515991, | |
| "logps/rejected": -1.62472403049469, | |
| "loss": -0.1078, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0689195394515991, | |
| "rewards/margins": 0.5558044910430908, | |
| "rewards/rejected": -1.62472403049469, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 7.302061093053783, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -3.064298152923584, | |
| "logits/rejected": -3.0547451972961426, | |
| "logps/chosen": -1.0759824514389038, | |
| "logps/rejected": -1.38835608959198, | |
| "loss": -0.089, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0759824514389038, | |
| "rewards/margins": 0.3123735785484314, | |
| "rewards/rejected": -1.38835608959198, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 6.576904824772524, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.995882272720337, | |
| "logits/rejected": -2.9872708320617676, | |
| "logps/chosen": -1.010962724685669, | |
| "logps/rejected": -1.5580971240997314, | |
| "loss": -0.0761, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.010962724685669, | |
| "rewards/margins": 0.547134518623352, | |
| "rewards/rejected": -1.5580971240997314, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 10.915099417857844, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.8915393352508545, | |
| "logits/rejected": -2.841787338256836, | |
| "logps/chosen": -1.0743449926376343, | |
| "logps/rejected": -1.400960922241211, | |
| "loss": -0.0844, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0743449926376343, | |
| "rewards/margins": 0.3266158401966095, | |
| "rewards/rejected": -1.400960922241211, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 7.5813693492052066, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.9785983562469482, | |
| "logits/rejected": -3.023461103439331, | |
| "logps/chosen": -1.0497643947601318, | |
| "logps/rejected": -1.5954610109329224, | |
| "loss": -0.0988, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0497643947601318, | |
| "rewards/margins": 0.5456965565681458, | |
| "rewards/rejected": -1.5954610109329224, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 7.880456151156036, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.9451870918273926, | |
| "logits/rejected": -2.9251372814178467, | |
| "logps/chosen": -1.1994966268539429, | |
| "logps/rejected": -1.6415138244628906, | |
| "loss": -0.0959, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1994966268539429, | |
| "rewards/margins": 0.442017138004303, | |
| "rewards/rejected": -1.6415138244628906, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 11.389751848981007, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.9757840633392334, | |
| "logits/rejected": -2.9875569343566895, | |
| "logps/chosen": -1.3223577737808228, | |
| "logps/rejected": -1.7084649801254272, | |
| "loss": -0.0795, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.3223577737808228, | |
| "rewards/margins": 0.38610729575157166, | |
| "rewards/rejected": -1.7084649801254272, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 12.240889169213316, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.884559392929077, | |
| "logits/rejected": -2.9181649684906006, | |
| "logps/chosen": -1.162393569946289, | |
| "logps/rejected": -1.6213098764419556, | |
| "loss": -0.0999, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.162393569946289, | |
| "rewards/margins": 0.4589163362979889, | |
| "rewards/rejected": -1.6213098764419556, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 13.394133386548184, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.988391160964966, | |
| "logits/rejected": -2.9592318534851074, | |
| "logps/chosen": -1.1906192302703857, | |
| "logps/rejected": -2.017512559890747, | |
| "loss": -0.0867, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1906192302703857, | |
| "rewards/margins": 0.8268932104110718, | |
| "rewards/rejected": -2.017512559890747, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 10.341335190859326, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.871455192565918, | |
| "logits/rejected": -2.853935718536377, | |
| "logps/chosen": -1.3897101879119873, | |
| "logps/rejected": -1.8974939584732056, | |
| "loss": -0.1116, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.3897101879119873, | |
| "rewards/margins": 0.5077837705612183, | |
| "rewards/rejected": -1.8974939584732056, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 14.583251944991732, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.824887752532959, | |
| "logits/rejected": -2.8534679412841797, | |
| "logps/chosen": -1.3409488201141357, | |
| "logps/rejected": -2.2745461463928223, | |
| "loss": -0.1046, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.3409488201141357, | |
| "rewards/margins": 0.933597207069397, | |
| "rewards/rejected": -2.2745461463928223, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 13.977236800680524, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.9754512310028076, | |
| "logits/rejected": -2.960723400115967, | |
| "logps/chosen": -1.255544662475586, | |
| "logps/rejected": -1.7897508144378662, | |
| "loss": -0.1166, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.255544662475586, | |
| "rewards/margins": 0.534206211566925, | |
| "rewards/rejected": -1.7897508144378662, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 21.73429928886912, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.92985200881958, | |
| "logits/rejected": -2.9375405311584473, | |
| "logps/chosen": -1.1773474216461182, | |
| "logps/rejected": -1.8664484024047852, | |
| "loss": -0.1043, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1773474216461182, | |
| "rewards/margins": 0.6891010999679565, | |
| "rewards/rejected": -1.8664484024047852, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 65.37359700720873, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.851064682006836, | |
| "logits/rejected": -2.87498140335083, | |
| "logps/chosen": -1.0109388828277588, | |
| "logps/rejected": -1.6845824718475342, | |
| "loss": -0.1099, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0109388828277588, | |
| "rewards/margins": 0.6736436486244202, | |
| "rewards/rejected": -1.6845824718475342, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 18.528930954008196, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.8931515216827393, | |
| "logits/rejected": -2.873427629470825, | |
| "logps/chosen": -1.1016333103179932, | |
| "logps/rejected": -1.9188123941421509, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1016333103179932, | |
| "rewards/margins": 0.8171789050102234, | |
| "rewards/rejected": -1.9188123941421509, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 7.80304044080231, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.8918251991271973, | |
| "logits/rejected": -2.9344098567962646, | |
| "logps/chosen": -1.0134468078613281, | |
| "logps/rejected": -1.6890709400177002, | |
| "loss": -0.1128, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0134468078613281, | |
| "rewards/margins": 0.6756240129470825, | |
| "rewards/rejected": -1.6890709400177002, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 7.7047015056100046, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.8743629455566406, | |
| "logits/rejected": -2.911881685256958, | |
| "logps/chosen": -1.0063416957855225, | |
| "logps/rejected": -1.913240671157837, | |
| "loss": -0.1187, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0063416957855225, | |
| "rewards/margins": 0.9068989753723145, | |
| "rewards/rejected": -1.913240671157837, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 7.811346817640016, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.9131577014923096, | |
| "logits/rejected": -2.89136004447937, | |
| "logps/chosen": -1.1495698690414429, | |
| "logps/rejected": -2.071730375289917, | |
| "loss": -0.0945, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.1495698690414429, | |
| "rewards/margins": 0.9221606254577637, | |
| "rewards/rejected": -2.071730375289917, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 11.061536990800207, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.813307523727417, | |
| "logits/rejected": -2.7539114952087402, | |
| "logps/chosen": -1.268898367881775, | |
| "logps/rejected": -1.69308602809906, | |
| "loss": -0.1163, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.268898367881775, | |
| "rewards/margins": 0.4241875112056732, | |
| "rewards/rejected": -1.69308602809906, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 15.040971267646988, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.8270034790039062, | |
| "logits/rejected": -2.855402946472168, | |
| "logps/chosen": -1.5387921333312988, | |
| "logps/rejected": -2.168403148651123, | |
| "loss": -0.1198, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.5387921333312988, | |
| "rewards/margins": 0.6296108961105347, | |
| "rewards/rejected": -2.168403148651123, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 18.3545573133183, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.818117618560791, | |
| "logits/rejected": -2.7833092212677, | |
| "logps/chosen": -1.293579339981079, | |
| "logps/rejected": -1.8654571771621704, | |
| "loss": -0.1002, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.293579339981079, | |
| "rewards/margins": 0.5718778371810913, | |
| "rewards/rejected": -1.8654571771621704, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 8.458661335807625, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.7233188152313232, | |
| "logits/rejected": -2.727560043334961, | |
| "logps/chosen": -1.3122286796569824, | |
| "logps/rejected": -2.4676425457000732, | |
| "loss": -0.1053, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.3122286796569824, | |
| "rewards/margins": 1.1554138660430908, | |
| "rewards/rejected": -2.4676425457000732, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 8.86883909656063, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.719998359680176, | |
| "logits/rejected": -2.7265241146087646, | |
| "logps/chosen": -1.129214882850647, | |
| "logps/rejected": -1.7588284015655518, | |
| "loss": -0.1146, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.129214882850647, | |
| "rewards/margins": 0.6296136975288391, | |
| "rewards/rejected": -1.7588284015655518, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 10.016548371714508, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.811687707901001, | |
| "logits/rejected": -2.7085366249084473, | |
| "logps/chosen": -1.3638594150543213, | |
| "logps/rejected": -2.3513691425323486, | |
| "loss": -0.1184, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.3638594150543213, | |
| "rewards/margins": 0.9875099062919617, | |
| "rewards/rejected": -2.3513691425323486, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 21.998434755048503, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.6830623149871826, | |
| "logits/rejected": -2.6364707946777344, | |
| "logps/chosen": -1.3768701553344727, | |
| "logps/rejected": -2.424539089202881, | |
| "loss": -0.1267, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3768701553344727, | |
| "rewards/margins": 1.0476689338684082, | |
| "rewards/rejected": -2.424539089202881, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 15.145805829963615, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.4627432823181152, | |
| "logits/rejected": -2.4076457023620605, | |
| "logps/chosen": -1.2042968273162842, | |
| "logps/rejected": -2.078500270843506, | |
| "loss": -0.1185, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2042968273162842, | |
| "rewards/margins": 0.8742035031318665, | |
| "rewards/rejected": -2.078500270843506, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 10.783814010063464, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.5319347381591797, | |
| "logits/rejected": -2.5071659088134766, | |
| "logps/chosen": -1.5030553340911865, | |
| "logps/rejected": -2.248628616333008, | |
| "loss": -0.1178, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.5030553340911865, | |
| "rewards/margins": 0.7455734014511108, | |
| "rewards/rejected": -2.248628616333008, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 17.35265824839872, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.629790782928467, | |
| "logits/rejected": -2.5811970233917236, | |
| "logps/chosen": -1.1765772104263306, | |
| "logps/rejected": -2.45619535446167, | |
| "loss": -0.1283, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1765772104263306, | |
| "rewards/margins": 1.2796180248260498, | |
| "rewards/rejected": -2.45619535446167, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 21.631571252123923, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.559217929840088, | |
| "logits/rejected": -2.4746146202087402, | |
| "logps/chosen": -1.4234548807144165, | |
| "logps/rejected": -2.4785306453704834, | |
| "loss": -0.1032, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.4234548807144165, | |
| "rewards/margins": 1.055075764656067, | |
| "rewards/rejected": -2.4785306453704834, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 10.665899447598967, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.662745237350464, | |
| "logits/rejected": -2.6075425148010254, | |
| "logps/chosen": -1.1387856006622314, | |
| "logps/rejected": -2.129990577697754, | |
| "loss": -0.1171, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.1387856006622314, | |
| "rewards/margins": 0.9912050366401672, | |
| "rewards/rejected": -2.129990577697754, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 30.24510546028332, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.5815982818603516, | |
| "logits/rejected": -2.5034749507904053, | |
| "logps/chosen": -1.3083416223526, | |
| "logps/rejected": -2.3620128631591797, | |
| "loss": -0.1357, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.3083416223526, | |
| "rewards/margins": 1.0536713600158691, | |
| "rewards/rejected": -2.3620128631591797, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 36.43768779189584, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.4912540912628174, | |
| "logits/rejected": -2.503256320953369, | |
| "logps/chosen": -1.527940034866333, | |
| "logps/rejected": -2.379206895828247, | |
| "loss": -0.1328, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.527940034866333, | |
| "rewards/margins": 0.8512667417526245, | |
| "rewards/rejected": -2.379206895828247, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 21.411949693790206, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.429790735244751, | |
| "logits/rejected": -2.451469898223877, | |
| "logps/chosen": -1.8636459112167358, | |
| "logps/rejected": -2.370575428009033, | |
| "loss": -0.1231, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.8636459112167358, | |
| "rewards/margins": 0.5069290399551392, | |
| "rewards/rejected": -2.370575428009033, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 26.28588094272757, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.484380006790161, | |
| "logits/rejected": -2.3554978370666504, | |
| "logps/chosen": -1.374985933303833, | |
| "logps/rejected": -3.2982125282287598, | |
| "loss": -0.1605, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.374985933303833, | |
| "rewards/margins": 1.9232263565063477, | |
| "rewards/rejected": -3.2982125282287598, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 15.64313913527933, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.5084352493286133, | |
| "logits/rejected": -2.379485607147217, | |
| "logps/chosen": -1.2756072282791138, | |
| "logps/rejected": -2.862361431121826, | |
| "loss": -0.1283, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2756072282791138, | |
| "rewards/margins": 1.586754560470581, | |
| "rewards/rejected": -2.862361431121826, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 31.160163808849045, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.464437484741211, | |
| "logits/rejected": -2.4234752655029297, | |
| "logps/chosen": -1.7923996448516846, | |
| "logps/rejected": -3.0464653968811035, | |
| "loss": -0.1107, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.7923996448516846, | |
| "rewards/margins": 1.2540655136108398, | |
| "rewards/rejected": -3.0464653968811035, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 14.525714466074616, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.3748879432678223, | |
| "logits/rejected": -2.277477741241455, | |
| "logps/chosen": -1.8306829929351807, | |
| "logps/rejected": -2.5916666984558105, | |
| "loss": -0.1171, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.8306829929351807, | |
| "rewards/margins": 0.7609835863113403, | |
| "rewards/rejected": -2.5916666984558105, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 12.205724246076834, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.5410079956054688, | |
| "logits/rejected": -2.368896484375, | |
| "logps/chosen": -1.3447208404541016, | |
| "logps/rejected": -2.198944091796875, | |
| "loss": -0.1286, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3447208404541016, | |
| "rewards/margins": 0.8542232513427734, | |
| "rewards/rejected": -2.198944091796875, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 13.048722996544631, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.3360071182250977, | |
| "logits/rejected": -2.238037347793579, | |
| "logps/chosen": -1.3523575067520142, | |
| "logps/rejected": -2.568472146987915, | |
| "loss": -0.1212, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.3523575067520142, | |
| "rewards/margins": 1.2161142826080322, | |
| "rewards/rejected": -2.568472146987915, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 14.965881571164392, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.4726943969726562, | |
| "logits/rejected": -2.4096648693084717, | |
| "logps/chosen": -1.4510114192962646, | |
| "logps/rejected": -2.0148894786834717, | |
| "loss": -0.1379, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.4510114192962646, | |
| "rewards/margins": 0.563878059387207, | |
| "rewards/rejected": -2.0148894786834717, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 14.196438929994413, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.434021472930908, | |
| "logits/rejected": -2.2209062576293945, | |
| "logps/chosen": -1.2371381521224976, | |
| "logps/rejected": -2.5744917392730713, | |
| "loss": -0.1344, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.2371381521224976, | |
| "rewards/margins": 1.3373535871505737, | |
| "rewards/rejected": -2.5744917392730713, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 13.34734425640694, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.482234477996826, | |
| "logits/rejected": -2.3337721824645996, | |
| "logps/chosen": -1.7002273797988892, | |
| "logps/rejected": -2.32104754447937, | |
| "loss": -0.1352, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.7002273797988892, | |
| "rewards/margins": 0.6208202838897705, | |
| "rewards/rejected": -2.32104754447937, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 54.856648124460555, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.4059951305389404, | |
| "logits/rejected": -2.2941792011260986, | |
| "logps/chosen": -1.7276633977890015, | |
| "logps/rejected": -2.791877031326294, | |
| "loss": -0.1395, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.7276633977890015, | |
| "rewards/margins": 1.0642132759094238, | |
| "rewards/rejected": -2.791877031326294, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 16.338313868801983, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.3474369049072266, | |
| "logits/rejected": -2.2156565189361572, | |
| "logps/chosen": -1.581926941871643, | |
| "logps/rejected": -3.4111123085021973, | |
| "loss": -0.1119, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.581926941871643, | |
| "rewards/margins": 1.8291854858398438, | |
| "rewards/rejected": -3.4111123085021973, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 27.599650373019013, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.424649238586426, | |
| "logits/rejected": -2.313764810562134, | |
| "logps/chosen": -1.4561196565628052, | |
| "logps/rejected": -2.5569052696228027, | |
| "loss": -0.1255, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4561196565628052, | |
| "rewards/margins": 1.1007858514785767, | |
| "rewards/rejected": -2.5569052696228027, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 13.71174939782389, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.182265520095825, | |
| "logits/rejected": -2.240220069885254, | |
| "logps/chosen": -2.133391857147217, | |
| "logps/rejected": -2.705057382583618, | |
| "loss": -0.1229, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.133391857147217, | |
| "rewards/margins": 0.571665346622467, | |
| "rewards/rejected": -2.705057382583618, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 22.094952760541535, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -2.351863384246826, | |
| "logits/rejected": -2.2399449348449707, | |
| "logps/chosen": -1.3272682428359985, | |
| "logps/rejected": -3.206008195877075, | |
| "loss": -0.1347, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3272682428359985, | |
| "rewards/margins": 1.8787400722503662, | |
| "rewards/rejected": -3.206008195877075, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 14.410515050019907, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.3990814685821533, | |
| "logits/rejected": -2.256584644317627, | |
| "logps/chosen": -1.4144983291625977, | |
| "logps/rejected": -2.471238851547241, | |
| "loss": -0.1205, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.4144983291625977, | |
| "rewards/margins": 1.056740641593933, | |
| "rewards/rejected": -2.471238851547241, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 17.737495980302032, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.3040690422058105, | |
| "logits/rejected": -1.9883592128753662, | |
| "logps/chosen": -1.8120450973510742, | |
| "logps/rejected": -2.608847141265869, | |
| "loss": -0.1314, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.8120450973510742, | |
| "rewards/margins": 0.7968018651008606, | |
| "rewards/rejected": -2.608847141265869, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 147.24723390344812, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -2.2860684394836426, | |
| "logits/rejected": -2.1697564125061035, | |
| "logps/chosen": -1.6789791584014893, | |
| "logps/rejected": -3.0929505825042725, | |
| "loss": -0.1132, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.6789791584014893, | |
| "rewards/margins": 1.413971185684204, | |
| "rewards/rejected": -3.0929505825042725, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 25.64073118103101, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -2.218724012374878, | |
| "logits/rejected": -2.1296398639678955, | |
| "logps/chosen": -1.6159321069717407, | |
| "logps/rejected": -2.5099380016326904, | |
| "loss": -0.1222, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.6159321069717407, | |
| "rewards/margins": 0.8940057754516602, | |
| "rewards/rejected": -2.5099380016326904, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 12.902928067743126, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.3579747676849365, | |
| "logits/rejected": -2.1200146675109863, | |
| "logps/chosen": -1.5285216569900513, | |
| "logps/rejected": -3.1276838779449463, | |
| "loss": -0.145, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.5285216569900513, | |
| "rewards/margins": 1.5991623401641846, | |
| "rewards/rejected": -3.1276838779449463, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 15.785906225773893, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.407064914703369, | |
| "logits/rejected": -2.30210542678833, | |
| "logps/chosen": -1.3575342893600464, | |
| "logps/rejected": -2.2857096195220947, | |
| "loss": -0.1468, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.3575342893600464, | |
| "rewards/margins": 0.9281753301620483, | |
| "rewards/rejected": -2.2857096195220947, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 66.73212890394835, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -2.2724781036376953, | |
| "logits/rejected": -2.2895565032958984, | |
| "logps/chosen": -1.549403429031372, | |
| "logps/rejected": -2.9475514888763428, | |
| "loss": -0.1203, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.549403429031372, | |
| "rewards/margins": 1.3981479406356812, | |
| "rewards/rejected": -2.9475514888763428, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 38.70646418643775, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.399785041809082, | |
| "logits/rejected": -2.3085193634033203, | |
| "logps/chosen": -1.3037949800491333, | |
| "logps/rejected": -2.2208690643310547, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.3037949800491333, | |
| "rewards/margins": 0.9170739054679871, | |
| "rewards/rejected": -2.2208690643310547, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 85.36661297405541, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.350160598754883, | |
| "logits/rejected": -2.304274559020996, | |
| "logps/chosen": -1.328881859779358, | |
| "logps/rejected": -2.426609516143799, | |
| "loss": -0.1355, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.328881859779358, | |
| "rewards/margins": 1.0977277755737305, | |
| "rewards/rejected": -2.426609516143799, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 21.914030829763547, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.41226863861084, | |
| "logits/rejected": -2.258676767349243, | |
| "logps/chosen": -1.3917608261108398, | |
| "logps/rejected": -2.255201578140259, | |
| "loss": -0.1082, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.3917608261108398, | |
| "rewards/margins": 0.8634408712387085, | |
| "rewards/rejected": -2.255201578140259, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 11.128742517489252, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.3416173458099365, | |
| "logits/rejected": -2.183840274810791, | |
| "logps/chosen": -1.4819704294204712, | |
| "logps/rejected": -2.7424230575561523, | |
| "loss": -0.1199, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4819704294204712, | |
| "rewards/margins": 1.2604527473449707, | |
| "rewards/rejected": -2.7424230575561523, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 13.499128871751621, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.4478988647460938, | |
| "logits/rejected": -2.2530007362365723, | |
| "logps/chosen": -1.503173589706421, | |
| "logps/rejected": -2.4205775260925293, | |
| "loss": -0.1196, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.503173589706421, | |
| "rewards/margins": 0.9174041748046875, | |
| "rewards/rejected": -2.4205775260925293, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 46.00477049565933, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.428011894226074, | |
| "logits/rejected": -2.195507526397705, | |
| "logps/chosen": -1.488608956336975, | |
| "logps/rejected": -3.2226758003234863, | |
| "loss": -0.1329, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.488608956336975, | |
| "rewards/margins": 1.7340667247772217, | |
| "rewards/rejected": -3.2226758003234863, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 15.147104463873568, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -2.3303403854370117, | |
| "logits/rejected": -2.2017440795898438, | |
| "logps/chosen": -1.4811475276947021, | |
| "logps/rejected": -3.1947004795074463, | |
| "loss": -0.1345, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.4811475276947021, | |
| "rewards/margins": 1.7135528326034546, | |
| "rewards/rejected": -3.1947004795074463, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 32.54043568171796, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -2.3127968311309814, | |
| "logits/rejected": -2.2588541507720947, | |
| "logps/chosen": -1.7812035083770752, | |
| "logps/rejected": -2.363581418991089, | |
| "loss": -0.1144, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.7812035083770752, | |
| "rewards/margins": 0.5823777318000793, | |
| "rewards/rejected": -2.363581418991089, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 16.364361772449453, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.2632575035095215, | |
| "logits/rejected": -2.1989002227783203, | |
| "logps/chosen": -1.3589812517166138, | |
| "logps/rejected": -2.665706157684326, | |
| "loss": -0.1402, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3589812517166138, | |
| "rewards/margins": 1.3067247867584229, | |
| "rewards/rejected": -2.665706157684326, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 24.459898730148318, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.4754738807678223, | |
| "logits/rejected": -2.2798328399658203, | |
| "logps/chosen": -1.1175788640975952, | |
| "logps/rejected": -2.699981212615967, | |
| "loss": -0.1488, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1175788640975952, | |
| "rewards/margins": 1.5824021100997925, | |
| "rewards/rejected": -2.699981212615967, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 20.15635024127512, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.3345370292663574, | |
| "logits/rejected": -2.0842413902282715, | |
| "logps/chosen": -1.416595458984375, | |
| "logps/rejected": -2.772968053817749, | |
| "loss": -0.1388, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.416595458984375, | |
| "rewards/margins": 1.3563728332519531, | |
| "rewards/rejected": -2.772968053817749, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 25.1930677516597, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -2.2844507694244385, | |
| "logits/rejected": -2.2151055335998535, | |
| "logps/chosen": -1.5157394409179688, | |
| "logps/rejected": -2.6189606189727783, | |
| "loss": -0.1335, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.5157394409179688, | |
| "rewards/margins": 1.1032211780548096, | |
| "rewards/rejected": -2.6189606189727783, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 94.3340138658481, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.3165175914764404, | |
| "logits/rejected": -2.216142177581787, | |
| "logps/chosen": -1.6333303451538086, | |
| "logps/rejected": -2.6600871086120605, | |
| "loss": -0.1365, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.6333303451538086, | |
| "rewards/margins": 1.0267568826675415, | |
| "rewards/rejected": -2.6600871086120605, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 37.68445074958579, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.173781394958496, | |
| "logits/rejected": -2.0770103931427, | |
| "logps/chosen": -1.7893623113632202, | |
| "logps/rejected": -2.782428026199341, | |
| "loss": -0.1362, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.7893623113632202, | |
| "rewards/margins": 0.9930654764175415, | |
| "rewards/rejected": -2.782428026199341, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 65.20343901413754, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.4339003562927246, | |
| "logits/rejected": -2.235538959503174, | |
| "logps/chosen": -1.1971713304519653, | |
| "logps/rejected": -2.028782367706299, | |
| "loss": -0.1358, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1971713304519653, | |
| "rewards/margins": 0.8316110372543335, | |
| "rewards/rejected": -2.028782367706299, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 29.43315569875579, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.406641721725464, | |
| "logits/rejected": -2.2101848125457764, | |
| "logps/chosen": -1.7016384601593018, | |
| "logps/rejected": -2.7031149864196777, | |
| "loss": -0.1135, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.7016384601593018, | |
| "rewards/margins": 1.0014766454696655, | |
| "rewards/rejected": -2.7031149864196777, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 58.96075147402077, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.5120296478271484, | |
| "logits/rejected": -2.0890145301818848, | |
| "logps/chosen": -1.2953853607177734, | |
| "logps/rejected": -3.554316759109497, | |
| "loss": -0.1399, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.2953853607177734, | |
| "rewards/margins": 2.2589316368103027, | |
| "rewards/rejected": -3.554316759109497, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 34.51709660208177, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.295931339263916, | |
| "logits/rejected": -2.02848219871521, | |
| "logps/chosen": -1.3973679542541504, | |
| "logps/rejected": -3.212092638015747, | |
| "loss": -0.145, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.3973679542541504, | |
| "rewards/margins": 1.814724326133728, | |
| "rewards/rejected": -3.212092638015747, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 11.228558025822675, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.295358180999756, | |
| "logits/rejected": -2.151076316833496, | |
| "logps/chosen": -1.4753532409667969, | |
| "logps/rejected": -2.6792643070220947, | |
| "loss": -0.131, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.4753532409667969, | |
| "rewards/margins": 1.2039110660552979, | |
| "rewards/rejected": -2.6792643070220947, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 15.547524058275151, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.3587679862976074, | |
| "logits/rejected": -2.1848464012145996, | |
| "logps/chosen": -1.336055874824524, | |
| "logps/rejected": -2.380079746246338, | |
| "loss": -0.1371, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.336055874824524, | |
| "rewards/margins": 1.044023871421814, | |
| "rewards/rejected": -2.380079746246338, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": -0.1067421951134, | |
| "train_runtime": 20026.0763, | |
| "train_samples_per_second": 3.053, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |