| { | |
| "best_metric": 0.03178340569138527, | |
| "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold4/checkpoint-1850", | |
| "epoch": 4.994228549442093, | |
| "eval_steps": 50, | |
| "global_step": 3245, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015390534821085032, | |
| "grad_norm": 3.720082998275757, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.9997, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 4.1043524742126465, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.9245, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.930521249771118, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.716, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 3.3468408584594727, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.378, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 2.080782413482666, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8709, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.6925050020217896, | |
| "eval_runtime": 182.7994, | |
| "eval_samples_per_second": 6.324, | |
| "eval_steps_per_second": 6.324, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 0.8935777544975281, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.59, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.2229251861572266, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.4074, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 0.6000741124153137, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.2131, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 0.9645354151725769, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.2181, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.6606675982475281, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1563, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.10799792408943176, | |
| "eval_runtime": 182.9398, | |
| "eval_samples_per_second": 6.319, | |
| "eval_steps_per_second": 6.319, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8444873094558716, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.1101, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.7903103828430176, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.1092, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.273266315460205, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0786, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.959189236164093, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0836, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.7009065747261047, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0875, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.07308054715394974, | |
| "eval_runtime": 183.0999, | |
| "eval_samples_per_second": 6.313, | |
| "eval_steps_per_second": 6.313, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.712720513343811, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0828, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 1.3354097604751587, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.0752, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.6948460340499878, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0595, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.6320282816886902, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0519, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.5802289247512817, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0732, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.0556374117732048, | |
| "eval_runtime": 183.0845, | |
| "eval_samples_per_second": 6.314, | |
| "eval_steps_per_second": 6.314, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.7064422965049744, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0641, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.7606561779975891, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0609, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.33633682131767273, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0464, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.7043708562850952, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0583, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.5120052099227905, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0549, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.05272047594189644, | |
| "eval_runtime": 183.1535, | |
| "eval_samples_per_second": 6.312, | |
| "eval_steps_per_second": 6.312, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.3508935868740082, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0526, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.815880537033081, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.0515, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.5120180249214172, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0574, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.3832987844944, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0619, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.44087255001068115, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0517, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.04957836493849754, | |
| "eval_runtime": 182.5082, | |
| "eval_samples_per_second": 6.334, | |
| "eval_steps_per_second": 6.334, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.4753435254096985, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.0577, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 1.0133485794067383, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0625, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.3338690996170044, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.0567, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.428244948387146, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0526, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.40518462657928467, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0464, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.047364190220832825, | |
| "eval_runtime": 182.7404, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 6.326, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.27636629343032837, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0586, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.4587933123111725, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0445, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.4389565885066986, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.0454, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.5720734596252441, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0418, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.2468489706516266, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0635, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.045144420117139816, | |
| "eval_runtime": 183.1142, | |
| "eval_samples_per_second": 6.313, | |
| "eval_steps_per_second": 6.313, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.1773131936788559, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0429, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.4588727355003357, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0535, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.5701034665107727, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.0472, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.34796223044395447, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0627, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.2773059010505676, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0357, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.0441606342792511, | |
| "eval_runtime": 183.2481, | |
| "eval_samples_per_second": 6.308, | |
| "eval_steps_per_second": 6.308, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.3536244034767151, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0368, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.26772475242614746, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.029, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5594661235809326, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0357, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.2639472186565399, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.0391, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.27159807085990906, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0298, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.040104784071445465, | |
| "eval_runtime": 183.3289, | |
| "eval_samples_per_second": 6.306, | |
| "eval_steps_per_second": 6.306, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.5231143832206726, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0312, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.178458109498024, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.049, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.4445004463195801, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0616, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.35495319962501526, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0425, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.2151714414358139, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.047, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.03841087967157364, | |
| "eval_runtime": 182.9781, | |
| "eval_samples_per_second": 6.318, | |
| "eval_steps_per_second": 6.318, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.20624691247940063, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.0317, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.33963027596473694, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.0322, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.346618115901947, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0383, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.2768615782260895, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0549, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.31985950469970703, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0384, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.037438128143548965, | |
| "eval_runtime": 182.1789, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 6.345, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.2605275809764862, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0356, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.18158534169197083, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.0395, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.35811084508895874, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.0433, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.6907298564910889, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0347, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.5655986070632935, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0318, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.03586694970726967, | |
| "eval_runtime": 182.4839, | |
| "eval_samples_per_second": 6.335, | |
| "eval_steps_per_second": 6.335, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.46735772490501404, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0375, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.4806405305862427, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0364, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.2836764454841614, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0339, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.46108153462409973, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.047, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.23082521557807922, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0362, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.03821026161313057, | |
| "eval_runtime": 182.862, | |
| "eval_samples_per_second": 6.322, | |
| "eval_steps_per_second": 6.322, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.3194074034690857, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.0316, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.1818566769361496, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0262, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.2859642207622528, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.0325, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.5254983305931091, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0379, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.2477121204137802, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.0296, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.03541037067770958, | |
| "eval_runtime": 183.1717, | |
| "eval_samples_per_second": 6.311, | |
| "eval_steps_per_second": 6.311, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.24987374246120453, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.0382, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.24398306012153625, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.0388, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.19983555376529694, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0327, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.29734015464782715, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0338, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.2929656505584717, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.0305, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.03470904752612114, | |
| "eval_runtime": 183.4668, | |
| "eval_samples_per_second": 6.301, | |
| "eval_steps_per_second": 6.301, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.16526123881340027, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.0368, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.22322843968868256, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0265, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.2852199375629425, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.031, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.4977399408817291, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0292, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.665130615234375, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.0333, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.035344552248716354, | |
| "eval_runtime": 183.1247, | |
| "eval_samples_per_second": 6.313, | |
| "eval_steps_per_second": 6.313, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.18849772214889526, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.0332, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.32548508048057556, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0307, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.12678252160549164, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0277, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.1886940598487854, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.036, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.24736231565475464, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0205, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03444148227572441, | |
| "eval_runtime": 182.6694, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 6.328, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.24484626948833466, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0301, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.2631525695323944, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0388, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.2534925639629364, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.027, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.4256260395050049, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0352, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.09553076326847076, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.041, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.03463003784418106, | |
| "eval_runtime": 182.9048, | |
| "eval_samples_per_second": 6.32, | |
| "eval_steps_per_second": 6.32, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.31411033868789673, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.0307, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.10352231562137604, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0244, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.3034602403640747, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.0338, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.21614131331443787, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0335, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.25026729702949524, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0321, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.03277269005775452, | |
| "eval_runtime": 183.2642, | |
| "eval_samples_per_second": 6.308, | |
| "eval_steps_per_second": 6.308, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.20271597802639008, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.0357, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.25522804260253906, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0396, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.150895893573761, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.0305, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.11997763067483902, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.0328, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.25409039855003357, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0332, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.03351612761616707, | |
| "eval_runtime": 183.4659, | |
| "eval_samples_per_second": 6.301, | |
| "eval_steps_per_second": 6.301, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.19263124465942383, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.0269, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.16586606204509735, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.0283, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 0.16381381452083588, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.025, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.488243967294693, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.03, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.19238832592964172, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.0326, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.03455475717782974, | |
| "eval_runtime": 183.3613, | |
| "eval_samples_per_second": 6.304, | |
| "eval_steps_per_second": 6.304, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.45936962962150574, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0368, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.20919299125671387, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.027, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.2479189932346344, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.0401, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.14464621245861053, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0472, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.5745062232017517, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0217, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.033430181443691254, | |
| "eval_runtime": 183.4179, | |
| "eval_samples_per_second": 6.303, | |
| "eval_steps_per_second": 6.303, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.15858586132526398, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.03, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.16054297983646393, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.0241, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.17755919694900513, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0238, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.3623185157775879, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.0381, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.08923501521348953, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0212, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.03504960238933563, | |
| "eval_runtime": 182.729, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 6.326, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.5454447269439697, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.031, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.12969723343849182, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0262, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.10606259107589722, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0299, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.2240438163280487, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.0359, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.3463599681854248, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.0297, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.03407027944922447, | |
| "eval_runtime": 183.0464, | |
| "eval_samples_per_second": 6.315, | |
| "eval_steps_per_second": 6.315, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.33675724267959595, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.0312, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.2068137675523758, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.0288, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.16806113719940186, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.0236, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.3267403841018677, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0323, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.2041923701763153, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.0217, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.031942326575517654, | |
| "eval_runtime": 183.3435, | |
| "eval_samples_per_second": 6.305, | |
| "eval_steps_per_second": 6.305, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.24685949087142944, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.0218, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.309602826833725, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.023, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.13260765373706818, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0228, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.16125592589378357, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.02, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.24640512466430664, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0162, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.0336560420691967, | |
| "eval_runtime": 183.6429, | |
| "eval_samples_per_second": 6.295, | |
| "eval_steps_per_second": 6.295, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.12094150483608246, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0208, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.5082912445068359, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.0197, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.277658075094223, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.0134, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.20811834931373596, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0157, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.5055046081542969, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0313, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.034869175404310226, | |
| "eval_runtime": 183.6997, | |
| "eval_samples_per_second": 6.293, | |
| "eval_steps_per_second": 6.293, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.10814691334962845, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0152, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.30697014927864075, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.025, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.3481878936290741, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.016, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.16028915345668793, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0208, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.10776527225971222, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.0132, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.03573461249470711, | |
| "eval_runtime": 183.3063, | |
| "eval_samples_per_second": 6.306, | |
| "eval_steps_per_second": 6.306, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.23579591512680054, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.019, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.3954004645347595, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0231, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.25841882824897766, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0233, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.1326594203710556, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.0242, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.2397865206003189, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0184, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.033922333270311356, | |
| "eval_runtime": 172.2025, | |
| "eval_samples_per_second": 6.713, | |
| "eval_steps_per_second": 6.713, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.21709144115447998, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.0214, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.12962335348129272, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0233, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.20671947300434113, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.0178, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.38715389370918274, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0195, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.1862838715314865, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.0156, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.03550899028778076, | |
| "eval_runtime": 156.0386, | |
| "eval_samples_per_second": 7.408, | |
| "eval_steps_per_second": 7.408, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.22983674705028534, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.0133, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.424269437789917, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0253, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.15247733891010284, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0212, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.22490614652633667, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.019, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.3363230526447296, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.0137, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.03481822833418846, | |
| "eval_runtime": 155.5945, | |
| "eval_samples_per_second": 7.43, | |
| "eval_steps_per_second": 7.43, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.30516526103019714, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.0161, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.1704840511083603, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.0209, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.18374107778072357, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0205, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.3123207092285156, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.0197, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.23016874492168427, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.0216, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.0334537997841835, | |
| "eval_runtime": 155.4657, | |
| "eval_samples_per_second": 7.436, | |
| "eval_steps_per_second": 7.436, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.39230310916900635, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.0178, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.3503061830997467, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0197, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.23184560239315033, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0232, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.4075738191604614, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0227, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.6792215704917908, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0195, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.032654643058776855, | |
| "eval_runtime": 155.3228, | |
| "eval_samples_per_second": 7.443, | |
| "eval_steps_per_second": 7.443, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.2639402151107788, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.019, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.1913764923810959, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.0314, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.3856119215488434, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.0245, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.11779230833053589, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.0173, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.16009725630283356, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0262, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.03214823827147484, | |
| "eval_runtime": 155.2086, | |
| "eval_samples_per_second": 7.448, | |
| "eval_steps_per_second": 7.448, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.20571549236774445, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0161, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.45876345038414, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0223, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.3662247061729431, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0257, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.1181124597787857, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.0193, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.30701759457588196, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.0252, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.0339156873524189, | |
| "eval_runtime": 155.2003, | |
| "eval_samples_per_second": 7.448, | |
| "eval_steps_per_second": 7.448, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.2601530849933624, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.0198, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.4243848919868469, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0252, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.518486738204956, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.023, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.25651493668556213, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.0264, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.21836812794208527, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.0248, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.03178340569138527, | |
| "eval_runtime": 155.2466, | |
| "eval_samples_per_second": 7.446, | |
| "eval_steps_per_second": 7.446, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.25536152720451355, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0151, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.2502846419811249, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0262, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.5007121562957764, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0134, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.19754254817962646, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.018, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.24531638622283936, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0173, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.033025115728378296, | |
| "eval_runtime": 155.1444, | |
| "eval_samples_per_second": 7.451, | |
| "eval_steps_per_second": 7.451, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 0.2479398548603058, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.0222, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.3831672668457031, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.018, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.13186602294445038, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0169, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.16748853027820587, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0138, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.24280278384685516, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.0285, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.03332563117146492, | |
| "eval_runtime": 155.1412, | |
| "eval_samples_per_second": 7.451, | |
| "eval_steps_per_second": 7.451, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.1093946024775505, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.0068, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.17604872584342957, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.0111, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.3216577172279358, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.0102, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.32190221548080444, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.0082, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.15662884712219238, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.0112, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.036320120096206665, | |
| "eval_runtime": 155.0278, | |
| "eval_samples_per_second": 7.457, | |
| "eval_steps_per_second": 7.457, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 0.7870141863822937, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.0069, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.060445357114076614, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0116, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.27109313011169434, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.012, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.13197797536849976, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.009, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.3487105071544647, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0091, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.03961965814232826, | |
| "eval_runtime": 155.0517, | |
| "eval_samples_per_second": 7.456, | |
| "eval_steps_per_second": 7.456, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.25426051020622253, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.0073, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 0.3215971887111664, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.0099, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.3518652319908142, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0136, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.37935787439346313, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.0097, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.3207817077636719, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.0086, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.03981826826930046, | |
| "eval_runtime": 155.0762, | |
| "eval_samples_per_second": 7.454, | |
| "eval_steps_per_second": 7.454, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.10169580578804016, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.0101, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.3294033706188202, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0117, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.3449319899082184, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0074, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.2027152180671692, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0055, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.4143899083137512, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0086, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.04120039939880371, | |
| "eval_runtime": 155.054, | |
| "eval_samples_per_second": 7.455, | |
| "eval_steps_per_second": 7.455, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.1887126863002777, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.0108, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.660940408706665, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.015, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.09298793971538544, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.0094, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.4668661653995514, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.0094, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.09208986163139343, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0156, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.03979753330349922, | |
| "eval_runtime": 155.1054, | |
| "eval_samples_per_second": 7.453, | |
| "eval_steps_per_second": 7.453, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.19043681025505066, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.0105, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.27596035599708557, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0095, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.2940841317176819, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.0124, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.20040611922740936, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0135, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.2633702754974365, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0097, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.03873945027589798, | |
| "eval_runtime": 155.0752, | |
| "eval_samples_per_second": 7.454, | |
| "eval_steps_per_second": 7.454, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.12973666191101074, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0092, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.16432422399520874, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.006, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.21306325495243073, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0139, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.518015444278717, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0076, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.255136102437973, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0103, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.03970556706190109, | |
| "eval_runtime": 154.977, | |
| "eval_samples_per_second": 7.459, | |
| "eval_steps_per_second": 7.459, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.0744418203830719, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0134, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.5227424502372742, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.0123, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.1410445272922516, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0083, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.2641276717185974, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0118, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.10524629801511765, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0059, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.04078695550560951, | |
| "eval_runtime": 155.0702, | |
| "eval_samples_per_second": 7.455, | |
| "eval_steps_per_second": 7.455, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.31813839077949524, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0124, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.21312487125396729, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.0066, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.22191260755062103, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.0113, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.30453941226005554, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.013, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.19308875501155853, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.0122, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.0401541143655777, | |
| "eval_runtime": 154.8862, | |
| "eval_samples_per_second": 7.464, | |
| "eval_steps_per_second": 7.464, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.5179212093353271, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.0097, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.2972368896007538, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0065, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.08873302489519119, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.009, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.3741326332092285, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0074, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.20551709830760956, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0075, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.04080299288034439, | |
| "eval_runtime": 154.9803, | |
| "eval_samples_per_second": 7.459, | |
| "eval_steps_per_second": 7.459, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.08597930520772934, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0098, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.24366773664951324, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0052, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.11822346597909927, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.0111, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.23733307421207428, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0111, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.2179519385099411, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.0104, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.04115796834230423, | |
| "eval_runtime": 155.0812, | |
| "eval_samples_per_second": 7.454, | |
| "eval_steps_per_second": 7.454, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.029365114867687225, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0086, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.36178985238075256, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0043, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.043734073638916016, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.0086, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.6358830332756042, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.0072, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.5515215992927551, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.0128, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.04165811836719513, | |
| "eval_runtime": 155.0845, | |
| "eval_samples_per_second": 7.454, | |
| "eval_steps_per_second": 7.454, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.4808307886123657, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0077, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.4030892550945282, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0053, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 1.4124490022659302, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0162, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.4085770547389984, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.0068, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.0908542349934578, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.0094, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.040961481630802155, | |
| "eval_runtime": 154.9989, | |
| "eval_samples_per_second": 7.458, | |
| "eval_steps_per_second": 7.458, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.12777447700500488, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0049, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.12393078953027725, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0018, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.11471406370401382, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.0026, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.024372773244976997, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0057, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.047470737248659134, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.0016, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.0423700213432312, | |
| "eval_runtime": 154.914, | |
| "eval_samples_per_second": 7.462, | |
| "eval_steps_per_second": 7.462, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.3048446774482727, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0046, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.21287034451961517, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.0045, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.0366392582654953, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0026, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.04145758971571922, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0047, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.039583995938301086, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0045, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.045189324766397476, | |
| "eval_runtime": 154.868, | |
| "eval_samples_per_second": 7.464, | |
| "eval_steps_per_second": 7.464, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.30602163076400757, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.0051, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.31484103202819824, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0043, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.041927698999643326, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0027, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 0.29563918709754944, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.0059, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.08490947633981705, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0027, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.046041812747716904, | |
| "eval_runtime": 154.9521, | |
| "eval_samples_per_second": 7.46, | |
| "eval_steps_per_second": 7.46, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.0900108739733696, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0023, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.023624056950211525, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.0018, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.3250238597393036, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.0031, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.08482615649700165, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0035, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.019452480599284172, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.0027, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.04714136943221092, | |
| "eval_runtime": 154.8788, | |
| "eval_samples_per_second": 7.464, | |
| "eval_steps_per_second": 7.464, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.023408198729157448, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0018, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.05028143152594566, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0028, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.13757863640785217, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.0033, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.32358190417289734, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0031, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.03358887881040573, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0036, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.04825592041015625, | |
| "eval_runtime": 154.8564, | |
| "eval_samples_per_second": 7.465, | |
| "eval_steps_per_second": 7.465, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.1749475747346878, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.002, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.07079064100980759, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.0012, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.07326746731996536, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0091, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.030964983627200127, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0014, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.3133019506931305, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0034, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.04851840063929558, | |
| "eval_runtime": 154.8125, | |
| "eval_samples_per_second": 7.467, | |
| "eval_steps_per_second": 7.467, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.07805794477462769, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.0071, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.06846173107624054, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0043, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.036680787801742554, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.0024, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.46223872900009155, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0035, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.018994424492120743, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0021, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.04854997619986534, | |
| "eval_runtime": 154.9074, | |
| "eval_samples_per_second": 7.463, | |
| "eval_steps_per_second": 7.463, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.05986038222908974, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0008, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 2.833479642868042, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.003, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.02678421512246132, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.0024, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.4688439965248108, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.0038, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 0.1824573129415512, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.0025, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.049106284976005554, | |
| "eval_runtime": 154.8445, | |
| "eval_samples_per_second": 7.466, | |
| "eval_steps_per_second": 7.466, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.2514187693595886, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0019, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.10211896151304245, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0037, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.22848080098628998, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.0062, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.5508701205253601, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0028, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.013921206817030907, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0022, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.04922712221741676, | |
| "eval_runtime": 154.8961, | |
| "eval_samples_per_second": 7.463, | |
| "eval_steps_per_second": 7.463, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.10601812601089478, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.004, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.48500388860702515, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0038, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.535514771938324, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0053, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.06666892021894455, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.0039, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.0649837926030159, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.0038, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.049353163689374924, | |
| "eval_runtime": 154.9618, | |
| "eval_samples_per_second": 7.46, | |
| "eval_steps_per_second": 7.46, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.0760839581489563, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.0062, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.2534802556037903, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.0015, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.04318990558385849, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0035, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.16355225443840027, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0019, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.28926822543144226, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0057, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.04941301792860031, | |
| "eval_runtime": 155.0884, | |
| "eval_samples_per_second": 7.454, | |
| "eval_steps_per_second": 7.454, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.11070690304040909, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0038, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.33111050724983215, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.0028, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.06555396318435669, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.003, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.17634208500385284, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0024, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.06867170333862305, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.0028, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.049279745668172836, | |
| "eval_runtime": 155.0371, | |
| "eval_samples_per_second": 7.456, | |
| "eval_steps_per_second": 7.456, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.03979424387216568, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0033, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.01702362857758999, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.0031, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.360327810049057, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0055, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.1880173534154892, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0037, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 7.696760527880847e+17, | |
| "train_loss": 0.0513197087359906, | |
| "train_runtime": 39476.3181, | |
| "train_samples_per_second": 1.317, | |
| "train_steps_per_second": 0.082 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3245, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.696760527880847e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |