| { | |
| "best_metric": 0.0313267707824707, | |
| "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold6/checkpoint-1350", | |
| "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": 2.7129907608032227, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.3358, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 2.8113174438476562, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.2708, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.1741812229156494, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0352, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 2.263507843017578, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.5544, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 1.1408025026321411, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2584, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.241627037525177, | |
| "eval_runtime": 176.7851, | |
| "eval_samples_per_second": 6.539, | |
| "eval_steps_per_second": 6.539, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 1.4696407318115234, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.1594, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.216342568397522, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.1273, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 1.221793532371521, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.1015, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 1.4902610778808594, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.1184, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.7272764444351196, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0852, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.06952265650033951, | |
| "eval_runtime": 176.7102, | |
| "eval_samples_per_second": 6.542, | |
| "eval_steps_per_second": 6.542, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8303190469741821, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.0646, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.8325642347335815, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.0725, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.1191056966781616, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0688, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.7813414931297302, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0791, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.9104443192481995, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0619, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.058522578328847885, | |
| "eval_runtime": 176.6753, | |
| "eval_samples_per_second": 6.543, | |
| "eval_steps_per_second": 6.543, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.7682419419288635, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0617, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 0.7752271890640259, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.0614, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.8143901228904724, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0526, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.36231929063796997, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0442, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.5930745601654053, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0573, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.05442158505320549, | |
| "eval_runtime": 176.7184, | |
| "eval_samples_per_second": 6.541, | |
| "eval_steps_per_second": 6.541, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.7608053088188171, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0553, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.8498407602310181, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0504, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3562508821487427, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0385, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.44236478209495544, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0459, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.5377424359321594, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0447, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.04289978742599487, | |
| "eval_runtime": 176.7719, | |
| "eval_samples_per_second": 6.539, | |
| "eval_steps_per_second": 6.539, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.595135509967804, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0404, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.39058539271354675, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.0412, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.5041882991790771, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0539, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.427709698677063, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0504, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.4061878025531769, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0395, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.046997975558042526, | |
| "eval_runtime": 176.7958, | |
| "eval_samples_per_second": 6.539, | |
| "eval_steps_per_second": 6.539, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.5137187242507935, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.0518, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 0.3105601370334625, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0459, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.4471800923347473, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.0477, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.3897547721862793, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0445, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.4644322395324707, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0436, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.04534762352705002, | |
| "eval_runtime": 176.7398, | |
| "eval_samples_per_second": 6.541, | |
| "eval_steps_per_second": 6.541, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.3490193784236908, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0559, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.41950008273124695, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0369, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.528809666633606, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.0376, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.4884808659553528, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0351, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.2686597406864166, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0482, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.04443967342376709, | |
| "eval_runtime": 176.807, | |
| "eval_samples_per_second": 6.538, | |
| "eval_steps_per_second": 6.538, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.28282681107521057, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0421, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.486635684967041, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0469, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.4719170331954956, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.0432, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.4644046723842621, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0583, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.3970758616924286, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0291, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.038791827857494354, | |
| "eval_runtime": 176.9416, | |
| "eval_samples_per_second": 6.533, | |
| "eval_steps_per_second": 6.533, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.30748650431632996, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0359, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.4109121263027191, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.031, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5491299629211426, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0395, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.2228752225637436, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.0455, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.31454986333847046, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0291, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.03798589110374451, | |
| "eval_runtime": 176.9376, | |
| "eval_samples_per_second": 6.533, | |
| "eval_steps_per_second": 6.533, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.4372515380382538, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0304, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.40634268522262573, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.0423, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.3712210953235626, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0537, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.3416385352611542, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0402, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.19281573593616486, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.0421, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.03445535525679588, | |
| "eval_runtime": 176.9207, | |
| "eval_samples_per_second": 6.534, | |
| "eval_steps_per_second": 6.534, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.25675076246261597, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.0356, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.20244090259075165, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.026, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.26542899012565613, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0334, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.25318223237991333, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0512, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.31446531414985657, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0363, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.03572468087077141, | |
| "eval_runtime": 177.0724, | |
| "eval_samples_per_second": 6.528, | |
| "eval_steps_per_second": 6.528, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.3364261984825134, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0322, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.3238294720649719, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.044, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.44701728224754333, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.043, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.23885679244995117, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0332, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.378052294254303, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0313, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.03557116910815239, | |
| "eval_runtime": 176.9535, | |
| "eval_samples_per_second": 6.533, | |
| "eval_steps_per_second": 6.533, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.372465044260025, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0337, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.23820656538009644, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0309, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.22136415541172028, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0333, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.5092695355415344, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.0441, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.3982639014720917, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0354, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.0369330570101738, | |
| "eval_runtime": 177.0125, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.38318920135498047, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.0323, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.2556602954864502, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0279, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.3347822427749634, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.029, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.5283674001693726, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0373, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.3553995192050934, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.0264, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.03539222106337547, | |
| "eval_runtime": 176.9814, | |
| "eval_samples_per_second": 6.532, | |
| "eval_steps_per_second": 6.532, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.25216934084892273, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.0341, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.21336272358894348, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.036, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.40105733275413513, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0339, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.3879629075527191, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0344, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.31842857599258423, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.028, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.034204475581645966, | |
| "eval_runtime": 177.0583, | |
| "eval_samples_per_second": 6.529, | |
| "eval_steps_per_second": 6.529, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.1974581778049469, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.0295, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.3911699056625366, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0272, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.2721879780292511, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.033, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.3825089633464813, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0259, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.36365213990211487, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.029, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.03416259214282036, | |
| "eval_runtime": 177.0108, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.2540026903152466, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.0264, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.26534292101860046, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0242, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.18031062185764313, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0301, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.30301544070243835, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.0382, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.31646642088890076, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0209, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03293142840266228, | |
| "eval_runtime": 176.9732, | |
| "eval_samples_per_second": 6.532, | |
| "eval_steps_per_second": 6.532, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.2676602303981781, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0301, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.31470590829849243, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0312, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.235820010304451, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.0282, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.538521945476532, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0299, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.13776828348636627, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.036, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.03404444456100464, | |
| "eval_runtime": 176.9806, | |
| "eval_samples_per_second": 6.532, | |
| "eval_steps_per_second": 6.532, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.2912769615650177, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.0295, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.11564439535140991, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0226, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.34356001019477844, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.0303, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.28519406914711, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0309, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.33766573667526245, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0319, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.03423403203487396, | |
| "eval_runtime": 176.9956, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.48815008997917175, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.0328, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.3189893960952759, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0329, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.24951376020908356, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.0285, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.42252716422080994, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.0267, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.24648497998714447, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0316, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.03275367617607117, | |
| "eval_runtime": 176.8832, | |
| "eval_samples_per_second": 6.535, | |
| "eval_steps_per_second": 6.535, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.17964158952236176, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.023, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.2720399498939514, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.028, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 1.056227684020996, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.0246, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.6856677532196045, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.0292, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.14827373623847961, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.0331, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.03661814332008362, | |
| "eval_runtime": 177.1274, | |
| "eval_samples_per_second": 6.526, | |
| "eval_steps_per_second": 6.526, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.4225656986236572, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0314, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.3269042372703552, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.0229, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.40620213747024536, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.0339, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.24237732589244843, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0389, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.16498889029026031, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0221, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.03357882797718048, | |
| "eval_runtime": 177.0586, | |
| "eval_samples_per_second": 6.529, | |
| "eval_steps_per_second": 6.529, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.162587508559227, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.0344, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.24337506294250488, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.0225, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.27764126658439636, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0248, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.3477858603000641, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.0294, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.14060847461223602, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0215, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.03731633350253105, | |
| "eval_runtime": 177.0625, | |
| "eval_samples_per_second": 6.529, | |
| "eval_steps_per_second": 6.529, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.3935520350933075, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.0283, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.17700353264808655, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0218, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.1391867846250534, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0283, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.3383353352546692, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.0337, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.39419957995414734, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.028, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.03271936997771263, | |
| "eval_runtime": 177.1043, | |
| "eval_samples_per_second": 6.527, | |
| "eval_steps_per_second": 6.527, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.6072343587875366, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.0295, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.23118914663791656, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.0307, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.19144384562969208, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.0237, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.2098597139120102, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0333, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.12983718514442444, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.0216, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.031436022371053696, | |
| "eval_runtime": 177.0238, | |
| "eval_samples_per_second": 6.53, | |
| "eval_steps_per_second": 6.53, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.19415806233882904, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.0176, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.3132944405078888, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.017, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.1949128657579422, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0181, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.07456086575984955, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.0203, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.13869090378284454, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0173, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.0313267707824707, | |
| "eval_runtime": 177.0027, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.13352008163928986, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0181, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.5723903179168701, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.016, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.20875513553619385, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.013, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.12426389753818512, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0143, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.3997185528278351, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0224, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.034030597656965256, | |
| "eval_runtime": 177.0265, | |
| "eval_samples_per_second": 6.53, | |
| "eval_steps_per_second": 6.53, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.08424781262874603, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0156, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.29116368293762207, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0216, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.15509343147277832, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.0115, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.447714239358902, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0206, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.12527887523174286, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.0096, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.038199178874492645, | |
| "eval_runtime": 177.0817, | |
| "eval_samples_per_second": 6.528, | |
| "eval_steps_per_second": 6.528, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.3768637478351593, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.0197, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.0920100063085556, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0216, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.3017866909503937, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0193, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.14904329180717468, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.0198, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.6704458594322205, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0191, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.0345410592854023, | |
| "eval_runtime": 176.9434, | |
| "eval_samples_per_second": 6.533, | |
| "eval_steps_per_second": 6.533, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.16943538188934326, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.0219, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.1667422503232956, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0162, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.09347015619277954, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.0149, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.2861078083515167, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0189, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.36650845408439636, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.0153, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.03815247118473053, | |
| "eval_runtime": 176.9191, | |
| "eval_samples_per_second": 6.534, | |
| "eval_steps_per_second": 6.534, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.2632177770137787, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.0157, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.41745585203170776, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0167, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.11307290941476822, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0178, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.30142202973365784, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.0174, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.22189970314502716, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.0133, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.03709259256720543, | |
| "eval_runtime": 176.98, | |
| "eval_samples_per_second": 6.532, | |
| "eval_steps_per_second": 6.532, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.2833783030509949, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.0173, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.26689910888671875, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.02, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.2401566207408905, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0154, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.41448310017585754, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.021, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.1988029032945633, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.0176, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.03363857790827751, | |
| "eval_runtime": 177.028, | |
| "eval_samples_per_second": 6.53, | |
| "eval_steps_per_second": 6.53, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.30356597900390625, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.0134, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.38142576813697815, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0171, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.5134309530258179, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0209, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.18491867184638977, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0206, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.3845498561859131, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0207, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.03425337374210358, | |
| "eval_runtime": 177.021, | |
| "eval_samples_per_second": 6.53, | |
| "eval_steps_per_second": 6.53, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.4090877175331116, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.0178, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.46939918398857117, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.0242, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.4078204035758972, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.0183, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.14355067908763885, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.0129, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.09979858994483948, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0232, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.03321119770407677, | |
| "eval_runtime": 177.0116, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.26243603229522705, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0127, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.347381055355072, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0182, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.5351781249046326, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0235, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.0745101198554039, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.016, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.3060047924518585, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.018, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.03361698240041733, | |
| "eval_runtime": 176.9895, | |
| "eval_samples_per_second": 6.531, | |
| "eval_steps_per_second": 6.531, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.48047566413879395, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.019, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.4708353281021118, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0232, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.4504677951335907, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.0205, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.2111469954252243, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.0196, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.23958446085453033, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.0196, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.032700322568416595, | |
| "eval_runtime": 177.2481, | |
| "eval_samples_per_second": 6.522, | |
| "eval_steps_per_second": 6.522, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.2881019115447998, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0141, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.29338282346725464, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0245, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.1548585146665573, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0139, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.1375659704208374, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.0166, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.2031036913394928, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0168, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.03366805613040924, | |
| "eval_runtime": 177.1654, | |
| "eval_samples_per_second": 6.525, | |
| "eval_steps_per_second": 6.525, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 0.5966081023216248, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.0212, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.45410460233688354, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.0195, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.09344927221536636, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0153, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.39616525173187256, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0135, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.15546825528144836, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.0226, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.03499778360128403, | |
| "eval_runtime": 177.0752, | |
| "eval_samples_per_second": 6.528, | |
| "eval_steps_per_second": 6.528, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.06762529164552689, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.0061, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.09853676706552505, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.0109, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.23364971578121185, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.0082, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.3123812973499298, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.006, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.12991108000278473, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.0087, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.03811512142419815, | |
| "eval_runtime": 177.296, | |
| "eval_samples_per_second": 6.52, | |
| "eval_steps_per_second": 6.52, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 0.3749006390571594, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.0052, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.11496545374393463, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0095, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.09478363394737244, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.009, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.11217588186264038, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.0086, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.8163807988166809, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0109, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.042027898132801056, | |
| "eval_runtime": 177.3243, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.4078664481639862, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.0045, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 0.33415350317955017, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.0071, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.24102692306041718, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0103, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.3822549283504486, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.0087, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.2982684075832367, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.0093, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.039782147854566574, | |
| "eval_runtime": 177.2325, | |
| "eval_samples_per_second": 6.523, | |
| "eval_steps_per_second": 6.523, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.08719394356012344, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.0078, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.41318830847740173, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0086, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.45956486463546753, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0083, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.08652245998382568, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0067, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.3944852948188782, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0063, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.04194685071706772, | |
| "eval_runtime": 177.3503, | |
| "eval_samples_per_second": 6.518, | |
| "eval_steps_per_second": 6.518, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.12498008459806442, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.0084, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.5928575992584229, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.0089, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.11140237003564835, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.0072, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.1740739494562149, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.0061, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.12572453916072845, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0131, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.03909638524055481, | |
| "eval_runtime": 177.3836, | |
| "eval_samples_per_second": 6.517, | |
| "eval_steps_per_second": 6.517, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.11590702086687088, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.0069, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.33815303444862366, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0064, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.18626803159713745, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.012, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.11907946318387985, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0104, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.18667781352996826, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0076, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.0381210520863533, | |
| "eval_runtime": 177.2897, | |
| "eval_samples_per_second": 6.52, | |
| "eval_steps_per_second": 6.52, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.03888513520359993, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0057, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.26933878660202026, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.0052, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.09857998043298721, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0086, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.3500402271747589, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0086, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.14311885833740234, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0102, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.039107028394937515, | |
| "eval_runtime": 177.3496, | |
| "eval_samples_per_second": 6.518, | |
| "eval_steps_per_second": 6.518, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.030803972855210304, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0089, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.47256898880004883, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.0092, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.15510740876197815, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0066, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.8535168170928955, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0119, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.15884172916412354, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0056, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.04255598038434982, | |
| "eval_runtime": 177.4359, | |
| "eval_samples_per_second": 6.515, | |
| "eval_steps_per_second": 6.515, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.3383089601993561, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0155, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.18983083963394165, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.0087, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.17381636798381805, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.0093, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.27495548129081726, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.0125, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.20102736353874207, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.011, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.03723160922527313, | |
| "eval_runtime": 177.2814, | |
| "eval_samples_per_second": 6.521, | |
| "eval_steps_per_second": 6.521, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.07937626540660858, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.0088, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.09707421064376831, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0061, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.03899567574262619, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.0085, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.10247146338224411, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0054, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.028264617547392845, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0062, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.03944387286901474, | |
| "eval_runtime": 177.3407, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.3208552598953247, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0058, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.5241620540618896, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0047, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.12019523233175278, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.0118, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.5147708058357239, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0085, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.15229442715644836, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.0071, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.04138202220201492, | |
| "eval_runtime": 177.3229, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.06029895693063736, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0097, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.2357356995344162, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0046, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.03440342843532562, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.0063, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.07327036559581757, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.0042, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.5318852066993713, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.009, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.0410628505051136, | |
| "eval_runtime": 177.3171, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.5346488952636719, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0078, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.17273636162281036, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0073, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 0.013756945729255676, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0114, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.3017001152038574, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.008, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.026331551373004913, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.009, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.03899988904595375, | |
| "eval_runtime": 177.2513, | |
| "eval_samples_per_second": 6.522, | |
| "eval_steps_per_second": 6.522, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.033303115516901016, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0032, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.12996241450309753, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0026, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.1452363282442093, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.002, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.021251849830150604, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0053, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.023803936317563057, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.002, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.04099424183368683, | |
| "eval_runtime": 177.3558, | |
| "eval_samples_per_second": 6.518, | |
| "eval_steps_per_second": 6.518, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.16137996315956116, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0053, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.046762533485889435, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.002, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.06787025928497314, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0028, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.010515443980693817, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0066, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.04508398845791817, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0072, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.043692074716091156, | |
| "eval_runtime": 177.3839, | |
| "eval_samples_per_second": 6.517, | |
| "eval_steps_per_second": 6.517, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.0291841272264719, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.0047, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.08017151802778244, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0085, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.042719267308712006, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0025, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 0.06686729192733765, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.0049, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.03811721131205559, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0018, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.04406580701470375, | |
| "eval_runtime": 177.3982, | |
| "eval_samples_per_second": 6.516, | |
| "eval_steps_per_second": 6.516, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.166035458445549, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0023, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.04751814901828766, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.0035, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.10477123409509659, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.0024, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.0692051351070404, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0033, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.0680622011423111, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.0022, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.045887645334005356, | |
| "eval_runtime": 177.341, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.10797956585884094, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0027, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.015229018405079842, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0015, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.05744533613324165, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.0019, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.0408879816532135, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0027, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.030890140682458878, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0023, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.0471736304461956, | |
| "eval_runtime": 177.3274, | |
| "eval_samples_per_second": 6.519, | |
| "eval_steps_per_second": 6.519, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.27945175766944885, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.0023, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.020555149763822556, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.0007, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.034013085067272186, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0075, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.01194527093321085, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0026, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.35130780935287476, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0042, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.04759151116013527, | |
| "eval_runtime": 177.3898, | |
| "eval_samples_per_second": 6.517, | |
| "eval_steps_per_second": 6.517, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.03287435695528984, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.0076, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.16559340059757233, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0037, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.03446711599826813, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.003, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.2722172737121582, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0056, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.01754995808005333, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0013, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.04682955518364906, | |
| "eval_runtime": 177.4266, | |
| "eval_samples_per_second": 6.515, | |
| "eval_steps_per_second": 6.515, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.27391576766967773, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0015, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 0.1101820319890976, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.0065, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.024084176868200302, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.002, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.4284050166606903, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.0024, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 1.1096396446228027, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.004, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.04701865836977959, | |
| "eval_runtime": 177.4067, | |
| "eval_samples_per_second": 6.516, | |
| "eval_steps_per_second": 6.516, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.17373262345790863, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0036, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.058048076927661896, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0017, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.11685014516115189, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.0042, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.41562795639038086, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0033, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.01105436123907566, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0035, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.047170307487249374, | |
| "eval_runtime": 177.3984, | |
| "eval_samples_per_second": 6.516, | |
| "eval_steps_per_second": 6.516, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.20917461812496185, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.0026, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.39377889037132263, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0029, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.2723056674003601, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0088, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.1515825390815735, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.0026, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.06972948461771011, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.0029, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.0472617968916893, | |
| "eval_runtime": 177.4908, | |
| "eval_samples_per_second": 6.513, | |
| "eval_steps_per_second": 6.513, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.07699483633041382, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.0075, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.26263517141342163, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.0026, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.006273223087191582, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0046, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.060451608151197433, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0015, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.07621429860591888, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0026, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.04714197665452957, | |
| "eval_runtime": 177.4777, | |
| "eval_samples_per_second": 6.513, | |
| "eval_steps_per_second": 6.513, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.03596797212958336, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0029, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.16178975999355316, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.003, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.07360837608575821, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.0021, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.019955629482865334, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0036, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.02730751968920231, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.0042, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.047300469130277634, | |
| "eval_runtime": 177.5512, | |
| "eval_samples_per_second": 6.511, | |
| "eval_steps_per_second": 6.511, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.028180785477161407, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0042, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.03916345536708832, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.0019, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.0514407716691494, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0057, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.06616293638944626, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0021, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 8.169046237996646e+17, | |
| "train_loss": 0.03518220033433478, | |
| "train_runtime": 39188.4528, | |
| "train_samples_per_second": 1.326, | |
| "train_steps_per_second": 0.083 | |
| } | |
| ], | |
| "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": 8.169046237996646e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |