| { | |
| "best_metric": 0.03135393559932709, | |
| "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold8/checkpoint-1750", | |
| "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": 3.667463541030884, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.9998, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 4.112172603607178, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.9259, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.9791252613067627, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.7155, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 3.399125337600708, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.3786, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 2.067168712615967, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8741, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.694180428981781, | |
| "eval_runtime": 153.441, | |
| "eval_samples_per_second": 7.534, | |
| "eval_steps_per_second": 7.534, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 0.8935492634773254, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5903, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.2200783491134644, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.4084, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 0.6045705676078796, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.2131, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 0.9763121008872986, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.2182, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.6636149287223816, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1568, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.10798542201519012, | |
| "eval_runtime": 153.6032, | |
| "eval_samples_per_second": 7.526, | |
| "eval_steps_per_second": 7.526, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8777512907981873, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.11, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.7971823811531067, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.1089, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.2814828157424927, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0781, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.9530104994773865, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0826, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.687839150428772, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0871, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.07265099138021469, | |
| "eval_runtime": 153.7109, | |
| "eval_samples_per_second": 7.521, | |
| "eval_steps_per_second": 7.521, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.7616832256317139, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0819, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 1.4258986711502075, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.0753, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.7855386137962341, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0604, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.6559249758720398, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0521, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.5517032146453857, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0731, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.05592866241931915, | |
| "eval_runtime": 153.7545, | |
| "eval_samples_per_second": 7.518, | |
| "eval_steps_per_second": 7.518, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.6920191645622253, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0645, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.8007709383964539, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0607, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3330081105232239, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0463, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.6785910725593567, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0589, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.4944949746131897, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0543, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.05296166613698006, | |
| "eval_runtime": 153.7699, | |
| "eval_samples_per_second": 7.518, | |
| "eval_steps_per_second": 7.518, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.3755505084991455, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0528, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.7801454663276672, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.0523, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.5076040029525757, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0574, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.330967515707016, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0615, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.4423432946205139, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0527, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.05020619556307793, | |
| "eval_runtime": 153.6882, | |
| "eval_samples_per_second": 7.522, | |
| "eval_steps_per_second": 7.522, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.47362765669822693, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.0578, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 1.2070564031600952, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0633, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.33177974820137024, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.0576, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.4434049725532532, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0537, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.3827102482318878, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0465, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.047334399074316025, | |
| "eval_runtime": 153.7917, | |
| "eval_samples_per_second": 7.517, | |
| "eval_steps_per_second": 7.517, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.26626113057136536, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0592, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.459421843290329, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0436, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.4921688735485077, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.046, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.5686693787574768, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0426, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.22994281351566315, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0624, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.04411206394433975, | |
| "eval_runtime": 153.7262, | |
| "eval_samples_per_second": 7.52, | |
| "eval_steps_per_second": 7.52, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.18209218978881836, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0419, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.4690866470336914, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0539, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.569894552230835, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.048, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.34885498881340027, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0596, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.28586235642433167, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0356, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.04399007931351662, | |
| "eval_runtime": 153.7689, | |
| "eval_samples_per_second": 7.518, | |
| "eval_steps_per_second": 7.518, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.34978532791137695, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0372, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.26985228061676025, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.0297, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5369043946266174, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0354, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.2718733847141266, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.0385, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.2562827169895172, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0298, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.04100799188017845, | |
| "eval_runtime": 153.8218, | |
| "eval_samples_per_second": 7.515, | |
| "eval_steps_per_second": 7.515, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.5368469953536987, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0327, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.20041915774345398, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.049, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.4386811852455139, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0627, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.3586682975292206, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0425, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.3473396599292755, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.0477, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.038224488496780396, | |
| "eval_runtime": 153.8909, | |
| "eval_samples_per_second": 7.512, | |
| "eval_steps_per_second": 7.512, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.22000104188919067, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.0318, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.34618744254112244, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.0308, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.3343406021595001, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0374, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.3576776683330536, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0525, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.3294927477836609, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0369, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.03695068880915642, | |
| "eval_runtime": 153.7653, | |
| "eval_samples_per_second": 7.518, | |
| "eval_steps_per_second": 7.518, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.25722983479499817, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0353, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.1812303215265274, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.0404, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.36584559082984924, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.0439, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.9148450493812561, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0367, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.6868040561676025, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0295, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.03621339797973633, | |
| "eval_runtime": 153.7927, | |
| "eval_samples_per_second": 7.517, | |
| "eval_steps_per_second": 7.517, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.42277103662490845, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0371, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.6032630801200867, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0367, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.27518221735954285, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0358, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.4526023864746094, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.0472, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.203059583902359, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0366, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.036694660782814026, | |
| "eval_runtime": 153.9032, | |
| "eval_samples_per_second": 7.511, | |
| "eval_steps_per_second": 7.511, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.31703072786331177, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.031, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.1820017397403717, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0267, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.28401657938957214, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.0322, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.49696090817451477, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0385, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.22875753045082092, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.0303, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.03541158512234688, | |
| "eval_runtime": 153.9637, | |
| "eval_samples_per_second": 7.508, | |
| "eval_steps_per_second": 7.508, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.23917922377586365, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.038, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.2626829743385315, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.0399, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.21504108607769012, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0329, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.28656354546546936, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0338, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.27761170268058777, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.031, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.03466470539569855, | |
| "eval_runtime": 153.9328, | |
| "eval_samples_per_second": 7.51, | |
| "eval_steps_per_second": 7.51, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.16848354041576385, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.0347, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.2347051501274109, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0264, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.2753904163837433, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.0309, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.4441526234149933, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0288, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.6059399247169495, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.0345, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.03497758135199547, | |
| "eval_runtime": 153.8687, | |
| "eval_samples_per_second": 7.513, | |
| "eval_steps_per_second": 7.513, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.20635059475898743, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.0327, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.3250363767147064, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0288, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.14383748173713684, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0275, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.16825367510318756, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.0367, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.23061300814151764, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0203, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03370543196797371, | |
| "eval_runtime": 153.8973, | |
| "eval_samples_per_second": 7.512, | |
| "eval_steps_per_second": 7.512, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.2180616408586502, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0282, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.2725543975830078, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0349, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.2383366823196411, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.0272, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.4063188135623932, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0345, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.08444732427597046, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.0406, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.03430265933275223, | |
| "eval_runtime": 153.9727, | |
| "eval_samples_per_second": 7.508, | |
| "eval_steps_per_second": 7.508, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.32592183351516724, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.029, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.072084441781044, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0233, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.33728575706481934, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.036, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.193743035197258, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0347, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.25290656089782715, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0326, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.03308522328734398, | |
| "eval_runtime": 153.9012, | |
| "eval_samples_per_second": 7.511, | |
| "eval_steps_per_second": 7.511, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.1948683261871338, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.035, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.2447928935289383, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0404, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.1428983509540558, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.0322, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.12207616120576859, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.032, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.26099836826324463, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0325, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.0332084558904171, | |
| "eval_runtime": 153.956, | |
| "eval_samples_per_second": 7.509, | |
| "eval_steps_per_second": 7.509, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.21652770042419434, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.0269, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.16382655501365662, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.028, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 0.17023272812366486, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.0252, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.4871385097503662, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.0294, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.21047188341617584, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.0328, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.03518792241811752, | |
| "eval_runtime": 153.9015, | |
| "eval_samples_per_second": 7.511, | |
| "eval_steps_per_second": 7.511, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.4814411699771881, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0382, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.19904109835624695, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.0261, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.25041642785072327, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.0415, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.1513228565454483, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0456, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.19501468539237976, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0198, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.032900188118219376, | |
| "eval_runtime": 154.0261, | |
| "eval_samples_per_second": 7.505, | |
| "eval_steps_per_second": 7.505, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.18908829987049103, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.0307, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.14665935933589935, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.0228, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.16647779941558838, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0237, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.35168758034706116, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.0369, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.09242203086614609, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0202, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.034483663737773895, | |
| "eval_runtime": 153.9818, | |
| "eval_samples_per_second": 7.507, | |
| "eval_steps_per_second": 7.507, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.3413452208042145, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.0299, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.14245063066482544, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0266, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.1168195828795433, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0312, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.21591340005397797, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.0354, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.3393869698047638, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.0301, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.03373435139656067, | |
| "eval_runtime": 154.0523, | |
| "eval_samples_per_second": 7.504, | |
| "eval_steps_per_second": 7.504, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.3212391436100006, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.032, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.19600151479244232, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.0289, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.1494583934545517, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.0235, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.2839687764644623, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0325, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.19894124567508698, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.022, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.03152144327759743, | |
| "eval_runtime": 153.8921, | |
| "eval_samples_per_second": 7.512, | |
| "eval_steps_per_second": 7.512, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.2692122161388397, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.021, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.2785106301307678, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.0225, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.12847158312797546, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0227, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.1553700566291809, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.0208, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.2573378384113312, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0159, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.032453399151563644, | |
| "eval_runtime": 153.8675, | |
| "eval_samples_per_second": 7.513, | |
| "eval_steps_per_second": 7.513, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.12254810333251953, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0203, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.48425406217575073, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.0182, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.25750547647476196, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.0136, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.24635617434978485, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0173, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.6029477715492249, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0304, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.03489888086915016, | |
| "eval_runtime": 154.0117, | |
| "eval_samples_per_second": 7.506, | |
| "eval_steps_per_second": 7.506, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.0881878212094307, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0155, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.35236889123916626, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0252, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.3474541902542114, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.0155, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.17613942921161652, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0207, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.10841204226016998, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.0145, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.03541386127471924, | |
| "eval_runtime": 153.9757, | |
| "eval_samples_per_second": 7.508, | |
| "eval_steps_per_second": 7.508, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.28811851143836975, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.0194, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.19602102041244507, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0212, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.2925778329372406, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0245, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.11785721778869629, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.025, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.19476380944252014, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0191, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.03342807665467262, | |
| "eval_runtime": 154.0542, | |
| "eval_samples_per_second": 7.504, | |
| "eval_steps_per_second": 7.504, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.20125311613082886, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.025, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.18337379395961761, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0225, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.2674618065357208, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.0199, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.31585049629211426, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0181, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.20296861231327057, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.015, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.03503217175602913, | |
| "eval_runtime": 154.0143, | |
| "eval_samples_per_second": 7.506, | |
| "eval_steps_per_second": 7.506, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.261443167924881, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.0132, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.3993077874183655, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0248, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.12978719174861908, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0213, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.180982768535614, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.0187, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.2891698479652405, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.0151, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.03429364413022995, | |
| "eval_runtime": 154.0684, | |
| "eval_samples_per_second": 7.503, | |
| "eval_steps_per_second": 7.503, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.2865266501903534, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.016, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.20926989614963531, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.0195, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.124493807554245, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0202, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.3267703652381897, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.0196, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.24480195343494415, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.022, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.03300287202000618, | |
| "eval_runtime": 154.0869, | |
| "eval_samples_per_second": 7.502, | |
| "eval_steps_per_second": 7.502, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.5107797384262085, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.0182, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.3029766082763672, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0186, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.28008517622947693, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0239, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.24223163723945618, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0227, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.3835962414741516, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0192, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.03280750289559364, | |
| "eval_runtime": 154.0503, | |
| "eval_samples_per_second": 7.504, | |
| "eval_steps_per_second": 7.504, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.2985157072544098, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.0204, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.18213863670825958, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.0294, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.38688215613365173, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.0222, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.13498535752296448, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.0178, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.2052980363368988, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0266, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.03135393559932709, | |
| "eval_runtime": 154.0345, | |
| "eval_samples_per_second": 7.505, | |
| "eval_steps_per_second": 7.505, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.39836385846138, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0145, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.7306950092315674, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0224, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.35386157035827637, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0247, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.0966259315609932, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.0192, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.3098321855068207, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.0246, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.0328693725168705, | |
| "eval_runtime": 154.0652, | |
| "eval_samples_per_second": 7.503, | |
| "eval_steps_per_second": 7.503, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.2583811283111572, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.0189, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.3612092435359955, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0245, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.6103046536445618, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.0234, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.24834753572940826, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.0269, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.24632981419563293, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.022, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.03177366778254509, | |
| "eval_runtime": 154.0675, | |
| "eval_samples_per_second": 7.503, | |
| "eval_steps_per_second": 7.503, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.26031494140625, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0141, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.3189842402935028, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0266, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.5599565505981445, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0125, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.2120751440525055, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.0184, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.26458740234375, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0172, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.03305280581116676, | |
| "eval_runtime": 154.111, | |
| "eval_samples_per_second": 7.501, | |
| "eval_steps_per_second": 7.501, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 0.3012411296367645, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.023, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.42970404028892517, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.019, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.12723307311534882, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0169, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.19563929736614227, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0148, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.2497878521680832, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.0294, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.03314288705587387, | |
| "eval_runtime": 154.1313, | |
| "eval_samples_per_second": 7.5, | |
| "eval_steps_per_second": 7.5, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.10688884556293488, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.007, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.19804410636425018, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.0106, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.24048706889152527, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.0099, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.30408522486686707, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.0088, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.16502350568771362, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.0117, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.035659901797771454, | |
| "eval_runtime": 154.1747, | |
| "eval_samples_per_second": 7.498, | |
| "eval_steps_per_second": 7.498, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 1.2794837951660156, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.0127, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.06127078831195831, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0103, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.16809481382369995, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.0095, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.08473172783851624, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.0075, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.3410157859325409, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0092, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.03898155316710472, | |
| "eval_runtime": 154.1765, | |
| "eval_samples_per_second": 7.498, | |
| "eval_steps_per_second": 7.498, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.1153949573636055, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.0094, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 0.6003677845001221, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.009, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.30562642216682434, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0098, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.3595854938030243, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.01, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.17397575080394745, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.0077, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.04023173823952675, | |
| "eval_runtime": 154.148, | |
| "eval_samples_per_second": 7.499, | |
| "eval_steps_per_second": 7.499, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.05835992470383644, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.0093, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.39425238966941833, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0132, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.33873605728149414, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0083, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.1726522296667099, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0058, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.29870375990867615, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.012, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.03809724748134613, | |
| "eval_runtime": 154.2438, | |
| "eval_samples_per_second": 7.495, | |
| "eval_steps_per_second": 7.495, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.08864596486091614, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.01, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.3720962703227997, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.0147, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.16348831355571747, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.009, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.22117219865322113, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.007, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.0957334116101265, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0122, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.038925137370824814, | |
| "eval_runtime": 154.1786, | |
| "eval_samples_per_second": 7.498, | |
| "eval_steps_per_second": 7.498, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.13717682659626007, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.0082, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.3247677981853485, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0068, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.07803032547235489, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.0116, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.2899368405342102, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0145, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.2591307759284973, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0099, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.03895587474107742, | |
| "eval_runtime": 154.1073, | |
| "eval_samples_per_second": 7.501, | |
| "eval_steps_per_second": 7.501, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.1508382260799408, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0086, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.5648091435432434, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.0049, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.3072786033153534, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0127, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.35864803194999695, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0073, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.11584164202213287, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0105, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.03940539062023163, | |
| "eval_runtime": 154.1567, | |
| "eval_samples_per_second": 7.499, | |
| "eval_steps_per_second": 7.499, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.08626662194728851, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0128, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.5059362053871155, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.0127, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.16148799657821655, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0096, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.23104353249073029, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0141, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.1036362498998642, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0063, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.04014433175325394, | |
| "eval_runtime": 154.1855, | |
| "eval_samples_per_second": 7.497, | |
| "eval_steps_per_second": 7.497, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.3482014238834381, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0113, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.16223925352096558, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.008, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.197907492518425, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.01, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.29179486632347107, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.0111, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.1589972823858261, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.0125, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.03913673758506775, | |
| "eval_runtime": 154.0952, | |
| "eval_samples_per_second": 7.502, | |
| "eval_steps_per_second": 7.502, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.437531977891922, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.0097, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.3047080338001251, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0066, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.19550862908363342, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.0088, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.6619084477424622, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0069, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.21010679006576538, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0072, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.04005194082856178, | |
| "eval_runtime": 154.215, | |
| "eval_samples_per_second": 7.496, | |
| "eval_steps_per_second": 7.496, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.2475212961435318, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0111, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.11438047885894775, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0041, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.1081487312912941, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.0112, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.3907530903816223, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0118, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.15736280381679535, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.008, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.040464118123054504, | |
| "eval_runtime": 154.2658, | |
| "eval_samples_per_second": 7.494, | |
| "eval_steps_per_second": 7.494, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.05687052756547928, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0087, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.44557449221611023, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0044, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.04779889062047005, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.0071, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.7044463157653809, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.007, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.6442089080810547, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.0134, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.04132843017578125, | |
| "eval_runtime": 154.2197, | |
| "eval_samples_per_second": 7.496, | |
| "eval_steps_per_second": 7.496, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.4363694190979004, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0062, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.5929672122001648, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0067, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 0.09736856818199158, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0133, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.4394150674343109, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.0071, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.096865713596344, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.0106, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.04112254083156586, | |
| "eval_runtime": 154.1912, | |
| "eval_samples_per_second": 7.497, | |
| "eval_steps_per_second": 7.497, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.1255452036857605, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0045, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.2016223669052124, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0017, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.18958726525306702, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.0021, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.035594213753938675, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0062, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.059392865747213364, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.0019, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.042438898235559464, | |
| "eval_runtime": 154.2088, | |
| "eval_samples_per_second": 7.496, | |
| "eval_steps_per_second": 7.496, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.30811774730682373, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0042, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.44721946120262146, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.0052, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.1397596150636673, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0022, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.04802631959319115, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0045, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.07587681710720062, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0046, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.04538372904062271, | |
| "eval_runtime": 154.2989, | |
| "eval_samples_per_second": 7.492, | |
| "eval_steps_per_second": 7.492, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.10571572929620743, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.0041, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.08995271474123001, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0036, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.04601708799600601, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0029, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 0.546953558921814, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.0044, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.11951048672199249, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0024, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.047019705176353455, | |
| "eval_runtime": 154.3277, | |
| "eval_samples_per_second": 7.491, | |
| "eval_steps_per_second": 7.491, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.19217205047607422, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0019, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.06735862791538239, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.002, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.3350626826286316, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.0032, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.07147694379091263, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0034, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.03639521077275276, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.0043, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.04782328009605408, | |
| "eval_runtime": 154.2087, | |
| "eval_samples_per_second": 7.496, | |
| "eval_steps_per_second": 7.496, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.02708691731095314, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0014, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.026878582313656807, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0027, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.1307045817375183, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.0033, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.440371036529541, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0036, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.028540313243865967, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0035, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.04876387491822243, | |
| "eval_runtime": 154.3007, | |
| "eval_samples_per_second": 7.492, | |
| "eval_steps_per_second": 7.492, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.07301197946071625, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.0016, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.13184145092964172, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.002, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.0641188994050026, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0092, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.022258473560214043, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0017, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.29959726333618164, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0043, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.048822417855262756, | |
| "eval_runtime": 154.2575, | |
| "eval_samples_per_second": 7.494, | |
| "eval_steps_per_second": 7.494, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.05938433110713959, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.007, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.030639637261629105, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0045, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.039124682545661926, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.003, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.3977506756782532, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0038, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.015186689794063568, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0019, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.048708293586969376, | |
| "eval_runtime": 154.2938, | |
| "eval_samples_per_second": 7.492, | |
| "eval_steps_per_second": 7.492, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.09761206805706024, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0011, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 0.06154545769095421, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.0037, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.02214459702372551, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.0021, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.34301137924194336, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.0045, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 0.27755483984947205, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.0026, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.04899110272526741, | |
| "eval_runtime": 154.2672, | |
| "eval_samples_per_second": 7.493, | |
| "eval_steps_per_second": 7.493, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.08789480477571487, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0017, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.08831636607646942, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0028, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.4458440840244293, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.0036, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.39524224400520325, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0032, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.009003978222608566, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0043, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.04908358305692673, | |
| "eval_runtime": 154.231, | |
| "eval_samples_per_second": 7.495, | |
| "eval_steps_per_second": 7.495, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.040726516395807266, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.0036, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.5703622102737427, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0039, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.6556193232536316, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0046, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.042362816631793976, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.003, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.055376190692186356, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.0028, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.04941676929593086, | |
| "eval_runtime": 154.2562, | |
| "eval_samples_per_second": 7.494, | |
| "eval_steps_per_second": 7.494, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.07842595130205154, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.0065, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.23734621703624725, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.0017, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.007092535961419344, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0036, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.012472891248762608, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0018, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.28663739562034607, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0056, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.049282677471637726, | |
| "eval_runtime": 154.2309, | |
| "eval_samples_per_second": 7.495, | |
| "eval_steps_per_second": 7.495, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.10973035544157028, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0037, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.5079404711723328, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.004, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.0561533123254776, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.003, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.05922123044729233, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0023, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.05117160826921463, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.002, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.049201030284166336, | |
| "eval_runtime": 154.3325, | |
| "eval_samples_per_second": 7.49, | |
| "eval_steps_per_second": 7.49, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.2753202021121979, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0042, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.044614214450120926, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.0029, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.3854585886001587, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0048, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.14771610498428345, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0033, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 7.696760527880847e+17, | |
| "train_loss": 0.051249439420411005, | |
| "train_runtime": 36405.1679, | |
| "train_samples_per_second": 1.428, | |
| "train_steps_per_second": 0.089 | |
| } | |
| ], | |
| "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 | |
| } | |