| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.526447931526391, | |
| "eval_steps": 500, | |
| "global_step": 79326, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.028530670470756064, | |
| "grad_norm": 4.224736213684082, | |
| "learning_rate": 9.936968963517636e-05, | |
| "loss": 7.9606, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05706134094151213, | |
| "grad_norm": 3.775930881500244, | |
| "learning_rate": 9.873937927035273e-05, | |
| "loss": 7.4609, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.08559201141226819, | |
| "grad_norm": 4.54003381729126, | |
| "learning_rate": 9.810906890552909e-05, | |
| "loss": 7.2264, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.11412268188302425, | |
| "grad_norm": 4.762158393859863, | |
| "learning_rate": 9.747875854070544e-05, | |
| "loss": 7.1154, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.14265335235378032, | |
| "grad_norm": 4.3742995262146, | |
| "learning_rate": 9.684844817588181e-05, | |
| "loss": 7.0396, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.17118402282453637, | |
| "grad_norm": 6.319996356964111, | |
| "learning_rate": 9.621813781105817e-05, | |
| "loss": 6.9091, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.19971469329529243, | |
| "grad_norm": 5.134556770324707, | |
| "learning_rate": 9.558782744623454e-05, | |
| "loss": 6.8436, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.2282453637660485, | |
| "grad_norm": 5.821471691131592, | |
| "learning_rate": 9.49575170814109e-05, | |
| "loss": 6.7296, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.25677603423680456, | |
| "grad_norm": 5.649785041809082, | |
| "learning_rate": 9.432720671658725e-05, | |
| "loss": 6.6347, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.28530670470756064, | |
| "grad_norm": 4.85307502746582, | |
| "learning_rate": 9.369689635176361e-05, | |
| "loss": 6.5734, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.31383737517831667, | |
| "grad_norm": 4.924481391906738, | |
| "learning_rate": 9.306658598693997e-05, | |
| "loss": 6.4956, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.34236804564907275, | |
| "grad_norm": 5.568727970123291, | |
| "learning_rate": 9.243627562211632e-05, | |
| "loss": 6.4667, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.37089871611982883, | |
| "grad_norm": 5.307117938995361, | |
| "learning_rate": 9.18059652572927e-05, | |
| "loss": 6.4108, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.39942938659058486, | |
| "grad_norm": 5.159631729125977, | |
| "learning_rate": 9.117565489246905e-05, | |
| "loss": 6.3248, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.42796005706134094, | |
| "grad_norm": 5.529886722564697, | |
| "learning_rate": 9.054534452764542e-05, | |
| "loss": 6.3069, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.456490727532097, | |
| "grad_norm": 9.25552749633789, | |
| "learning_rate": 8.991503416282178e-05, | |
| "loss": 6.2188, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.48502139800285304, | |
| "grad_norm": 7.253251075744629, | |
| "learning_rate": 8.928472379799813e-05, | |
| "loss": 6.0356, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5135520684736091, | |
| "grad_norm": 8.699528694152832, | |
| "learning_rate": 8.86544134331745e-05, | |
| "loss": 6.0202, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.5420827389443652, | |
| "grad_norm": 7.144530296325684, | |
| "learning_rate": 8.802410306835086e-05, | |
| "loss": 5.8857, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.5706134094151213, | |
| "grad_norm": 7.254436492919922, | |
| "learning_rate": 8.739379270352722e-05, | |
| "loss": 5.8363, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.5991440798858774, | |
| "grad_norm": 7.687765121459961, | |
| "learning_rate": 8.676348233870359e-05, | |
| "loss": 5.7397, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.6276747503566333, | |
| "grad_norm": 6.547056674957275, | |
| "learning_rate": 8.613317197387994e-05, | |
| "loss": 5.6764, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.6562054208273894, | |
| "grad_norm": 9.949172973632812, | |
| "learning_rate": 8.55028616090563e-05, | |
| "loss": 5.6305, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.6847360912981455, | |
| "grad_norm": 8.107893943786621, | |
| "learning_rate": 8.487255124423267e-05, | |
| "loss": 5.5956, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.7132667617689016, | |
| "grad_norm": 6.340846538543701, | |
| "learning_rate": 8.424224087940903e-05, | |
| "loss": 5.5119, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.7417974322396577, | |
| "grad_norm": 7.98532247543335, | |
| "learning_rate": 8.361193051458538e-05, | |
| "loss": 5.4685, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.7703281027104137, | |
| "grad_norm": 8.349475860595703, | |
| "learning_rate": 8.298162014976175e-05, | |
| "loss": 5.4685, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.7988587731811697, | |
| "grad_norm": 8.270821571350098, | |
| "learning_rate": 8.23513097849381e-05, | |
| "loss": 5.4529, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.8273894436519258, | |
| "grad_norm": 7.967436790466309, | |
| "learning_rate": 8.172099942011447e-05, | |
| "loss": 5.369, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.8559201141226819, | |
| "grad_norm": 6.520368576049805, | |
| "learning_rate": 8.109068905529082e-05, | |
| "loss": 5.3417, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.884450784593438, | |
| "grad_norm": 7.0926713943481445, | |
| "learning_rate": 8.046037869046718e-05, | |
| "loss": 5.3051, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.912981455064194, | |
| "grad_norm": 7.306639194488525, | |
| "learning_rate": 7.983006832564355e-05, | |
| "loss": 5.275, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.9415121255349501, | |
| "grad_norm": 7.494438171386719, | |
| "learning_rate": 7.919975796081991e-05, | |
| "loss": 5.2383, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.9700427960057061, | |
| "grad_norm": 7.95282506942749, | |
| "learning_rate": 7.856944759599628e-05, | |
| "loss": 5.2275, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.9985734664764622, | |
| "grad_norm": 8.143848419189453, | |
| "learning_rate": 7.793913723117263e-05, | |
| "loss": 5.1984, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 5.088438034057617, | |
| "eval_runtime": 60.5124, | |
| "eval_samples_per_second": 817.683, | |
| "eval_steps_per_second": 51.113, | |
| "step": 17525 | |
| }, | |
| { | |
| "epoch": 1.0271041369472182, | |
| "grad_norm": 7.758360862731934, | |
| "learning_rate": 7.730882686634899e-05, | |
| "loss": 5.1319, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.0556348074179742, | |
| "grad_norm": 6.7139177322387695, | |
| "learning_rate": 7.667851650152536e-05, | |
| "loss": 5.1087, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.0841654778887304, | |
| "grad_norm": 6.584017753601074, | |
| "learning_rate": 7.604820613670172e-05, | |
| "loss": 5.0938, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.1126961483594864, | |
| "grad_norm": 7.792596817016602, | |
| "learning_rate": 7.541789577187807e-05, | |
| "loss": 5.0448, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.1412268188302426, | |
| "grad_norm": 7.243467807769775, | |
| "learning_rate": 7.478758540705444e-05, | |
| "loss": 5.0423, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.1697574893009985, | |
| "grad_norm": 8.1492338180542, | |
| "learning_rate": 7.41572750422308e-05, | |
| "loss": 4.9913, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.1982881597717547, | |
| "grad_norm": 7.754183292388916, | |
| "learning_rate": 7.352696467740716e-05, | |
| "loss": 4.9983, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.2268188302425107, | |
| "grad_norm": 7.245093822479248, | |
| "learning_rate": 7.289665431258353e-05, | |
| "loss": 4.9729, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.2553495007132667, | |
| "grad_norm": 7.720768451690674, | |
| "learning_rate": 7.226634394775988e-05, | |
| "loss": 4.9499, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.2838801711840229, | |
| "grad_norm": 7.034431457519531, | |
| "learning_rate": 7.163603358293624e-05, | |
| "loss": 4.9368, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.3124108416547788, | |
| "grad_norm": 9.341329574584961, | |
| "learning_rate": 7.100572321811261e-05, | |
| "loss": 4.8745, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.340941512125535, | |
| "grad_norm": 6.874212265014648, | |
| "learning_rate": 7.037541285328895e-05, | |
| "loss": 4.877, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.369472182596291, | |
| "grad_norm": 8.832662582397461, | |
| "learning_rate": 6.974510248846532e-05, | |
| "loss": 4.8839, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.3980028530670472, | |
| "grad_norm": 6.884407997131348, | |
| "learning_rate": 6.911479212364168e-05, | |
| "loss": 4.8456, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.4265335235378032, | |
| "grad_norm": 8.421597480773926, | |
| "learning_rate": 6.848448175881804e-05, | |
| "loss": 4.8491, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.4550641940085591, | |
| "grad_norm": 8.932446479797363, | |
| "learning_rate": 6.78541713939944e-05, | |
| "loss": 4.7936, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.4835948644793153, | |
| "grad_norm": 9.981544494628906, | |
| "learning_rate": 6.722386102917076e-05, | |
| "loss": 4.816, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.5121255349500713, | |
| "grad_norm": 7.8625922203063965, | |
| "learning_rate": 6.659355066434713e-05, | |
| "loss": 4.7983, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.5406562054208273, | |
| "grad_norm": 9.113426208496094, | |
| "learning_rate": 6.596324029952349e-05, | |
| "loss": 4.7438, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.5691868758915835, | |
| "grad_norm": 8.4766206741333, | |
| "learning_rate": 6.533292993469985e-05, | |
| "loss": 4.7544, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.5977175463623396, | |
| "grad_norm": 9.448740005493164, | |
| "learning_rate": 6.470261956987622e-05, | |
| "loss": 4.7244, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.6262482168330956, | |
| "grad_norm": 7.334268093109131, | |
| "learning_rate": 6.407230920505257e-05, | |
| "loss": 4.7225, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.6547788873038516, | |
| "grad_norm": 8.931607246398926, | |
| "learning_rate": 6.344199884022893e-05, | |
| "loss": 4.6495, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.6833095577746078, | |
| "grad_norm": 7.774843215942383, | |
| "learning_rate": 6.28116884754053e-05, | |
| "loss": 4.7006, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.7118402282453637, | |
| "grad_norm": 7.911451816558838, | |
| "learning_rate": 6.218137811058166e-05, | |
| "loss": 4.6525, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.7403708987161197, | |
| "grad_norm": 7.005200386047363, | |
| "learning_rate": 6.155106774575801e-05, | |
| "loss": 4.6765, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.768901569186876, | |
| "grad_norm": 8.889507293701172, | |
| "learning_rate": 6.0920757380934376e-05, | |
| "loss": 4.6397, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.797432239657632, | |
| "grad_norm": 8.807517051696777, | |
| "learning_rate": 6.029044701611074e-05, | |
| "loss": 4.6291, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.825962910128388, | |
| "grad_norm": 10.036596298217773, | |
| "learning_rate": 5.96601366512871e-05, | |
| "loss": 4.6033, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.854493580599144, | |
| "grad_norm": 8.668290138244629, | |
| "learning_rate": 5.902982628646345e-05, | |
| "loss": 4.617, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.8830242510699002, | |
| "grad_norm": 7.631191253662109, | |
| "learning_rate": 5.8399515921639816e-05, | |
| "loss": 4.591, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.9115549215406562, | |
| "grad_norm": 7.901882171630859, | |
| "learning_rate": 5.776920555681617e-05, | |
| "loss": 4.593, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.9400855920114122, | |
| "grad_norm": 8.088603973388672, | |
| "learning_rate": 5.7138895191992536e-05, | |
| "loss": 4.575, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.9686162624821684, | |
| "grad_norm": 13.519835472106934, | |
| "learning_rate": 5.65085848271689e-05, | |
| "loss": 4.5674, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.9971469329529246, | |
| "grad_norm": 7.497794151306152, | |
| "learning_rate": 5.587827446234526e-05, | |
| "loss": 4.5896, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4.483283996582031, | |
| "eval_runtime": 55.5401, | |
| "eval_samples_per_second": 890.889, | |
| "eval_steps_per_second": 55.69, | |
| "step": 35050 | |
| }, | |
| { | |
| "epoch": 2.0256776034236803, | |
| "grad_norm": 9.086140632629395, | |
| "learning_rate": 5.524796409752162e-05, | |
| "loss": 4.4731, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 2.0542082738944365, | |
| "grad_norm": 7.284142017364502, | |
| "learning_rate": 5.461765373269798e-05, | |
| "loss": 4.5119, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 2.0827389443651927, | |
| "grad_norm": 9.4158296585083, | |
| "learning_rate": 5.3987343367874346e-05, | |
| "loss": 4.4564, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 2.1112696148359484, | |
| "grad_norm": 10.107216835021973, | |
| "learning_rate": 5.33570330030507e-05, | |
| "loss": 4.5029, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.1398002853067046, | |
| "grad_norm": 8.506832122802734, | |
| "learning_rate": 5.2726722638227066e-05, | |
| "loss": 4.4487, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 2.168330955777461, | |
| "grad_norm": 10.495549201965332, | |
| "learning_rate": 5.209641227340343e-05, | |
| "loss": 4.433, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.196861626248217, | |
| "grad_norm": 8.59913444519043, | |
| "learning_rate": 5.146610190857979e-05, | |
| "loss": 4.4377, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 2.2253922967189728, | |
| "grad_norm": 7.879130840301514, | |
| "learning_rate": 5.083579154375615e-05, | |
| "loss": 4.3888, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.253922967189729, | |
| "grad_norm": 8.134923934936523, | |
| "learning_rate": 5.020548117893251e-05, | |
| "loss": 4.4215, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 2.282453637660485, | |
| "grad_norm": 8.942688941955566, | |
| "learning_rate": 4.957517081410887e-05, | |
| "loss": 4.4249, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.310984308131241, | |
| "grad_norm": 12.744963645935059, | |
| "learning_rate": 4.894486044928523e-05, | |
| "loss": 4.3957, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 2.339514978601997, | |
| "grad_norm": 10.202946662902832, | |
| "learning_rate": 4.831455008446159e-05, | |
| "loss": 4.4239, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 2.3680456490727533, | |
| "grad_norm": 7.9722418785095215, | |
| "learning_rate": 4.768423971963795e-05, | |
| "loss": 4.4196, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 2.3965763195435095, | |
| "grad_norm": 8.571127891540527, | |
| "learning_rate": 4.7053929354814315e-05, | |
| "loss": 4.3274, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.425106990014265, | |
| "grad_norm": 9.048084259033203, | |
| "learning_rate": 4.642361898999067e-05, | |
| "loss": 4.3422, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 2.4536376604850214, | |
| "grad_norm": 9.029339790344238, | |
| "learning_rate": 4.5793308625167035e-05, | |
| "loss": 4.3474, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 2.4821683309557776, | |
| "grad_norm": 8.281103134155273, | |
| "learning_rate": 4.51629982603434e-05, | |
| "loss": 4.3849, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 2.5106990014265333, | |
| "grad_norm": 7.619591236114502, | |
| "learning_rate": 4.4532687895519755e-05, | |
| "loss": 4.3154, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.5392296718972895, | |
| "grad_norm": 10.685304641723633, | |
| "learning_rate": 4.390237753069611e-05, | |
| "loss": 4.2803, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 2.5677603423680457, | |
| "grad_norm": 8.622846603393555, | |
| "learning_rate": 4.3272067165872475e-05, | |
| "loss": 4.2911, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.596291012838802, | |
| "grad_norm": 10.364981651306152, | |
| "learning_rate": 4.264175680104884e-05, | |
| "loss": 4.3172, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 2.6248216833095577, | |
| "grad_norm": 6.996392726898193, | |
| "learning_rate": 4.20114464362252e-05, | |
| "loss": 4.3123, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 2.653352353780314, | |
| "grad_norm": 9.445304870605469, | |
| "learning_rate": 4.138113607140156e-05, | |
| "loss": 4.2621, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 2.68188302425107, | |
| "grad_norm": 9.185279846191406, | |
| "learning_rate": 4.075082570657792e-05, | |
| "loss": 4.2983, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 2.710413694721826, | |
| "grad_norm": 7.654629230499268, | |
| "learning_rate": 4.0120515341754285e-05, | |
| "loss": 4.2994, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 2.738944365192582, | |
| "grad_norm": 8.687335014343262, | |
| "learning_rate": 3.949020497693065e-05, | |
| "loss": 4.2667, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.767475035663338, | |
| "grad_norm": 9.457321166992188, | |
| "learning_rate": 3.8859894612107e-05, | |
| "loss": 4.2709, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 2.7960057061340944, | |
| "grad_norm": 9.891905784606934, | |
| "learning_rate": 3.822958424728336e-05, | |
| "loss": 4.217, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 2.82453637660485, | |
| "grad_norm": 7.958982467651367, | |
| "learning_rate": 3.7599273882459725e-05, | |
| "loss": 4.219, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 2.8530670470756063, | |
| "grad_norm": 9.362072944641113, | |
| "learning_rate": 3.696896351763609e-05, | |
| "loss": 4.2212, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.881597717546362, | |
| "grad_norm": 9.81021499633789, | |
| "learning_rate": 3.6338653152812445e-05, | |
| "loss": 4.1786, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 2.9101283880171183, | |
| "grad_norm": 9.377098083496094, | |
| "learning_rate": 3.570834278798881e-05, | |
| "loss": 4.1874, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 2.9386590584878745, | |
| "grad_norm": 7.9157562255859375, | |
| "learning_rate": 3.507803242316517e-05, | |
| "loss": 4.2021, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 2.9671897289586306, | |
| "grad_norm": 9.986918449401855, | |
| "learning_rate": 3.444772205834153e-05, | |
| "loss": 4.2068, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.995720399429387, | |
| "grad_norm": 9.590192794799805, | |
| "learning_rate": 3.381741169351789e-05, | |
| "loss": 4.2007, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 4.12980842590332, | |
| "eval_runtime": 54.7837, | |
| "eval_samples_per_second": 903.188, | |
| "eval_steps_per_second": 56.458, | |
| "step": 52575 | |
| }, | |
| { | |
| "epoch": 3.0242510699001426, | |
| "grad_norm": 10.125175476074219, | |
| "learning_rate": 3.3187101328694254e-05, | |
| "loss": 4.1689, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 3.0527817403708988, | |
| "grad_norm": 9.876838684082031, | |
| "learning_rate": 3.255679096387061e-05, | |
| "loss": 4.1606, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 3.081312410841655, | |
| "grad_norm": 8.361190795898438, | |
| "learning_rate": 3.192648059904697e-05, | |
| "loss": 4.1792, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 3.1098430813124107, | |
| "grad_norm": 11.448074340820312, | |
| "learning_rate": 3.129617023422333e-05, | |
| "loss": 4.1463, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 3.138373751783167, | |
| "grad_norm": 9.612030982971191, | |
| "learning_rate": 3.0665859869399694e-05, | |
| "loss": 4.1013, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.166904422253923, | |
| "grad_norm": 8.74341106414795, | |
| "learning_rate": 3.0035549504576054e-05, | |
| "loss": 4.1097, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 3.195435092724679, | |
| "grad_norm": 8.951421737670898, | |
| "learning_rate": 2.9405239139752418e-05, | |
| "loss": 4.1154, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 3.223965763195435, | |
| "grad_norm": 9.474740028381348, | |
| "learning_rate": 2.8774928774928778e-05, | |
| "loss": 4.1132, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 3.2524964336661912, | |
| "grad_norm": 10.41540241241455, | |
| "learning_rate": 2.8144618410105137e-05, | |
| "loss": 4.1416, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 3.281027104136947, | |
| "grad_norm": 10.558377265930176, | |
| "learning_rate": 2.75143080452815e-05, | |
| "loss": 4.1209, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 3.309557774607703, | |
| "grad_norm": 8.718255996704102, | |
| "learning_rate": 2.6883997680457857e-05, | |
| "loss": 4.0549, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 3.3380884450784594, | |
| "grad_norm": 8.449244499206543, | |
| "learning_rate": 2.6253687315634217e-05, | |
| "loss": 4.1046, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 3.3666191155492156, | |
| "grad_norm": 10.529504776000977, | |
| "learning_rate": 2.562337695081058e-05, | |
| "loss": 4.1058, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 3.3951497860199713, | |
| "grad_norm": 8.345918655395508, | |
| "learning_rate": 2.499306658598694e-05, | |
| "loss": 4.0668, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 3.4236804564907275, | |
| "grad_norm": 10.951416015625, | |
| "learning_rate": 2.43627562211633e-05, | |
| "loss": 4.1145, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 3.4522111269614837, | |
| "grad_norm": 8.29075813293457, | |
| "learning_rate": 2.3732445856339664e-05, | |
| "loss": 4.0765, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 3.4807417974322394, | |
| "grad_norm": 9.023518562316895, | |
| "learning_rate": 2.3102135491516024e-05, | |
| "loss": 4.0608, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 3.5092724679029956, | |
| "grad_norm": 10.263666152954102, | |
| "learning_rate": 2.2471825126692384e-05, | |
| "loss": 4.0361, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 3.537803138373752, | |
| "grad_norm": 10.308072090148926, | |
| "learning_rate": 2.1841514761868744e-05, | |
| "loss": 4.0801, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 3.566333808844508, | |
| "grad_norm": 11.750104904174805, | |
| "learning_rate": 2.1211204397045107e-05, | |
| "loss": 4.0166, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 3.5948644793152638, | |
| "grad_norm": 9.711214065551758, | |
| "learning_rate": 2.0580894032221467e-05, | |
| "loss": 4.1, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 3.62339514978602, | |
| "grad_norm": 9.461030006408691, | |
| "learning_rate": 1.9950583667397827e-05, | |
| "loss": 4.0546, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 3.651925820256776, | |
| "grad_norm": 9.695231437683105, | |
| "learning_rate": 1.9320273302574187e-05, | |
| "loss": 3.9823, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 3.680456490727532, | |
| "grad_norm": 8.181650161743164, | |
| "learning_rate": 1.868996293775055e-05, | |
| "loss": 4.0467, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 3.708987161198288, | |
| "grad_norm": 10.602164268493652, | |
| "learning_rate": 1.805965257292691e-05, | |
| "loss": 4.0006, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 3.7375178316690443, | |
| "grad_norm": 8.176977157592773, | |
| "learning_rate": 1.7429342208103274e-05, | |
| "loss": 4.0594, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 3.7660485021398005, | |
| "grad_norm": 11.314457893371582, | |
| "learning_rate": 1.679903184327963e-05, | |
| "loss": 3.9929, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 3.794579172610556, | |
| "grad_norm": 10.995957374572754, | |
| "learning_rate": 1.6168721478455993e-05, | |
| "loss": 4.0191, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 3.8231098430813124, | |
| "grad_norm": 8.929314613342285, | |
| "learning_rate": 1.5538411113632353e-05, | |
| "loss": 4.0284, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 3.8516405135520686, | |
| "grad_norm": 9.887913703918457, | |
| "learning_rate": 1.4908100748808715e-05, | |
| "loss": 3.9552, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 3.8801711840228243, | |
| "grad_norm": 12.043901443481445, | |
| "learning_rate": 1.4277790383985077e-05, | |
| "loss": 4.0139, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 3.9087018544935805, | |
| "grad_norm": 8.444652557373047, | |
| "learning_rate": 1.3647480019161435e-05, | |
| "loss": 3.9746, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 3.9372325249643367, | |
| "grad_norm": 9.180087089538574, | |
| "learning_rate": 1.3017169654337797e-05, | |
| "loss": 3.9663, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 3.965763195435093, | |
| "grad_norm": 11.611577987670898, | |
| "learning_rate": 1.2386859289514158e-05, | |
| "loss": 3.939, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 3.9942938659058487, | |
| "grad_norm": 9.529749870300293, | |
| "learning_rate": 1.1756548924690518e-05, | |
| "loss": 3.9911, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.9361355304718018, | |
| "eval_runtime": 55.3368, | |
| "eval_samples_per_second": 894.161, | |
| "eval_steps_per_second": 55.894, | |
| "step": 70100 | |
| }, | |
| { | |
| "epoch": 4.022824536376604, | |
| "grad_norm": 10.018484115600586, | |
| "learning_rate": 1.112623855986688e-05, | |
| "loss": 3.9744, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 4.051355206847361, | |
| "grad_norm": 11.023027420043945, | |
| "learning_rate": 1.049592819504324e-05, | |
| "loss": 3.9143, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 4.079885877318117, | |
| "grad_norm": 11.09890079498291, | |
| "learning_rate": 9.8656178302196e-06, | |
| "loss": 3.967, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 4.108416547788873, | |
| "grad_norm": 11.603713035583496, | |
| "learning_rate": 9.235307465395961e-06, | |
| "loss": 3.9525, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 4.136947218259629, | |
| "grad_norm": 8.631877899169922, | |
| "learning_rate": 8.604997100572321e-06, | |
| "loss": 3.9658, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 4.165477888730385, | |
| "grad_norm": 7.5662407875061035, | |
| "learning_rate": 7.974686735748683e-06, | |
| "loss": 3.9118, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 4.194008559201142, | |
| "grad_norm": 10.707962036132812, | |
| "learning_rate": 7.344376370925044e-06, | |
| "loss": 3.948, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 4.222539229671897, | |
| "grad_norm": 11.31724739074707, | |
| "learning_rate": 6.714066006101405e-06, | |
| "loss": 3.9319, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 4.251069900142653, | |
| "grad_norm": 9.398261070251465, | |
| "learning_rate": 6.083755641277765e-06, | |
| "loss": 3.9621, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 4.279600570613409, | |
| "grad_norm": 11.119857788085938, | |
| "learning_rate": 5.453445276454126e-06, | |
| "loss": 3.9429, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 4.3081312410841655, | |
| "grad_norm": 11.433586120605469, | |
| "learning_rate": 4.823134911630487e-06, | |
| "loss": 3.9126, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 4.336661911554922, | |
| "grad_norm": 10.043067932128906, | |
| "learning_rate": 4.192824546806848e-06, | |
| "loss": 3.9218, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 4.365192582025678, | |
| "grad_norm": 9.148512840270996, | |
| "learning_rate": 3.5625141819832086e-06, | |
| "loss": 3.9124, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 4.393723252496434, | |
| "grad_norm": 9.347081184387207, | |
| "learning_rate": 2.9322038171595694e-06, | |
| "loss": 3.8917, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 4.422253922967189, | |
| "grad_norm": 8.787015914916992, | |
| "learning_rate": 2.30189345233593e-06, | |
| "loss": 3.9091, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 4.4507845934379455, | |
| "grad_norm": 8.788630485534668, | |
| "learning_rate": 1.671583087512291e-06, | |
| "loss": 3.9211, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 4.479315263908702, | |
| "grad_norm": 8.639442443847656, | |
| "learning_rate": 1.0412727226886518e-06, | |
| "loss": 3.9139, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 4.507845934379458, | |
| "grad_norm": 9.101580619812012, | |
| "learning_rate": 4.1096235786501275e-07, | |
| "loss": 3.911, | |
| "step": 79000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 79326, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 8192, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6441172712886272e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |