| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.8716681835912103, |
| "eval_steps": 100, |
| "global_step": 1800, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.005200884150305552, |
| "grad_norm": 4.82585334777832, |
| "learning_rate": 5.6140350877192985e-06, |
| "loss": 0.7676299571990967, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.010401768300611104, |
| "grad_norm": 3.292309522628784, |
| "learning_rate": 1.263157894736842e-05, |
| "loss": 0.631245470046997, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.015602652450916656, |
| "grad_norm": 3.0301995277404785, |
| "learning_rate": 1.9649122807017544e-05, |
| "loss": 0.6749040126800537, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.02080353660122221, |
| "grad_norm": 2.853067398071289, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.7889204978942871, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.02600442075152776, |
| "grad_norm": 2.4595491886138916, |
| "learning_rate": 3.368421052631579e-05, |
| "loss": 0.8261689186096192, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.031205304901833313, |
| "grad_norm": 3.285839319229126, |
| "learning_rate": 4.070175438596492e-05, |
| "loss": 0.8855669021606445, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.03640618905213886, |
| "grad_norm": 2.9381635189056396, |
| "learning_rate": 4.771929824561404e-05, |
| "loss": 0.8640634536743164, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.04160707320244442, |
| "grad_norm": 2.142974376678467, |
| "learning_rate": 5.4736842105263165e-05, |
| "loss": 0.8938538551330566, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.046807957352749965, |
| "grad_norm": 3.4374115467071533, |
| "learning_rate": 6.175438596491228e-05, |
| "loss": 0.9762144088745117, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.05200884150305552, |
| "grad_norm": 2.2888057231903076, |
| "learning_rate": 6.87719298245614e-05, |
| "loss": 1.0100465774536134, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.05720972565336107, |
| "grad_norm": 2.101193428039551, |
| "learning_rate": 7.578947368421052e-05, |
| "loss": 1.0064237594604493, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.062410609803666625, |
| "grad_norm": 2.3019754886627197, |
| "learning_rate": 7.999977348305978e-05, |
| "loss": 1.0382986068725586, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.06761149395397217, |
| "grad_norm": 2.0822298526763916, |
| "learning_rate": 7.999722519694521e-05, |
| "loss": 1.1403823852539063, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.07281237810427772, |
| "grad_norm": 2.109672784805298, |
| "learning_rate": 7.999184565952555e-05, |
| "loss": 1.0954228401184083, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.07801326225458328, |
| "grad_norm": 1.9484328031539917, |
| "learning_rate": 7.998363525159777e-05, |
| "loss": 1.125577735900879, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.08321414640488883, |
| "grad_norm": 2.0304486751556396, |
| "learning_rate": 7.997259455434539e-05, |
| "loss": 1.078431987762451, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.08841503055519438, |
| "grad_norm": 1.8764644861221313, |
| "learning_rate": 7.995872434929734e-05, |
| "loss": 1.0825774192810058, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.09361591470549993, |
| "grad_norm": 2.521047353744507, |
| "learning_rate": 7.994202561827263e-05, |
| "loss": 1.0628202438354493, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.0988167988558055, |
| "grad_norm": 1.7737245559692383, |
| "learning_rate": 7.992249954331088e-05, |
| "loss": 1.0672310829162597, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.10401768300611104, |
| "grad_norm": 2.0016262531280518, |
| "learning_rate": 7.990014750658856e-05, |
| "loss": 1.0844260215759278, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.10401768300611104, |
| "eval_loss": 1.090377688407898, |
| "eval_runtime": 131.495, |
| "eval_samples_per_second": 1.947, |
| "eval_steps_per_second": 1.947, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.10921856715641659, |
| "grad_norm": 2.2041330337524414, |
| "learning_rate": 7.987497109032128e-05, |
| "loss": 1.095380973815918, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.11441945130672214, |
| "grad_norm": 1.7770912647247314, |
| "learning_rate": 7.98469720766517e-05, |
| "loss": 1.076313304901123, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.11962033545702769, |
| "grad_norm": 2.0055956840515137, |
| "learning_rate": 7.98161524475234e-05, |
| "loss": 1.069803810119629, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.12482121960733325, |
| "grad_norm": 1.9808892011642456, |
| "learning_rate": 7.978251438454063e-05, |
| "loss": 1.062901496887207, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.13002210375763879, |
| "grad_norm": 1.7519869804382324, |
| "learning_rate": 7.974606026881378e-05, |
| "loss": 1.03425931930542, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.13522298790794435, |
| "grad_norm": 1.8879916667938232, |
| "learning_rate": 7.970679268079092e-05, |
| "loss": 1.085061264038086, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.1404238720582499, |
| "grad_norm": 1.855212926864624, |
| "learning_rate": 7.966471440007514e-05, |
| "loss": 1.0653216361999511, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.14562475620855544, |
| "grad_norm": 1.9829131364822388, |
| "learning_rate": 7.961982840522773e-05, |
| "loss": 1.0679091453552245, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.150825640358861, |
| "grad_norm": 1.8631235361099243, |
| "learning_rate": 7.957213787355738e-05, |
| "loss": 1.0336699485778809, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.15602652450916657, |
| "grad_norm": 1.7199710607528687, |
| "learning_rate": 7.95216461808953e-05, |
| "loss": 1.085853385925293, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.1612274086594721, |
| "grad_norm": 1.8429338932037354, |
| "learning_rate": 7.946835690135617e-05, |
| "loss": 1.039409065246582, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.16642829280977767, |
| "grad_norm": 1.9319918155670166, |
| "learning_rate": 7.941227380708522e-05, |
| "loss": 1.0467087745666503, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.1716291769600832, |
| "grad_norm": 2.0441744327545166, |
| "learning_rate": 7.93534008679912e-05, |
| "loss": 1.0675994873046875, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.17683006111038876, |
| "grad_norm": 1.847270131111145, |
| "learning_rate": 7.929174225146534e-05, |
| "loss": 1.0682602882385255, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.18203094526069433, |
| "grad_norm": 1.8770182132720947, |
| "learning_rate": 7.922730232208637e-05, |
| "loss": 1.0655933380126954, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.18723182941099986, |
| "grad_norm": 1.9199661016464233, |
| "learning_rate": 7.916008564131158e-05, |
| "loss": 1.0514616012573241, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.19243271356130542, |
| "grad_norm": 1.7040131092071533, |
| "learning_rate": 7.909009696715391e-05, |
| "loss": 1.0357669830322265, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.197633597711611, |
| "grad_norm": 1.7878484725952148, |
| "learning_rate": 7.901734125384515e-05, |
| "loss": 1.034726905822754, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.20283448186191652, |
| "grad_norm": 1.7647621631622314, |
| "learning_rate": 7.89418236514853e-05, |
| "loss": 1.0258402824401855, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.20803536601222208, |
| "grad_norm": 1.7211089134216309, |
| "learning_rate": 7.886354950567789e-05, |
| "loss": 1.0237075805664062, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.20803536601222208, |
| "eval_loss": 1.0211308002471924, |
| "eval_runtime": 130.7508, |
| "eval_samples_per_second": 1.958, |
| "eval_steps_per_second": 1.958, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.21323625016252762, |
| "grad_norm": 1.9167125225067139, |
| "learning_rate": 7.878252435715178e-05, |
| "loss": 1.0424039840698243, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.21843713431283318, |
| "grad_norm": 1.6589795351028442, |
| "learning_rate": 7.869875394136873e-05, |
| "loss": 0.985629940032959, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.22363801846313874, |
| "grad_norm": 1.7960469722747803, |
| "learning_rate": 7.861224418811757e-05, |
| "loss": 1.0503170013427734, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.22883890261344428, |
| "grad_norm": 1.876910924911499, |
| "learning_rate": 7.852300122109443e-05, |
| "loss": 1.0263469696044922, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.23403978676374984, |
| "grad_norm": 1.7034902572631836, |
| "learning_rate": 7.843103135746916e-05, |
| "loss": 1.0396705627441407, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.23924067091405538, |
| "grad_norm": 1.7886664867401123, |
| "learning_rate": 7.833634110743832e-05, |
| "loss": 1.0125335693359374, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.24444155506436094, |
| "grad_norm": 1.7878754138946533, |
| "learning_rate": 7.823893717376422e-05, |
| "loss": 0.9978487968444825, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.2496424392146665, |
| "grad_norm": 1.6163357496261597, |
| "learning_rate": 7.813882645130049e-05, |
| "loss": 1.0093013763427734, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.25484332336497206, |
| "grad_norm": 1.7179474830627441, |
| "learning_rate": 7.80360160265041e-05, |
| "loss": 0.9946321487426758, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.26004420751527757, |
| "grad_norm": 1.828262209892273, |
| "learning_rate": 7.793051317693359e-05, |
| "loss": 1.0461179733276367, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.26524509166558313, |
| "grad_norm": 1.837222933769226, |
| "learning_rate": 7.782232537073406e-05, |
| "loss": 1.0099411964416505, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.2704459758158887, |
| "grad_norm": 1.9459391832351685, |
| "learning_rate": 7.771146026610844e-05, |
| "loss": 0.9811776161193848, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.27564685996619426, |
| "grad_norm": 1.8060634136199951, |
| "learning_rate": 7.759792571077545e-05, |
| "loss": 0.9858491897583008, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.2808477441164998, |
| "grad_norm": 1.8257906436920166, |
| "learning_rate": 7.748172974141401e-05, |
| "loss": 1.0260663032531738, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.2860486282668054, |
| "grad_norm": 1.7271229028701782, |
| "learning_rate": 7.736288058309448e-05, |
| "loss": 1.0253376007080077, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.2912495124171109, |
| "grad_norm": 1.6389424800872803, |
| "learning_rate": 7.724138664869628e-05, |
| "loss": 0.9872174263000488, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.29645039656741645, |
| "grad_norm": 1.7230466604232788, |
| "learning_rate": 7.711725653831255e-05, |
| "loss": 1.0354929924011231, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.301651280717722, |
| "grad_norm": 1.7749247550964355, |
| "learning_rate": 7.699049903864125e-05, |
| "loss": 0.9916903495788574, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.3068521648680276, |
| "grad_norm": 1.7340270280838013, |
| "learning_rate": 7.68611231223632e-05, |
| "loss": 1.0025782585144043, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.31205304901833314, |
| "grad_norm": 1.5753134489059448, |
| "learning_rate": 7.672913794750699e-05, |
| "loss": 0.9782845497131347, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.31205304901833314, |
| "eval_loss": 1.01508629322052, |
| "eval_runtime": 130.89, |
| "eval_samples_per_second": 1.956, |
| "eval_steps_per_second": 1.956, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.31725393316863865, |
| "grad_norm": 1.6529031991958618, |
| "learning_rate": 7.659455285680071e-05, |
| "loss": 0.9918001174926758, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.3224548173189442, |
| "grad_norm": 1.6655994653701782, |
| "learning_rate": 7.645737737701054e-05, |
| "loss": 0.9847027778625488, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.32765570146924977, |
| "grad_norm": 1.7864056825637817, |
| "learning_rate": 7.631762121826653e-05, |
| "loss": 0.9591518402099609, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.33285658561955533, |
| "grad_norm": 1.7958587408065796, |
| "learning_rate": 7.617529427337507e-05, |
| "loss": 0.9712434768676758, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.3380574697698609, |
| "grad_norm": 1.7462350130081177, |
| "learning_rate": 7.603040661711879e-05, |
| "loss": 1.0182976722717285, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.3432583539201664, |
| "grad_norm": 1.73703932762146, |
| "learning_rate": 7.588296850554326e-05, |
| "loss": 0.978699779510498, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.34845923807047197, |
| "grad_norm": 1.781213402748108, |
| "learning_rate": 7.573299037523115e-05, |
| "loss": 1.0263344764709472, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.35366012222077753, |
| "grad_norm": 1.7940926551818848, |
| "learning_rate": 7.558048284256334e-05, |
| "loss": 1.039241409301758, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.3588610063710831, |
| "grad_norm": 1.725530982017517, |
| "learning_rate": 7.542545670296747e-05, |
| "loss": 0.9716745376586914, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.36406189052138865, |
| "grad_norm": 1.665209174156189, |
| "learning_rate": 7.526792293015382e-05, |
| "loss": 0.9396204948425293, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.36926277467169416, |
| "grad_norm": 1.8031824827194214, |
| "learning_rate": 7.51078926753384e-05, |
| "loss": 0.9682992935180664, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.3744636588219997, |
| "grad_norm": 1.7742233276367188, |
| "learning_rate": 7.494537726645375e-05, |
| "loss": 0.9833150863647461, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.3796645429723053, |
| "grad_norm": 1.7864766120910645, |
| "learning_rate": 7.478038820734698e-05, |
| "loss": 0.9690051078796387, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.38486542712261085, |
| "grad_norm": 1.7735403776168823, |
| "learning_rate": 7.461293717696549e-05, |
| "loss": 0.9864785194396972, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.3900663112729164, |
| "grad_norm": 1.6049963235855103, |
| "learning_rate": 7.444303602853024e-05, |
| "loss": 0.9804511070251465, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.395267195423222, |
| "grad_norm": 1.6558054685592651, |
| "learning_rate": 7.427069678869672e-05, |
| "loss": 0.9362913131713867, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.4004680795735275, |
| "grad_norm": 1.6555157899856567, |
| "learning_rate": 7.409593165670364e-05, |
| "loss": 0.948128604888916, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.40566896372383304, |
| "grad_norm": 1.812888741493225, |
| "learning_rate": 7.391875300350937e-05, |
| "loss": 0.9553811073303222, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.4108698478741386, |
| "grad_norm": 1.6602718830108643, |
| "learning_rate": 7.373917337091626e-05, |
| "loss": 0.940127944946289, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.41607073202444417, |
| "grad_norm": 1.7474323511123657, |
| "learning_rate": 7.355720547068288e-05, |
| "loss": 0.9683933258056641, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.41607073202444417, |
| "eval_loss": 0.9856516122817993, |
| "eval_runtime": 130.2348, |
| "eval_samples_per_second": 1.966, |
| "eval_steps_per_second": 1.966, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.42127161617474973, |
| "grad_norm": 1.7211471796035767, |
| "learning_rate": 7.337286218362413e-05, |
| "loss": 0.9713785171508789, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.42647250032505524, |
| "grad_norm": 1.839705467224121, |
| "learning_rate": 7.31861565586995e-05, |
| "loss": 0.9828580856323242, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.4316733844753608, |
| "grad_norm": 1.6094111204147339, |
| "learning_rate": 7.299710181208943e-05, |
| "loss": 0.9652169227600098, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.43687426862566636, |
| "grad_norm": 1.8238450288772583, |
| "learning_rate": 7.280571132625971e-05, |
| "loss": 0.9175550460815429, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.4420751527759719, |
| "grad_norm": 1.7738935947418213, |
| "learning_rate": 7.261199864901423e-05, |
| "loss": 0.959293270111084, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.4472760369262775, |
| "grad_norm": 1.7823359966278076, |
| "learning_rate": 7.241597749253593e-05, |
| "loss": 0.9460945129394531, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.452476921076583, |
| "grad_norm": 1.852073073387146, |
| "learning_rate": 7.221766173241629e-05, |
| "loss": 0.9572076797485352, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.45767780522688856, |
| "grad_norm": 1.7884498834609985, |
| "learning_rate": 7.201706540667293e-05, |
| "loss": 0.980253028869629, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.4628786893771941, |
| "grad_norm": 1.6045715808868408, |
| "learning_rate": 7.181420271475616e-05, |
| "loss": 0.9448029518127441, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.4680795735274997, |
| "grad_norm": 1.703908920288086, |
| "learning_rate": 7.160908801654358e-05, |
| "loss": 0.9464167594909668, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.47328045767780524, |
| "grad_norm": 1.5748329162597656, |
| "learning_rate": 7.14017358313239e-05, |
| "loss": 0.9481819152832032, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.47848134182811075, |
| "grad_norm": 1.62021803855896, |
| "learning_rate": 7.11921608367689e-05, |
| "loss": 0.9535123825073242, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.4836822259784163, |
| "grad_norm": 1.6786726713180542, |
| "learning_rate": 7.09803778678946e-05, |
| "loss": 0.9362807273864746, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.4888831101287219, |
| "grad_norm": 1.602408528327942, |
| "learning_rate": 7.076640191601113e-05, |
| "loss": 0.9352068901062012, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.49408399427902744, |
| "grad_norm": 1.710305094718933, |
| "learning_rate": 7.055024812766153e-05, |
| "loss": 0.9479847908020019, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.499284878429333, |
| "grad_norm": 1.7393652200698853, |
| "learning_rate": 7.03319318035496e-05, |
| "loss": 0.9210905075073242, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.5044857625796385, |
| "grad_norm": 1.6267069578170776, |
| "learning_rate": 7.011146839745681e-05, |
| "loss": 0.935267448425293, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.5096866467299441, |
| "grad_norm": 1.5649526119232178, |
| "learning_rate": 6.98888735151484e-05, |
| "loss": 0.950043773651123, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.5148875308802496, |
| "grad_norm": 1.7210514545440674, |
| "learning_rate": 6.966416291326867e-05, |
| "loss": 0.9338683128356934, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.5200884150305551, |
| "grad_norm": 1.7836523056030273, |
| "learning_rate": 6.943735249822567e-05, |
| "loss": 0.9206833839416504, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5200884150305551, |
| "eval_loss": 0.9551063776016235, |
| "eval_runtime": 130.2472, |
| "eval_samples_per_second": 1.965, |
| "eval_steps_per_second": 1.965, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.5252892991808608, |
| "grad_norm": 1.6640737056732178, |
| "learning_rate": 6.920845832506523e-05, |
| "loss": 0.9343124389648437, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.5304901833311663, |
| "grad_norm": 1.5780658721923828, |
| "learning_rate": 6.897749659633448e-05, |
| "loss": 0.9712604522705078, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.5356910674814719, |
| "grad_norm": 1.6826986074447632, |
| "learning_rate": 6.874448366093497e-05, |
| "loss": 0.9570392608642578, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.5408919516317774, |
| "grad_norm": 1.6517504453659058, |
| "learning_rate": 6.850943601296529e-05, |
| "loss": 0.9325905799865722, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.546092835782083, |
| "grad_norm": 1.8107248544692993, |
| "learning_rate": 6.827237029055364e-05, |
| "loss": 0.9377958297729492, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.5512937199323885, |
| "grad_norm": 1.6924495697021484, |
| "learning_rate": 6.803330327468006e-05, |
| "loss": 0.9227448463439941, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.556494604082694, |
| "grad_norm": 1.6941320896148682, |
| "learning_rate": 6.779225188798846e-05, |
| "loss": 0.9221724510192871, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.5616954882329996, |
| "grad_norm": 1.605466604232788, |
| "learning_rate": 6.754923319358883e-05, |
| "loss": 0.9474879264831543, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.5668963723833051, |
| "grad_norm": 1.6994117498397827, |
| "learning_rate": 6.730426439384939e-05, |
| "loss": 0.9544130325317383, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.5720972565336108, |
| "grad_norm": 1.6269444227218628, |
| "learning_rate": 6.705736282917892e-05, |
| "loss": 0.8944686889648438, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.5772981406839163, |
| "grad_norm": 1.7221931219100952, |
| "learning_rate": 6.680854597679924e-05, |
| "loss": 0.9075075149536133, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.5824990248342218, |
| "grad_norm": 1.7379156351089478, |
| "learning_rate": 6.655783144950809e-05, |
| "loss": 0.9554330825805664, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.5876999089845274, |
| "grad_norm": 1.7438715696334839, |
| "learning_rate": 6.63052369944324e-05, |
| "loss": 0.9195778846740723, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.5929007931348329, |
| "grad_norm": 1.6393606662750244, |
| "learning_rate": 6.605078049177205e-05, |
| "loss": 0.915708065032959, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.5981016772851385, |
| "grad_norm": 1.7054882049560547, |
| "learning_rate": 6.579447995353416e-05, |
| "loss": 0.9384199142456054, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.603302561435444, |
| "grad_norm": 1.5726091861724854, |
| "learning_rate": 6.553635352225812e-05, |
| "loss": 0.9182180404663086, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.6085034455857495, |
| "grad_norm": 1.827543020248413, |
| "learning_rate": 6.52764194697313e-05, |
| "loss": 0.9172782897949219, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.6137043297360552, |
| "grad_norm": 1.541303038597107, |
| "learning_rate": 6.501469619569579e-05, |
| "loss": 0.9366864204406739, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.6189052138863607, |
| "grad_norm": 1.6896532773971558, |
| "learning_rate": 6.475120222654573e-05, |
| "loss": 0.91693115234375, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.6241060980366663, |
| "grad_norm": 1.8263614177703857, |
| "learning_rate": 6.448595621401615e-05, |
| "loss": 0.9207332611083985, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.6241060980366663, |
| "eval_loss": 0.9301373958587646, |
| "eval_runtime": 129.9867, |
| "eval_samples_per_second": 1.969, |
| "eval_steps_per_second": 1.969, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.6293069821869718, |
| "grad_norm": 2.5737788677215576, |
| "learning_rate": 6.421897693386245e-05, |
| "loss": 1.8008251190185547, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.6345078663372773, |
| "grad_norm": 2.3786563873291016, |
| "learning_rate": 6.395028328453154e-05, |
| "loss": 1.0148398399353027, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.6397087504875829, |
| "grad_norm": 2.1619725227355957, |
| "learning_rate": 6.367989428582395e-05, |
| "loss": 0.9035258293151855, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.6449096346378884, |
| "grad_norm": 1.8177882432937622, |
| "learning_rate": 6.340782907754754e-05, |
| "loss": 0.9431497573852539, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.650110518788194, |
| "grad_norm": 2.0389435291290283, |
| "learning_rate": 6.313410691816269e-05, |
| "loss": 0.9617778778076171, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.6553114029384995, |
| "grad_norm": 1.8454713821411133, |
| "learning_rate": 6.285874718341898e-05, |
| "loss": 0.9618912696838379, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.660512287088805, |
| "grad_norm": 1.9317169189453125, |
| "learning_rate": 6.25817693649838e-05, |
| "loss": 1.020989227294922, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.6657131712391107, |
| "grad_norm": 2.249952554702759, |
| "learning_rate": 6.230319306906246e-05, |
| "loss": 1.0368837356567382, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.6709140553894162, |
| "grad_norm": 1.8525416851043701, |
| "learning_rate": 6.202303801501045e-05, |
| "loss": 0.9958114624023438, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.6761149395397218, |
| "grad_norm": 1.99785578250885, |
| "learning_rate": 6.174132403393749e-05, |
| "loss": 1.0103933334350585, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.6813158236900273, |
| "grad_norm": 2.1036858558654785, |
| "learning_rate": 6.145807106730384e-05, |
| "loss": 1.009110927581787, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.6865167078403328, |
| "grad_norm": 1.8276114463806152, |
| "learning_rate": 6.11732991655087e-05, |
| "loss": 1.0265812873840332, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.6917175919906384, |
| "grad_norm": 1.8524454832077026, |
| "learning_rate": 6.088702848647087e-05, |
| "loss": 1.0076023101806642, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.6969184761409439, |
| "grad_norm": 2.044529914855957, |
| "learning_rate": 6.059927929420191e-05, |
| "loss": 1.0623496055603028, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.7021193602912496, |
| "grad_norm": 1.9014661312103271, |
| "learning_rate": 6.031007195737173e-05, |
| "loss": 1.0505518913269043, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.7073202444415551, |
| "grad_norm": 1.7579931020736694, |
| "learning_rate": 6.0019426947866685e-05, |
| "loss": 1.0260109901428223, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.7125211285918606, |
| "grad_norm": 1.9789166450500488, |
| "learning_rate": 5.972736483934053e-05, |
| "loss": 1.0057721138000488, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.7177220127421662, |
| "grad_norm": 1.7603486776351929, |
| "learning_rate": 5.943390630575808e-05, |
| "loss": 0.9560928344726562, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.7229228968924717, |
| "grad_norm": 1.723509669303894, |
| "learning_rate": 5.9139072119931735e-05, |
| "loss": 1.0287532806396484, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.7281237810427773, |
| "grad_norm": 1.880836844444275, |
| "learning_rate": 5.884288315205113e-05, |
| "loss": 1.0159936904907227, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.7281237810427773, |
| "eval_loss": 1.0172600746154785, |
| "eval_runtime": 130.5328, |
| "eval_samples_per_second": 1.961, |
| "eval_steps_per_second": 1.961, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.7333246651930828, |
| "grad_norm": 1.7332347631454468, |
| "learning_rate": 5.8545360368205704e-05, |
| "loss": 0.9914582252502442, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.7385255493433883, |
| "grad_norm": 1.6425188779830933, |
| "learning_rate": 5.824652482890067e-05, |
| "loss": 0.9969341278076171, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.7437264334936939, |
| "grad_norm": 1.868652105331421, |
| "learning_rate": 5.794639768756625e-05, |
| "loss": 0.9733437538146973, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.7489273176439994, |
| "grad_norm": 1.7285988330841064, |
| "learning_rate": 5.764500018906019e-05, |
| "loss": 0.992225170135498, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.7541282017943051, |
| "grad_norm": 1.7920562028884888, |
| "learning_rate": 5.734235366816398e-05, |
| "loss": 0.9502806663513184, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.7593290859446106, |
| "grad_norm": 1.7255525588989258, |
| "learning_rate": 5.703847954807273e-05, |
| "loss": 0.9720046997070313, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.7645299700949162, |
| "grad_norm": 1.8702120780944824, |
| "learning_rate": 5.673339933887854e-05, |
| "loss": 1.0045533180236816, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.7697308542452217, |
| "grad_norm": 1.7282917499542236, |
| "learning_rate": 5.642713463604804e-05, |
| "loss": 0.998747444152832, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.7749317383955272, |
| "grad_norm": 1.6850981712341309, |
| "learning_rate": 5.6119707118893566e-05, |
| "loss": 0.9491232872009278, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.7801326225458328, |
| "grad_norm": 1.8323360681533813, |
| "learning_rate": 5.581113854903873e-05, |
| "loss": 0.9483734130859375, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.7853335066961383, |
| "grad_norm": 1.7287927865982056, |
| "learning_rate": 5.5501450768877895e-05, |
| "loss": 0.9630053520202637, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.790534390846444, |
| "grad_norm": 1.8735415935516357, |
| "learning_rate": 5.519066570003006e-05, |
| "loss": 0.9625276565551758, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.7957352749967495, |
| "grad_norm": 1.7333344221115112, |
| "learning_rate": 5.4878805341787106e-05, |
| "loss": 0.956482982635498, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.800936159147055, |
| "grad_norm": 1.7283087968826294, |
| "learning_rate": 5.456589176955656e-05, |
| "loss": 0.9767995834350586, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.8061370432973606, |
| "grad_norm": 1.6415756940841675, |
| "learning_rate": 5.4251947133298954e-05, |
| "loss": 0.9446447372436524, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.8113379274476661, |
| "grad_norm": 1.6519927978515625, |
| "learning_rate": 5.393699365595995e-05, |
| "loss": 0.9424878120422363, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.8165388115979717, |
| "grad_norm": 1.647696852684021, |
| "learning_rate": 5.3621053631897175e-05, |
| "loss": 0.9464174270629883, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.8217396957482772, |
| "grad_norm": 1.8417209386825562, |
| "learning_rate": 5.330414942530217e-05, |
| "loss": 0.9534976959228516, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.8269405798985827, |
| "grad_norm": 1.8787815570831299, |
| "learning_rate": 5.298630346861732e-05, |
| "loss": 0.902367115020752, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.8321414640488883, |
| "grad_norm": 1.68453848361969, |
| "learning_rate": 5.266753826094787e-05, |
| "loss": 0.9392575263977051, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.8321414640488883, |
| "eval_loss": 0.9664498567581177, |
| "eval_runtime": 129.775, |
| "eval_samples_per_second": 1.973, |
| "eval_steps_per_second": 1.973, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.8373423481991938, |
| "grad_norm": 1.501584529876709, |
| "learning_rate": 5.234787636646932e-05, |
| "loss": 0.9557664871215821, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.8425432323494995, |
| "grad_norm": 1.5360844135284424, |
| "learning_rate": 5.202734041283029e-05, |
| "loss": 0.9075163841247559, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.847744116499805, |
| "grad_norm": 1.65811288356781, |
| "learning_rate": 5.1705953089550675e-05, |
| "loss": 0.9179675102233886, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.8529450006501105, |
| "grad_norm": 1.61856210231781, |
| "learning_rate": 5.138373714641558e-05, |
| "loss": 0.9237114906311035, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.8581458848004161, |
| "grad_norm": 1.7458878755569458, |
| "learning_rate": 5.1060715391864965e-05, |
| "loss": 0.8636340141296387, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.8633467689507216, |
| "grad_norm": 1.6920363903045654, |
| "learning_rate": 5.073691069137912e-05, |
| "loss": 0.9312037467956543, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.8685476531010272, |
| "grad_norm": 1.714153528213501, |
| "learning_rate": 5.041234596586009e-05, |
| "loss": 0.9352224349975586, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.8737485372513327, |
| "grad_norm": 1.6470508575439453, |
| "learning_rate": 5.0087044190009206e-05, |
| "loss": 0.8943827629089356, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.8789494214016382, |
| "grad_norm": 1.622175693511963, |
| "learning_rate": 4.9761028390700754e-05, |
| "loss": 0.9128511428833008, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.8841503055519438, |
| "grad_norm": 1.6196850538253784, |
| "learning_rate": 4.9434321645352084e-05, |
| "loss": 0.9311873435974121, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.8893511897022494, |
| "grad_norm": 1.7249358892440796, |
| "learning_rate": 4.9106947080289925e-05, |
| "loss": 0.9176848411560059, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.894552073852555, |
| "grad_norm": 1.6511820554733276, |
| "learning_rate": 4.8778927869113445e-05, |
| "loss": 0.9112163543701172, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.8997529580028605, |
| "grad_norm": 1.5577549934387207, |
| "learning_rate": 4.845028723105389e-05, |
| "loss": 0.8965622901916503, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.904953842153166, |
| "grad_norm": 1.6360331773757935, |
| "learning_rate": 4.812104842933087e-05, |
| "loss": 0.913609504699707, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.9101547263034716, |
| "grad_norm": 1.451485276222229, |
| "learning_rate": 4.779123476950581e-05, |
| "loss": 0.8944340705871582, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.9153556104537771, |
| "grad_norm": 1.5654209852218628, |
| "learning_rate": 4.746086959783209e-05, |
| "loss": 0.8596482276916504, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.9205564946040827, |
| "grad_norm": 1.5695706605911255, |
| "learning_rate": 4.712997629960253e-05, |
| "loss": 0.8860456466674804, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.9257573787543882, |
| "grad_norm": 1.7602401971817017, |
| "learning_rate": 4.679857829749404e-05, |
| "loss": 0.8744911193847656, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.9309582629046937, |
| "grad_norm": 1.6938329935073853, |
| "learning_rate": 4.6466699049909584e-05, |
| "loss": 0.9156827926635742, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.9361591470549994, |
| "grad_norm": 1.4965558052062988, |
| "learning_rate": 4.6134362049317625e-05, |
| "loss": 0.9088706016540528, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.9361591470549994, |
| "eval_loss": 0.9005925059318542, |
| "eval_runtime": 129.9874, |
| "eval_samples_per_second": 1.969, |
| "eval_steps_per_second": 1.969, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.9413600312053049, |
| "grad_norm": 1.5755003690719604, |
| "learning_rate": 4.58015908205893e-05, |
| "loss": 0.8684419631958008, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.9465609153556105, |
| "grad_norm": 1.5753810405731201, |
| "learning_rate": 4.5468408919333025e-05, |
| "loss": 0.8938017845153808, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.951761799505916, |
| "grad_norm": 1.5218995809555054, |
| "learning_rate": 4.513483993022719e-05, |
| "loss": 0.851717185974121, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.9569626836562215, |
| "grad_norm": 1.6325594186782837, |
| "learning_rate": 4.4800907465350664e-05, |
| "loss": 0.8885617256164551, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.9621635678065271, |
| "grad_norm": 1.4952536821365356, |
| "learning_rate": 4.446663516251138e-05, |
| "loss": 0.8511702537536621, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.9673644519568326, |
| "grad_norm": 1.5268827676773071, |
| "learning_rate": 4.413204668357312e-05, |
| "loss": 0.8577158927917481, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.9725653361071382, |
| "grad_norm": 1.4323556423187256, |
| "learning_rate": 4.379716571278054e-05, |
| "loss": 0.8342367172241211, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.9777662202574438, |
| "grad_norm": 1.5065745115280151, |
| "learning_rate": 4.3462015955082665e-05, |
| "loss": 0.894658374786377, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.9829671044077494, |
| "grad_norm": 1.4767268896102905, |
| "learning_rate": 4.3126621134455e-05, |
| "loss": 0.8451394081115723, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.9881679885580549, |
| "grad_norm": 1.5763777494430542, |
| "learning_rate": 4.279100499222004e-05, |
| "loss": 0.8588390350341797, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.9933688727083604, |
| "grad_norm": 1.5762237310409546, |
| "learning_rate": 4.2455191285366845e-05, |
| "loss": 0.8754203796386719, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.998569756858666, |
| "grad_norm": 1.5145306587219238, |
| "learning_rate": 4.2119203784869304e-05, |
| "loss": 0.8561000823974609, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.0031205304901833, |
| "grad_norm": 1.470058798789978, |
| "learning_rate": 4.178306627400347e-05, |
| "loss": 0.7531153202056885, |
| "step": 965 |
| }, |
| { |
| "epoch": 1.0083214146404889, |
| "grad_norm": 1.6469708681106567, |
| "learning_rate": 4.144680254666413e-05, |
| "loss": 0.7166049003601074, |
| "step": 970 |
| }, |
| { |
| "epoch": 1.0135222987907944, |
| "grad_norm": 1.5783348083496094, |
| "learning_rate": 4.1110436405680376e-05, |
| "loss": 0.6519159317016602, |
| "step": 975 |
| }, |
| { |
| "epoch": 1.0187231829411, |
| "grad_norm": 1.494524598121643, |
| "learning_rate": 4.0773991661130806e-05, |
| "loss": 0.6752121925354004, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.0239240670914056, |
| "grad_norm": 1.523681640625, |
| "learning_rate": 4.043749212865811e-05, |
| "loss": 0.6791035175323487, |
| "step": 985 |
| }, |
| { |
| "epoch": 1.029124951241711, |
| "grad_norm": 1.4255659580230713, |
| "learning_rate": 4.010096162778309e-05, |
| "loss": 0.7047647476196289, |
| "step": 990 |
| }, |
| { |
| "epoch": 1.0343258353920166, |
| "grad_norm": 1.512510061264038, |
| "learning_rate": 3.97644239802188e-05, |
| "loss": 0.6946704387664795, |
| "step": 995 |
| }, |
| { |
| "epoch": 1.0395267195423221, |
| "grad_norm": 1.4743566513061523, |
| "learning_rate": 3.942790300818415e-05, |
| "loss": 0.676310396194458, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.0395267195423221, |
| "eval_loss": 0.8658661246299744, |
| "eval_runtime": 130.1356, |
| "eval_samples_per_second": 1.967, |
| "eval_steps_per_second": 1.967, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.0447276036926278, |
| "grad_norm": 1.5579502582550049, |
| "learning_rate": 3.90914225327176e-05, |
| "loss": 0.6901298522949219, |
| "step": 1005 |
| }, |
| { |
| "epoch": 1.0499284878429334, |
| "grad_norm": 1.5494235754013062, |
| "learning_rate": 3.875500637199108e-05, |
| "loss": 0.6718050003051758, |
| "step": 1010 |
| }, |
| { |
| "epoch": 1.0551293719932389, |
| "grad_norm": 1.5289891958236694, |
| "learning_rate": 3.8418678339623876e-05, |
| "loss": 0.6892761707305908, |
| "step": 1015 |
| }, |
| { |
| "epoch": 1.0603302561435444, |
| "grad_norm": 1.4304561614990234, |
| "learning_rate": 3.8082462242997025e-05, |
| "loss": 0.6954478263854981, |
| "step": 1020 |
| }, |
| { |
| "epoch": 1.0655311402938499, |
| "grad_norm": 1.4161802530288696, |
| "learning_rate": 3.774638188156808e-05, |
| "loss": 0.6717328548431396, |
| "step": 1025 |
| }, |
| { |
| "epoch": 1.0707320244441556, |
| "grad_norm": 1.438629150390625, |
| "learning_rate": 3.741046104518636e-05, |
| "loss": 0.6766324043273926, |
| "step": 1030 |
| }, |
| { |
| "epoch": 1.0759329085944611, |
| "grad_norm": 1.5406919717788696, |
| "learning_rate": 3.707472351240907e-05, |
| "loss": 0.7088249206542969, |
| "step": 1035 |
| }, |
| { |
| "epoch": 1.0811337927447666, |
| "grad_norm": 1.3865864276885986, |
| "learning_rate": 3.6739193048818e-05, |
| "loss": 0.6776373386383057, |
| "step": 1040 |
| }, |
| { |
| "epoch": 1.0863346768950721, |
| "grad_norm": 1.5101228952407837, |
| "learning_rate": 3.640389340533736e-05, |
| "loss": 0.660363245010376, |
| "step": 1045 |
| }, |
| { |
| "epoch": 1.0915355610453776, |
| "grad_norm": 1.4708372354507446, |
| "learning_rate": 3.6068848316552415e-05, |
| "loss": 0.6472445487976074, |
| "step": 1050 |
| }, |
| { |
| "epoch": 1.0967364451956834, |
| "grad_norm": 1.5044161081314087, |
| "learning_rate": 3.573408149902955e-05, |
| "loss": 0.7127279758453369, |
| "step": 1055 |
| }, |
| { |
| "epoch": 1.1019373293459889, |
| "grad_norm": 1.4291512966156006, |
| "learning_rate": 3.53996166496373e-05, |
| "loss": 0.6595888614654541, |
| "step": 1060 |
| }, |
| { |
| "epoch": 1.1071382134962944, |
| "grad_norm": 1.5285484790802002, |
| "learning_rate": 3.506547744386911e-05, |
| "loss": 0.6698886394500733, |
| "step": 1065 |
| }, |
| { |
| "epoch": 1.1123390976465999, |
| "grad_norm": 1.3300918340682983, |
| "learning_rate": 3.473168753416728e-05, |
| "loss": 0.6541372776031494, |
| "step": 1070 |
| }, |
| { |
| "epoch": 1.1175399817969054, |
| "grad_norm": 1.329465389251709, |
| "learning_rate": 3.439827054824881e-05, |
| "loss": 0.6664098739624024, |
| "step": 1075 |
| }, |
| { |
| "epoch": 1.1227408659472111, |
| "grad_norm": 1.3707212209701538, |
| "learning_rate": 3.406525008743285e-05, |
| "loss": 0.6693881034851075, |
| "step": 1080 |
| }, |
| { |
| "epoch": 1.1279417500975166, |
| "grad_norm": 1.474888563156128, |
| "learning_rate": 3.373264972497002e-05, |
| "loss": 0.6865148067474365, |
| "step": 1085 |
| }, |
| { |
| "epoch": 1.1331426342478221, |
| "grad_norm": 1.370944857597351, |
| "learning_rate": 3.3400493004373815e-05, |
| "loss": 0.6525755405426026, |
| "step": 1090 |
| }, |
| { |
| "epoch": 1.1383435183981276, |
| "grad_norm": 1.5666426420211792, |
| "learning_rate": 3.306880343775396e-05, |
| "loss": 0.6941968917846679, |
| "step": 1095 |
| }, |
| { |
| "epoch": 1.1435444025484331, |
| "grad_norm": 1.3775230646133423, |
| "learning_rate": 3.273760450415218e-05, |
| "loss": 0.6021693706512451, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.1435444025484331, |
| "eval_loss": 0.8314206600189209, |
| "eval_runtime": 130.5631, |
| "eval_samples_per_second": 1.961, |
| "eval_steps_per_second": 1.961, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.1487452866987389, |
| "grad_norm": 1.4845924377441406, |
| "learning_rate": 3.240691964788012e-05, |
| "loss": 0.6344992160797119, |
| "step": 1105 |
| }, |
| { |
| "epoch": 1.1539461708490444, |
| "grad_norm": 1.515212059020996, |
| "learning_rate": 3.207677227685989e-05, |
| "loss": 0.6564799785614014, |
| "step": 1110 |
| }, |
| { |
| "epoch": 1.15914705499935, |
| "grad_norm": 1.382857084274292, |
| "learning_rate": 3.1747185760966985e-05, |
| "loss": 0.6568068981170654, |
| "step": 1115 |
| }, |
| { |
| "epoch": 1.1643479391496554, |
| "grad_norm": 1.4786866903305054, |
| "learning_rate": 3.141818343037621e-05, |
| "loss": 0.6519091129302979, |
| "step": 1120 |
| }, |
| { |
| "epoch": 1.169548823299961, |
| "grad_norm": 1.5029971599578857, |
| "learning_rate": 3.108978857391002e-05, |
| "loss": 0.6406647682189941, |
| "step": 1125 |
| }, |
| { |
| "epoch": 1.1747497074502666, |
| "grad_norm": 1.4615740776062012, |
| "learning_rate": 3.076202443739014e-05, |
| "loss": 0.6860923290252685, |
| "step": 1130 |
| }, |
| { |
| "epoch": 1.1799505916005721, |
| "grad_norm": 1.4388511180877686, |
| "learning_rate": 3.0434914221992032e-05, |
| "loss": 0.6385154724121094, |
| "step": 1135 |
| }, |
| { |
| "epoch": 1.1851514757508776, |
| "grad_norm": 1.4187670946121216, |
| "learning_rate": 3.0108481082602517e-05, |
| "loss": 0.6268198013305664, |
| "step": 1140 |
| }, |
| { |
| "epoch": 1.1903523599011832, |
| "grad_norm": 1.5264593362808228, |
| "learning_rate": 2.9782748126180885e-05, |
| "loss": 0.6531328678131103, |
| "step": 1145 |
| }, |
| { |
| "epoch": 1.1955532440514887, |
| "grad_norm": 1.4612548351287842, |
| "learning_rate": 2.945773841012303e-05, |
| "loss": 0.669494915008545, |
| "step": 1150 |
| }, |
| { |
| "epoch": 1.2007541282017944, |
| "grad_norm": 1.267225742340088, |
| "learning_rate": 2.913347494062949e-05, |
| "loss": 0.6084520816802979, |
| "step": 1155 |
| }, |
| { |
| "epoch": 1.2059550123521, |
| "grad_norm": 1.449588418006897, |
| "learning_rate": 2.8809980671076803e-05, |
| "loss": 0.6461035728454589, |
| "step": 1160 |
| }, |
| { |
| "epoch": 1.2111558965024054, |
| "grad_norm": 1.3878182172775269, |
| "learning_rate": 2.848727850039281e-05, |
| "loss": 0.6444483280181885, |
| "step": 1165 |
| }, |
| { |
| "epoch": 1.216356780652711, |
| "grad_norm": 1.510225772857666, |
| "learning_rate": 2.8165391271435667e-05, |
| "loss": 0.6362089157104492, |
| "step": 1170 |
| }, |
| { |
| "epoch": 1.2215576648030164, |
| "grad_norm": 1.3843779563903809, |
| "learning_rate": 2.7844341769376888e-05, |
| "loss": 0.6338253974914551, |
| "step": 1175 |
| }, |
| { |
| "epoch": 1.2267585489533221, |
| "grad_norm": 1.414486289024353, |
| "learning_rate": 2.752415272008853e-05, |
| "loss": 0.6381827354431152, |
| "step": 1180 |
| }, |
| { |
| "epoch": 1.2319594331036277, |
| "grad_norm": 1.3414344787597656, |
| "learning_rate": 2.720484678853445e-05, |
| "loss": 0.6552750587463378, |
| "step": 1185 |
| }, |
| { |
| "epoch": 1.2371603172539332, |
| "grad_norm": 1.470218539237976, |
| "learning_rate": 2.6886446577165987e-05, |
| "loss": 0.6548665523529053, |
| "step": 1190 |
| }, |
| { |
| "epoch": 1.2423612014042387, |
| "grad_norm": 1.4118874073028564, |
| "learning_rate": 2.656897462432193e-05, |
| "loss": 0.6231658935546875, |
| "step": 1195 |
| }, |
| { |
| "epoch": 1.2475620855545442, |
| "grad_norm": 1.407113790512085, |
| "learning_rate": 2.6252453402633284e-05, |
| "loss": 0.5831446647644043, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.2475620855545442, |
| "eval_loss": 0.7889392971992493, |
| "eval_runtime": 130.1098, |
| "eval_samples_per_second": 1.968, |
| "eval_steps_per_second": 1.968, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.25276296970485, |
| "grad_norm": 1.5195738077163696, |
| "learning_rate": 2.593690531743231e-05, |
| "loss": 0.6311056613922119, |
| "step": 1205 |
| }, |
| { |
| "epoch": 1.2579638538551554, |
| "grad_norm": 1.4188272953033447, |
| "learning_rate": 2.5622352705166725e-05, |
| "loss": 0.6084880352020263, |
| "step": 1210 |
| }, |
| { |
| "epoch": 1.263164738005461, |
| "grad_norm": 1.3883174657821655, |
| "learning_rate": 2.5308817831818447e-05, |
| "loss": 0.6124576568603516, |
| "step": 1215 |
| }, |
| { |
| "epoch": 1.2683656221557664, |
| "grad_norm": 1.4531702995300293, |
| "learning_rate": 2.4996322891327576e-05, |
| "loss": 0.6091352462768554, |
| "step": 1220 |
| }, |
| { |
| "epoch": 1.273566506306072, |
| "grad_norm": 1.4332900047302246, |
| "learning_rate": 2.46848900040213e-05, |
| "loss": 0.6466999053955078, |
| "step": 1225 |
| }, |
| { |
| "epoch": 1.2787673904563777, |
| "grad_norm": 1.4958285093307495, |
| "learning_rate": 2.437454121504809e-05, |
| "loss": 0.616027545928955, |
| "step": 1230 |
| }, |
| { |
| "epoch": 1.2839682746066832, |
| "grad_norm": 1.4578135013580322, |
| "learning_rate": 2.406529849281726e-05, |
| "loss": 0.6332333087921143, |
| "step": 1235 |
| }, |
| { |
| "epoch": 1.2891691587569887, |
| "grad_norm": 1.4748591184616089, |
| "learning_rate": 2.375718372744384e-05, |
| "loss": 0.6314262866973877, |
| "step": 1240 |
| }, |
| { |
| "epoch": 1.2943700429072942, |
| "grad_norm": 1.4996991157531738, |
| "learning_rate": 2.3450218729199108e-05, |
| "loss": 0.6231923580169678, |
| "step": 1245 |
| }, |
| { |
| "epoch": 1.2995709270575997, |
| "grad_norm": 1.4916212558746338, |
| "learning_rate": 2.314442522696669e-05, |
| "loss": 0.6058928966522217, |
| "step": 1250 |
| }, |
| { |
| "epoch": 1.3047718112079054, |
| "grad_norm": 1.3732458353042603, |
| "learning_rate": 2.283982486670449e-05, |
| "loss": 0.5989131450653076, |
| "step": 1255 |
| }, |
| { |
| "epoch": 1.309972695358211, |
| "grad_norm": 1.3525115251541138, |
| "learning_rate": 2.2536439209912393e-05, |
| "loss": 0.6314842700958252, |
| "step": 1260 |
| }, |
| { |
| "epoch": 1.3151735795085164, |
| "grad_norm": 1.4011585712432861, |
| "learning_rate": 2.2234289732106102e-05, |
| "loss": 0.6161143779754639, |
| "step": 1265 |
| }, |
| { |
| "epoch": 1.320374463658822, |
| "grad_norm": 1.2732415199279785, |
| "learning_rate": 2.193339782129685e-05, |
| "loss": 0.5840548515319824, |
| "step": 1270 |
| }, |
| { |
| "epoch": 1.3255753478091274, |
| "grad_norm": 1.5150249004364014, |
| "learning_rate": 2.1633784776477585e-05, |
| "loss": 0.625645112991333, |
| "step": 1275 |
| }, |
| { |
| "epoch": 1.3307762319594332, |
| "grad_norm": 1.4052765369415283, |
| "learning_rate": 2.1335471806115117e-05, |
| "loss": 0.599894666671753, |
| "step": 1280 |
| }, |
| { |
| "epoch": 1.3359771161097387, |
| "grad_norm": 1.4317781925201416, |
| "learning_rate": 2.1038480026648945e-05, |
| "loss": 0.584019947052002, |
| "step": 1285 |
| }, |
| { |
| "epoch": 1.3411780002600442, |
| "grad_norm": 1.4249917268753052, |
| "learning_rate": 2.074283046099654e-05, |
| "loss": 0.6218175411224365, |
| "step": 1290 |
| }, |
| { |
| "epoch": 1.3463788844103497, |
| "grad_norm": 1.450347661972046, |
| "learning_rate": 2.044854403706509e-05, |
| "loss": 0.6018783569335937, |
| "step": 1295 |
| }, |
| { |
| "epoch": 1.3515797685606552, |
| "grad_norm": 1.3482204675674438, |
| "learning_rate": 2.0155641586270286e-05, |
| "loss": 0.5710643291473388, |
| "step": 1300 |
| }, |
| { |
| "epoch": 1.3515797685606552, |
| "eval_loss": 0.7516965270042419, |
| "eval_runtime": 129.9408, |
| "eval_samples_per_second": 1.97, |
| "eval_steps_per_second": 1.97, |
| "step": 1300 |
| }, |
| { |
| "epoch": 1.356780652710961, |
| "grad_norm": 1.4780831336975098, |
| "learning_rate": 1.9864143842061504e-05, |
| "loss": 0.6006845951080322, |
| "step": 1305 |
| }, |
| { |
| "epoch": 1.3619815368612664, |
| "grad_norm": 1.6098695993423462, |
| "learning_rate": 1.9574071438454388e-05, |
| "loss": 0.5974199295043945, |
| "step": 1310 |
| }, |
| { |
| "epoch": 1.367182421011572, |
| "grad_norm": 1.4274905920028687, |
| "learning_rate": 1.9285444908570064e-05, |
| "loss": 0.5996578216552735, |
| "step": 1315 |
| }, |
| { |
| "epoch": 1.3723833051618775, |
| "grad_norm": 1.4278326034545898, |
| "learning_rate": 1.8998284683181824e-05, |
| "loss": 0.6217168807983399, |
| "step": 1320 |
| }, |
| { |
| "epoch": 1.377584189312183, |
| "grad_norm": 1.3714914321899414, |
| "learning_rate": 1.871261108926882e-05, |
| "loss": 0.6057654857635498, |
| "step": 1325 |
| }, |
| { |
| "epoch": 1.3827850734624887, |
| "grad_norm": 1.468660831451416, |
| "learning_rate": 1.842844434857724e-05, |
| "loss": 0.6053857803344727, |
| "step": 1330 |
| }, |
| { |
| "epoch": 1.3879859576127942, |
| "grad_norm": 1.3222684860229492, |
| "learning_rate": 1.814580457618885e-05, |
| "loss": 0.5948523044586181, |
| "step": 1335 |
| }, |
| { |
| "epoch": 1.3931868417630997, |
| "grad_norm": 1.3946906328201294, |
| "learning_rate": 1.7864711779097122e-05, |
| "loss": 0.5642759323120117, |
| "step": 1340 |
| }, |
| { |
| "epoch": 1.3983877259134052, |
| "grad_norm": 1.404165506362915, |
| "learning_rate": 1.7585185854791076e-05, |
| "loss": 0.5775787353515625, |
| "step": 1345 |
| }, |
| { |
| "epoch": 1.4035886100637107, |
| "grad_norm": 1.370401382446289, |
| "learning_rate": 1.7307246589846767e-05, |
| "loss": 0.5661319732666016, |
| "step": 1350 |
| }, |
| { |
| "epoch": 1.4087894942140164, |
| "grad_norm": 1.3934425115585327, |
| "learning_rate": 1.7030913658526675e-05, |
| "loss": 0.5760215282440185, |
| "step": 1355 |
| }, |
| { |
| "epoch": 1.413990378364322, |
| "grad_norm": 1.389297366142273, |
| "learning_rate": 1.6756206621387e-05, |
| "loss": 0.5774462699890137, |
| "step": 1360 |
| }, |
| { |
| "epoch": 1.4191912625146275, |
| "grad_norm": 1.511326789855957, |
| "learning_rate": 1.6483144923893135e-05, |
| "loss": 0.585289192199707, |
| "step": 1365 |
| }, |
| { |
| "epoch": 1.424392146664933, |
| "grad_norm": 1.4418237209320068, |
| "learning_rate": 1.62117478950431e-05, |
| "loss": 0.5913166999816895, |
| "step": 1370 |
| }, |
| { |
| "epoch": 1.4295930308152385, |
| "grad_norm": 1.46451997756958, |
| "learning_rate": 1.5942034745999383e-05, |
| "loss": 0.5850035667419433, |
| "step": 1375 |
| }, |
| { |
| "epoch": 1.4347939149655442, |
| "grad_norm": 1.308406114578247, |
| "learning_rate": 1.5674024568729036e-05, |
| "loss": 0.5444441795349121, |
| "step": 1380 |
| }, |
| { |
| "epoch": 1.4399947991158497, |
| "grad_norm": 1.340357780456543, |
| "learning_rate": 1.5407736334652214e-05, |
| "loss": 0.5663641929626465, |
| "step": 1385 |
| }, |
| { |
| "epoch": 1.4451956832661552, |
| "grad_norm": 1.3704862594604492, |
| "learning_rate": 1.5143188893299266e-05, |
| "loss": 0.579612398147583, |
| "step": 1390 |
| }, |
| { |
| "epoch": 1.4503965674164607, |
| "grad_norm": 1.418916940689087, |
| "learning_rate": 1.4880400970976428e-05, |
| "loss": 0.5483696460723877, |
| "step": 1395 |
| }, |
| { |
| "epoch": 1.4555974515667662, |
| "grad_norm": 1.4289562702178955, |
| "learning_rate": 1.461939116944032e-05, |
| "loss": 0.5578598976135254, |
| "step": 1400 |
| }, |
| { |
| "epoch": 1.4555974515667662, |
| "eval_loss": 0.7209516763687134, |
| "eval_runtime": 129.7511, |
| "eval_samples_per_second": 1.973, |
| "eval_steps_per_second": 1.973, |
| "step": 1400 |
| }, |
| { |
| "epoch": 1.460798335717072, |
| "grad_norm": 1.397242546081543, |
| "learning_rate": 1.4360177964581156e-05, |
| "loss": 0.5415009498596192, |
| "step": 1405 |
| }, |
| { |
| "epoch": 1.4659992198673775, |
| "grad_norm": 1.3476572036743164, |
| "learning_rate": 1.4102779705114907e-05, |
| "loss": 0.5409503936767578, |
| "step": 1410 |
| }, |
| { |
| "epoch": 1.471200104017683, |
| "grad_norm": 1.4320838451385498, |
| "learning_rate": 1.3847214611284439e-05, |
| "loss": 0.5809547424316406, |
| "step": 1415 |
| }, |
| { |
| "epoch": 1.4764009881679885, |
| "grad_norm": 1.4180504083633423, |
| "learning_rate": 1.3593500773569854e-05, |
| "loss": 0.5446939945220948, |
| "step": 1420 |
| }, |
| { |
| "epoch": 1.481601872318294, |
| "grad_norm": 1.4159796237945557, |
| "learning_rate": 1.334165615140783e-05, |
| "loss": 0.5414989471435547, |
| "step": 1425 |
| }, |
| { |
| "epoch": 1.4868027564685997, |
| "grad_norm": 1.3080824613571167, |
| "learning_rate": 1.3091698571920422e-05, |
| "loss": 0.5712019443511963, |
| "step": 1430 |
| }, |
| { |
| "epoch": 1.4920036406189052, |
| "grad_norm": 1.2765617370605469, |
| "learning_rate": 1.284364572865311e-05, |
| "loss": 0.5223927974700928, |
| "step": 1435 |
| }, |
| { |
| "epoch": 1.4972045247692107, |
| "grad_norm": 1.459304928779602, |
| "learning_rate": 1.2597515180322368e-05, |
| "loss": 0.5872461318969726, |
| "step": 1440 |
| }, |
| { |
| "epoch": 1.5024054089195165, |
| "grad_norm": 1.3618521690368652, |
| "learning_rate": 1.23533243495727e-05, |
| "loss": 0.5361846446990967, |
| "step": 1445 |
| }, |
| { |
| "epoch": 1.5076062930698217, |
| "grad_norm": 1.3767704963684082, |
| "learning_rate": 1.2111090521743374e-05, |
| "loss": 0.5445142269134522, |
| "step": 1450 |
| }, |
| { |
| "epoch": 1.5128071772201275, |
| "grad_norm": 1.2690768241882324, |
| "learning_rate": 1.1870830843644917e-05, |
| "loss": 0.5451547622680664, |
| "step": 1455 |
| }, |
| { |
| "epoch": 1.518008061370433, |
| "grad_norm": 1.356768012046814, |
| "learning_rate": 1.1632562322345292e-05, |
| "loss": 0.5559448719024658, |
| "step": 1460 |
| }, |
| { |
| "epoch": 1.5232089455207385, |
| "grad_norm": 1.4047377109527588, |
| "learning_rate": 1.1396301823966054e-05, |
| "loss": 0.5453744888305664, |
| "step": 1465 |
| }, |
| { |
| "epoch": 1.5284098296710442, |
| "grad_norm": 1.4262696504592896, |
| "learning_rate": 1.1162066072488434e-05, |
| "loss": 0.5406304359436035, |
| "step": 1470 |
| }, |
| { |
| "epoch": 1.5336107138213495, |
| "grad_norm": 1.4099568128585815, |
| "learning_rate": 1.0929871648569591e-05, |
| "loss": 0.5617681980133057, |
| "step": 1475 |
| }, |
| { |
| "epoch": 1.5388115979716552, |
| "grad_norm": 1.317654013633728, |
| "learning_rate": 1.0699734988368817e-05, |
| "loss": 0.5236352443695068, |
| "step": 1480 |
| }, |
| { |
| "epoch": 1.5440124821219607, |
| "grad_norm": 1.3843480348587036, |
| "learning_rate": 1.0471672382384185e-05, |
| "loss": 0.5739215850830078, |
| "step": 1485 |
| }, |
| { |
| "epoch": 1.5492133662722662, |
| "grad_norm": 1.4144139289855957, |
| "learning_rate": 1.0245699974299356e-05, |
| "loss": 0.5243452548980713, |
| "step": 1490 |
| }, |
| { |
| "epoch": 1.554414250422572, |
| "grad_norm": 1.4148069620132446, |
| "learning_rate": 1.002183375984085e-05, |
| "loss": 0.5655048847198486, |
| "step": 1495 |
| }, |
| { |
| "epoch": 1.5596151345728773, |
| "grad_norm": 1.4749091863632202, |
| "learning_rate": 9.800089585645733e-06, |
| "loss": 0.5409163951873779, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.5596151345728773, |
| "eval_loss": 0.6939327120780945, |
| "eval_runtime": 129.7422, |
| "eval_samples_per_second": 1.973, |
| "eval_steps_per_second": 1.973, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.564816018723183, |
| "grad_norm": 1.397222876548767, |
| "learning_rate": 9.58048314813992e-06, |
| "loss": 0.5466369152069092, |
| "step": 1505 |
| }, |
| { |
| "epoch": 1.5700169028734885, |
| "grad_norm": 1.4299346208572388, |
| "learning_rate": 9.363029992427112e-06, |
| "loss": 0.5456615447998047, |
| "step": 1510 |
| }, |
| { |
| "epoch": 1.575217787023794, |
| "grad_norm": 1.4388434886932373, |
| "learning_rate": 9.147745511188363e-06, |
| "loss": 0.5131537914276123, |
| "step": 1515 |
| }, |
| { |
| "epoch": 1.5804186711740997, |
| "grad_norm": 1.3658560514450073, |
| "learning_rate": 8.934644943592551e-06, |
| "loss": 0.5328524112701416, |
| "step": 1520 |
| }, |
| { |
| "epoch": 1.585619555324405, |
| "grad_norm": 1.4248464107513428, |
| "learning_rate": 8.723743374217566e-06, |
| "loss": 0.5406161785125733, |
| "step": 1525 |
| }, |
| { |
| "epoch": 1.5908204394747107, |
| "grad_norm": 1.2822633981704712, |
| "learning_rate": 8.51505573198264e-06, |
| "loss": 0.5242678642272949, |
| "step": 1530 |
| }, |
| { |
| "epoch": 1.5960213236250163, |
| "grad_norm": 1.3976045846939087, |
| "learning_rate": 8.308596789091492e-06, |
| "loss": 0.5548820495605469, |
| "step": 1535 |
| }, |
| { |
| "epoch": 1.6012222077753218, |
| "grad_norm": 1.3557575941085815, |
| "learning_rate": 8.104381159986717e-06, |
| "loss": 0.5298625946044921, |
| "step": 1540 |
| }, |
| { |
| "epoch": 1.6064230919256275, |
| "grad_norm": 1.343584656715393, |
| "learning_rate": 7.902423300315271e-06, |
| "loss": 0.5180022239685058, |
| "step": 1545 |
| }, |
| { |
| "epoch": 1.6116239760759328, |
| "grad_norm": 1.4940485954284668, |
| "learning_rate": 7.70273750590521e-06, |
| "loss": 0.5428160667419434, |
| "step": 1550 |
| }, |
| { |
| "epoch": 1.6168248602262385, |
| "grad_norm": 1.3675767183303833, |
| "learning_rate": 7.5053379117537136e-06, |
| "loss": 0.516777229309082, |
| "step": 1555 |
| }, |
| { |
| "epoch": 1.622025744376544, |
| "grad_norm": 1.4647380113601685, |
| "learning_rate": 7.310238491026567e-06, |
| "loss": 0.5383810043334961, |
| "step": 1560 |
| }, |
| { |
| "epoch": 1.6272266285268495, |
| "grad_norm": 1.482635736465454, |
| "learning_rate": 7.117453054069039e-06, |
| "loss": 0.5144073009490967, |
| "step": 1565 |
| }, |
| { |
| "epoch": 1.6324275126771552, |
| "grad_norm": 1.5026439428329468, |
| "learning_rate": 6.926995247428285e-06, |
| "loss": 0.5365920066833496, |
| "step": 1570 |
| }, |
| { |
| "epoch": 1.6376283968274605, |
| "grad_norm": 1.4037001132965088, |
| "learning_rate": 6.7388785528874e-06, |
| "loss": 0.539414644241333, |
| "step": 1575 |
| }, |
| { |
| "epoch": 1.6428292809777663, |
| "grad_norm": 1.3790786266326904, |
| "learning_rate": 6.553116286511012e-06, |
| "loss": 0.5314173221588134, |
| "step": 1580 |
| }, |
| { |
| "epoch": 1.6480301651280718, |
| "grad_norm": 1.3834137916564941, |
| "learning_rate": 6.369721597702802e-06, |
| "loss": 0.5361592292785644, |
| "step": 1585 |
| }, |
| { |
| "epoch": 1.6532310492783773, |
| "grad_norm": 1.3719836473464966, |
| "learning_rate": 6.188707468274597e-06, |
| "loss": 0.5356816291809082, |
| "step": 1590 |
| }, |
| { |
| "epoch": 1.658431933428683, |
| "grad_norm": 1.5612096786499023, |
| "learning_rate": 6.010086711527519e-06, |
| "loss": 0.522746467590332, |
| "step": 1595 |
| }, |
| { |
| "epoch": 1.6636328175789883, |
| "grad_norm": 1.3957282304763794, |
| "learning_rate": 5.833871971344947e-06, |
| "loss": 0.5154781341552734, |
| "step": 1600 |
| }, |
| { |
| "epoch": 1.6636328175789883, |
| "eval_loss": 0.6775081157684326, |
| "eval_runtime": 129.5251, |
| "eval_samples_per_second": 1.976, |
| "eval_steps_per_second": 1.976, |
| "step": 1600 |
| }, |
| { |
| "epoch": 1.668833701729294, |
| "grad_norm": 1.3207283020019531, |
| "learning_rate": 5.6600757212975065e-06, |
| "loss": 0.5063146591186524, |
| "step": 1605 |
| }, |
| { |
| "epoch": 1.6740345858795995, |
| "grad_norm": 1.4711432456970215, |
| "learning_rate": 5.4887102637600995e-06, |
| "loss": 0.5405515193939209, |
| "step": 1610 |
| }, |
| { |
| "epoch": 1.679235470029905, |
| "grad_norm": 1.4866151809692383, |
| "learning_rate": 5.319787729041084e-06, |
| "loss": 0.5130389213562012, |
| "step": 1615 |
| }, |
| { |
| "epoch": 1.6844363541802108, |
| "grad_norm": 1.3158960342407227, |
| "learning_rate": 5.153320074523618e-06, |
| "loss": 0.5351908206939697, |
| "step": 1620 |
| }, |
| { |
| "epoch": 1.689637238330516, |
| "grad_norm": 1.3680565357208252, |
| "learning_rate": 4.9893190838192415e-06, |
| "loss": 0.4998152256011963, |
| "step": 1625 |
| }, |
| { |
| "epoch": 1.6948381224808218, |
| "grad_norm": 1.4631091356277466, |
| "learning_rate": 4.827796365933752e-06, |
| "loss": 0.5004712104797363, |
| "step": 1630 |
| }, |
| { |
| "epoch": 1.7000390066311273, |
| "grad_norm": 1.3971714973449707, |
| "learning_rate": 4.668763354445425e-06, |
| "loss": 0.5030968666076661, |
| "step": 1635 |
| }, |
| { |
| "epoch": 1.7052398907814328, |
| "grad_norm": 1.2639859914779663, |
| "learning_rate": 4.5122313066957355e-06, |
| "loss": 0.5040280818939209, |
| "step": 1640 |
| }, |
| { |
| "epoch": 1.7104407749317385, |
| "grad_norm": 1.3765854835510254, |
| "learning_rate": 4.358211302992414e-06, |
| "loss": 0.503818941116333, |
| "step": 1645 |
| }, |
| { |
| "epoch": 1.7156416590820438, |
| "grad_norm": 1.3985488414764404, |
| "learning_rate": 4.206714245825207e-06, |
| "loss": 0.5131355285644531, |
| "step": 1650 |
| }, |
| { |
| "epoch": 1.7208425432323495, |
| "grad_norm": 1.3251243829727173, |
| "learning_rate": 4.057750859094029e-06, |
| "loss": 0.5185832500457763, |
| "step": 1655 |
| }, |
| { |
| "epoch": 1.726043427382655, |
| "grad_norm": 1.3081040382385254, |
| "learning_rate": 3.911331687349957e-06, |
| "loss": 0.5041447639465332, |
| "step": 1660 |
| }, |
| { |
| "epoch": 1.7312443115329605, |
| "grad_norm": 1.2962945699691772, |
| "learning_rate": 3.7674670950487337e-06, |
| "loss": 0.49753355979919434, |
| "step": 1665 |
| }, |
| { |
| "epoch": 1.7364451956832663, |
| "grad_norm": 1.4440361261367798, |
| "learning_rate": 3.626167265817162e-06, |
| "loss": 0.5241962432861328, |
| "step": 1670 |
| }, |
| { |
| "epoch": 1.7416460798335716, |
| "grad_norm": 1.3910887241363525, |
| "learning_rate": 3.487442201732245e-06, |
| "loss": 0.5163005352020263, |
| "step": 1675 |
| }, |
| { |
| "epoch": 1.7468469639838773, |
| "grad_norm": 1.4579485654830933, |
| "learning_rate": 3.3513017226131537e-06, |
| "loss": 0.5024435520172119, |
| "step": 1680 |
| }, |
| { |
| "epoch": 1.7520478481341828, |
| "grad_norm": 1.3572936058044434, |
| "learning_rate": 3.217755465326127e-06, |
| "loss": 0.5091588020324707, |
| "step": 1685 |
| }, |
| { |
| "epoch": 1.7572487322844883, |
| "grad_norm": 1.3362730741500854, |
| "learning_rate": 3.0868128831023123e-06, |
| "loss": 0.4787764549255371, |
| "step": 1690 |
| }, |
| { |
| "epoch": 1.762449616434794, |
| "grad_norm": 1.5451667308807373, |
| "learning_rate": 2.9584832448686216e-06, |
| "loss": 0.5192002773284912, |
| "step": 1695 |
| }, |
| { |
| "epoch": 1.7676505005850993, |
| "grad_norm": 1.3013676404953003, |
| "learning_rate": 2.832775634591589e-06, |
| "loss": 0.4944888114929199, |
| "step": 1700 |
| }, |
| { |
| "epoch": 1.7676505005850993, |
| "eval_loss": 0.6653155088424683, |
| "eval_runtime": 129.5395, |
| "eval_samples_per_second": 1.976, |
| "eval_steps_per_second": 1.976, |
| "step": 1700 |
| }, |
| { |
| "epoch": 1.772851384735405, |
| "grad_norm": 1.4618582725524902, |
| "learning_rate": 2.709698950634412e-06, |
| "loss": 0.5045624732971191, |
| "step": 1705 |
| }, |
| { |
| "epoch": 1.7780522688857106, |
| "grad_norm": 1.4680557250976562, |
| "learning_rate": 2.5892619051269873e-06, |
| "loss": 0.5313224792480469, |
| "step": 1710 |
| }, |
| { |
| "epoch": 1.783253153036016, |
| "grad_norm": 1.441771149635315, |
| "learning_rate": 2.471473023349269e-06, |
| "loss": 0.5317369937896729, |
| "step": 1715 |
| }, |
| { |
| "epoch": 1.7884540371863218, |
| "grad_norm": 1.377977728843689, |
| "learning_rate": 2.356340643127797e-06, |
| "loss": 0.49757208824157717, |
| "step": 1720 |
| }, |
| { |
| "epoch": 1.793654921336627, |
| "grad_norm": 1.376712679862976, |
| "learning_rate": 2.243872914245442e-06, |
| "loss": 0.5022801399230957, |
| "step": 1725 |
| }, |
| { |
| "epoch": 1.7988558054869328, |
| "grad_norm": 1.5128220319747925, |
| "learning_rate": 2.1340777978646045e-06, |
| "loss": 0.5198845863342285, |
| "step": 1730 |
| }, |
| { |
| "epoch": 1.8040566896372383, |
| "grad_norm": 1.3930203914642334, |
| "learning_rate": 2.026963065963572e-06, |
| "loss": 0.502904748916626, |
| "step": 1735 |
| }, |
| { |
| "epoch": 1.8092575737875438, |
| "grad_norm": 1.3942198753356934, |
| "learning_rate": 1.9225363007864483e-06, |
| "loss": 0.48986196517944336, |
| "step": 1740 |
| }, |
| { |
| "epoch": 1.8144584579378495, |
| "grad_norm": 1.238763689994812, |
| "learning_rate": 1.8208048943063872e-06, |
| "loss": 0.48829169273376466, |
| "step": 1745 |
| }, |
| { |
| "epoch": 1.8196593420881548, |
| "grad_norm": 1.3654608726501465, |
| "learning_rate": 1.7217760477023659e-06, |
| "loss": 0.5089236259460449, |
| "step": 1750 |
| }, |
| { |
| "epoch": 1.8248602262384606, |
| "grad_norm": 1.3893874883651733, |
| "learning_rate": 1.6254567708494517e-06, |
| "loss": 0.5198930740356446, |
| "step": 1755 |
| }, |
| { |
| "epoch": 1.830061110388766, |
| "grad_norm": 1.4330559968948364, |
| "learning_rate": 1.531853881822567e-06, |
| "loss": 0.5197700500488281, |
| "step": 1760 |
| }, |
| { |
| "epoch": 1.8352619945390716, |
| "grad_norm": 1.4249515533447266, |
| "learning_rate": 1.4409740064138844e-06, |
| "loss": 0.502013874053955, |
| "step": 1765 |
| }, |
| { |
| "epoch": 1.8404628786893773, |
| "grad_norm": 1.3588781356811523, |
| "learning_rate": 1.3528235776638111e-06, |
| "loss": 0.48603029251098634, |
| "step": 1770 |
| }, |
| { |
| "epoch": 1.8456637628396826, |
| "grad_norm": 1.3336181640625, |
| "learning_rate": 1.2674088354056148e-06, |
| "loss": 0.5047059059143066, |
| "step": 1775 |
| }, |
| { |
| "epoch": 1.8508646469899883, |
| "grad_norm": 1.3933433294296265, |
| "learning_rate": 1.1847358258237329e-06, |
| "loss": 0.5210882663726807, |
| "step": 1780 |
| }, |
| { |
| "epoch": 1.8560655311402938, |
| "grad_norm": 1.4772483110427856, |
| "learning_rate": 1.104810401025791e-06, |
| "loss": 0.512470006942749, |
| "step": 1785 |
| }, |
| { |
| "epoch": 1.8612664152905993, |
| "grad_norm": 1.2868046760559082, |
| "learning_rate": 1.0276382186283285e-06, |
| "loss": 0.4919888973236084, |
| "step": 1790 |
| }, |
| { |
| "epoch": 1.866467299440905, |
| "grad_norm": 1.4422026872634888, |
| "learning_rate": 9.532247413563555e-07, |
| "loss": 0.5298680782318115, |
| "step": 1795 |
| }, |
| { |
| "epoch": 1.8716681835912103, |
| "grad_norm": 1.4196135997772217, |
| "learning_rate": 8.815752366566399e-07, |
| "loss": 0.5006773948669434, |
| "step": 1800 |
| }, |
| { |
| "epoch": 1.8716681835912103, |
| "eval_loss": 0.6575887203216553, |
| "eval_runtime": 130.8561, |
| "eval_samples_per_second": 1.956, |
| "eval_steps_per_second": 1.956, |
| "step": 1800 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 1924, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.897885910540295e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
