| { | |
| "best_metric": 0.031774770468473434, | |
| "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold3/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": 3.6774675846099854, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.9984, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 4.073786735534668, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.9282, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.947826623916626, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.7182, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 3.3447530269622803, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.38, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 2.073380947113037, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8727, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.6953915357589722, | |
| "eval_runtime": 182.9068, | |
| "eval_samples_per_second": 6.32, | |
| "eval_steps_per_second": 6.32, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 0.8990628719329834, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.59, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.209133267402649, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.4083, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 0.6037677526473999, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.2135, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 0.9614138007164001, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.2176, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.675909698009491, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1562, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.10779810696840286, | |
| "eval_runtime": 182.8835, | |
| "eval_samples_per_second": 6.321, | |
| "eval_steps_per_second": 6.321, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8575292229652405, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.11, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.8497035503387451, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.109, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.2980672121047974, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0787, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.951867938041687, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0833, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.7117636799812317, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0876, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.07320743054151535, | |
| "eval_runtime": 183.0956, | |
| "eval_samples_per_second": 6.314, | |
| "eval_steps_per_second": 6.314, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.7358025908470154, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0824, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 1.396098256111145, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.075, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.7224653959274292, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0595, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.6202625632286072, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0522, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.581307053565979, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0737, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.05594949796795845, | |
| "eval_runtime": 183.0703, | |
| "eval_samples_per_second": 6.315, | |
| "eval_steps_per_second": 6.315, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.6913634538650513, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0643, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.7677487730979919, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0611, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3285107910633087, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0467, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.6444545388221741, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0577, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.48404598236083984, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0549, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.052511610090732574, | |
| "eval_runtime": 182.7374, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 6.326, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.3902552127838135, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0514, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.8752898573875427, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.0521, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.4953712821006775, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0576, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.37031790614128113, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0614, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.45014697313308716, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0519, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.04904678836464882, | |
| "eval_runtime": 182.3668, | |
| "eval_samples_per_second": 6.339, | |
| "eval_steps_per_second": 6.339, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.467668354511261, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.057, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 0.7374036908149719, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0606, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.3613051176071167, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.0567, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.4317915737628937, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0539, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.41593390703201294, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0448, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.046798210591077805, | |
| "eval_runtime": 182.5651, | |
| "eval_samples_per_second": 6.332, | |
| "eval_steps_per_second": 6.332, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.29354193806648254, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0583, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.4364352822303772, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0431, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.46638795733451843, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.0466, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.5713707208633423, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0415, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.23616552352905273, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0605, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.04365602135658264, | |
| "eval_runtime": 182.8728, | |
| "eval_samples_per_second": 6.321, | |
| "eval_steps_per_second": 6.321, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.16720706224441528, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0416, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.47889116406440735, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0529, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.5881320238113403, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.0484, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.3344607353210449, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0608, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.29933053255081177, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0356, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.04331237077713013, | |
| "eval_runtime": 183.1079, | |
| "eval_samples_per_second": 6.313, | |
| "eval_steps_per_second": 6.313, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.3562895953655243, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0368, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.31500017642974854, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.0298, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5254421830177307, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0352, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.24430716037750244, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.0402, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.24502789974212646, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0304, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.03936327248811722, | |
| "eval_runtime": 182.8827, | |
| "eval_samples_per_second": 6.321, | |
| "eval_steps_per_second": 6.321, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.4982253909111023, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0312, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.1713370382785797, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.0505, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.5369380712509155, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0627, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.36026087403297424, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0433, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.2174559235572815, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.0469, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.03778400272130966, | |
| "eval_runtime": 182.6098, | |
| "eval_samples_per_second": 6.33, | |
| "eval_steps_per_second": 6.33, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.1945793628692627, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.031, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.32261940836906433, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.0312, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.3344806432723999, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0378, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.29067111015319824, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0529, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.3147498071193695, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0368, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.03696472570300102, | |
| "eval_runtime": 182.0254, | |
| "eval_samples_per_second": 6.351, | |
| "eval_steps_per_second": 6.351, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.26502376794815063, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0357, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.18783558905124664, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.0397, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.3690519332885742, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.0443, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.7493480443954468, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0345, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.7668882608413696, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0333, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.036111775785684586, | |
| "eval_runtime": 182.3347, | |
| "eval_samples_per_second": 6.34, | |
| "eval_steps_per_second": 6.34, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.38832399249076843, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0374, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.43809109926223755, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0379, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.24600400030612946, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0325, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.47243162989616394, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.0504, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.24714922904968262, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0359, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.03797317296266556, | |
| "eval_runtime": 182.6494, | |
| "eval_samples_per_second": 6.329, | |
| "eval_steps_per_second": 6.329, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.28389057517051697, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.0321, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.1833420842885971, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0264, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.28029078245162964, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.0316, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.5390480160713196, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0374, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.2820804715156555, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.0296, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.0354355163872242, | |
| "eval_runtime": 182.8939, | |
| "eval_samples_per_second": 6.321, | |
| "eval_steps_per_second": 6.321, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.22800962626934052, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.0378, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.2685460150241852, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.0387, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.20110183954238892, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0334, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.2733299434185028, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0338, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.29376664757728577, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.0299, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.03445694223046303, | |
| "eval_runtime": 182.7984, | |
| "eval_samples_per_second": 6.324, | |
| "eval_steps_per_second": 6.324, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.16524243354797363, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.0359, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.22941583395004272, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0266, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.2582172453403473, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.0306, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.4427955746650696, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0285, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.6793798208236694, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.0333, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.03575538098812103, | |
| "eval_runtime": 182.6241, | |
| "eval_samples_per_second": 6.33, | |
| "eval_steps_per_second": 6.33, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.1999335139989853, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.0336, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.33107635378837585, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0295, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.1308780163526535, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0275, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.18942958116531372, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.0357, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.2868015468120575, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0202, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03390350192785263, | |
| "eval_runtime": 182.1932, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 6.345, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.22393953800201416, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0293, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.2709175944328308, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0394, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.2407248467206955, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.0267, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.37008678913116455, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0356, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.0995490550994873, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.0393, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.0345032662153244, | |
| "eval_runtime": 182.7802, | |
| "eval_samples_per_second": 6.325, | |
| "eval_steps_per_second": 6.325, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.3091396689414978, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.0301, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.07844063639640808, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0245, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.32270291447639465, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.0344, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.2041865885257721, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0336, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.2232264280319214, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0338, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.03348607197403908, | |
| "eval_runtime": 182.7226, | |
| "eval_samples_per_second": 6.327, | |
| "eval_steps_per_second": 6.327, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.20493818819522858, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.0361, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.24103085696697235, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0376, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.16034604609012604, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.0299, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.122935950756073, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.0333, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.24790926277637482, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0333, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.03358389437198639, | |
| "eval_runtime": 183.0943, | |
| "eval_samples_per_second": 6.314, | |
| "eval_steps_per_second": 6.314, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.20058514177799225, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.028, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.16348162293434143, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.0286, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 0.1269695907831192, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.0281, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.48118749260902405, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.03, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.21809779107570648, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.032, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.0352640300989151, | |
| "eval_runtime": 182.9402, | |
| "eval_samples_per_second": 6.319, | |
| "eval_steps_per_second": 6.319, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.45644205808639526, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0392, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.6009358167648315, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.0281, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.23524214327335358, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.0394, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.13619941473007202, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0488, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.2308131754398346, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0206, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.03377874568104744, | |
| "eval_runtime": 182.2847, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 6.342, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.16370758414268494, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.0302, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.15502361953258514, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.0242, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.16096149384975433, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0241, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.3754902184009552, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.0378, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.09612976759672165, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0206, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.03473026305437088, | |
| "eval_runtime": 182.3618, | |
| "eval_samples_per_second": 6.339, | |
| "eval_steps_per_second": 6.339, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.41802310943603516, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.0288, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.16000649333000183, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0267, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.10824336111545563, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0305, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.23428383469581604, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.035, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.3466808497905731, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.0285, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.034273166209459305, | |
| "eval_runtime": 182.4812, | |
| "eval_samples_per_second": 6.335, | |
| "eval_steps_per_second": 6.335, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.35819748044013977, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.0326, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.2065792828798294, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.0287, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.17215484380722046, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.0239, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.20868314802646637, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0321, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.13842304050922394, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.0217, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.03190917521715164, | |
| "eval_runtime": 182.8616, | |
| "eval_samples_per_second": 6.322, | |
| "eval_steps_per_second": 6.322, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.2511175870895386, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.0215, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.32441824674606323, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.0216, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.12878277897834778, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0219, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.15594542026519775, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.0209, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.25756609439849854, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0162, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.03346768394112587, | |
| "eval_runtime": 183.0599, | |
| "eval_samples_per_second": 6.315, | |
| "eval_steps_per_second": 6.315, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.1232733502984047, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0199, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.47351959347724915, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.0169, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.3003310561180115, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.0139, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.2135530561208725, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0173, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.6665906310081482, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0283, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.03566494956612587, | |
| "eval_runtime": 182.7325, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 6.326, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.09189430624246597, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0154, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.31826403737068176, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0234, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.367217093706131, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.015, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.19675622880458832, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0207, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.11378202587366104, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.0131, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.03628995269536972, | |
| "eval_runtime": 182.2877, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 6.342, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.19877927005290985, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.0201, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.16368044912815094, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0229, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.3174315094947815, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0245, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.15592573583126068, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.0238, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.2459307610988617, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0194, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.03373586758971214, | |
| "eval_runtime": 182.3201, | |
| "eval_samples_per_second": 6.34, | |
| "eval_steps_per_second": 6.34, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.21740512549877167, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.0221, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.12407861649990082, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0228, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.2391296774148941, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.0181, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.3536985218524933, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0176, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.23116756975650787, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.0155, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.0360429510474205, | |
| "eval_runtime": 182.6406, | |
| "eval_samples_per_second": 6.329, | |
| "eval_steps_per_second": 6.329, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.22570303082466125, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.0122, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.39931413531303406, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0239, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.13724231719970703, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0204, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.18864579498767853, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.0194, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.6650933027267456, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.017, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.03401138260960579, | |
| "eval_runtime": 182.8067, | |
| "eval_samples_per_second": 6.324, | |
| "eval_steps_per_second": 6.324, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.2827061116695404, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.0168, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.14448344707489014, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.0204, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.20898085832595825, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0179, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.356749027967453, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.0192, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.18224909901618958, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.0215, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.03248431161046028, | |
| "eval_runtime": 182.9367, | |
| "eval_samples_per_second": 6.319, | |
| "eval_steps_per_second": 6.319, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.38756829500198364, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.0172, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.3050745129585266, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0187, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.20349016785621643, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0228, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.23515108227729797, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0232, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.27538764476776123, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0185, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.03288661688566208, | |
| "eval_runtime": 182.5686, | |
| "eval_samples_per_second": 6.332, | |
| "eval_steps_per_second": 6.332, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.2373654991388321, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.0183, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.2687448263168335, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.031, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.38675692677497864, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.0234, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.11699690669775009, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.016, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.18256603181362152, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0275, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.031774770468473434, | |
| "eval_runtime": 182.1987, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 6.345, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.31536683440208435, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0159, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.32807672023773193, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0193, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.35243338346481323, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0235, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.13235580921173096, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.019, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.3544095456600189, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.0243, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.03448236361145973, | |
| "eval_runtime": 182.0949, | |
| "eval_samples_per_second": 6.348, | |
| "eval_steps_per_second": 6.348, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.26006463170051575, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.0203, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.32358068227767944, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0244, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.5204956531524658, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.0232, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.2777179479598999, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.0262, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.19366903603076935, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.0231, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.031777121126651764, | |
| "eval_runtime": 182.545, | |
| "eval_samples_per_second": 6.333, | |
| "eval_steps_per_second": 6.333, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.24932289123535156, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0138, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.3153086304664612, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0267, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.48333457112312317, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0123, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.2106030434370041, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.0186, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.2382618486881256, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0157, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.032552145421504974, | |
| "eval_runtime": 182.9652, | |
| "eval_samples_per_second": 6.318, | |
| "eval_steps_per_second": 6.318, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 0.25687718391418457, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.0217, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.3501730263233185, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.0183, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.15959833562374115, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0161, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.1952725201845169, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0125, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.27031782269477844, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.0287, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.03239323943853378, | |
| "eval_runtime": 181.5029, | |
| "eval_samples_per_second": 6.369, | |
| "eval_steps_per_second": 6.369, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.1635352373123169, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.007, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.20652925968170166, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.011, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.2952287197113037, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.0111, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.32101571559906006, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.0084, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.17091065645217896, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.0123, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.03469814360141754, | |
| "eval_runtime": 182.4502, | |
| "eval_samples_per_second": 6.336, | |
| "eval_steps_per_second": 6.336, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 0.6616946458816528, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.0068, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.04183431714773178, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0107, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.10999016463756561, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.0118, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.07972876727581024, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.0075, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.23311194777488708, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0079, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.037809696048498154, | |
| "eval_runtime": 182.1826, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 6.345, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.10034877806901932, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.007, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 1.8158587217330933, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.0085, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.650327742099762, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0106, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.29771697521209717, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.0082, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.10653995722532272, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.008, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.039970699697732925, | |
| "eval_runtime": 182.0809, | |
| "eval_samples_per_second": 6.349, | |
| "eval_steps_per_second": 6.349, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.11117534339427948, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.0112, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.33262163400650024, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0113, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.4333590269088745, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0083, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.26029932498931885, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0062, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.31432023644447327, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.011, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.039253540337085724, | |
| "eval_runtime": 182.3027, | |
| "eval_samples_per_second": 6.341, | |
| "eval_steps_per_second": 6.341, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.1568172574043274, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.0094, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.3937229812145233, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.015, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.4656408429145813, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.0093, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.4326402246952057, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.0079, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.11908203363418579, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0141, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.038429729640483856, | |
| "eval_runtime": 182.4763, | |
| "eval_samples_per_second": 6.335, | |
| "eval_steps_per_second": 6.335, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.2271566540002823, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.0088, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.32780489325523376, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0073, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.16878120601177216, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.012, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.36453279852867126, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0127, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.2168504148721695, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0088, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.03840179741382599, | |
| "eval_runtime": 182.6167, | |
| "eval_samples_per_second": 6.33, | |
| "eval_steps_per_second": 6.33, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.12594561278820038, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0083, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.2667974531650543, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.0058, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.280409038066864, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0127, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.4261855483055115, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0066, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.07291776686906815, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0113, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.037894509732723236, | |
| "eval_runtime": 182.7863, | |
| "eval_samples_per_second": 6.324, | |
| "eval_steps_per_second": 6.324, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.1499006748199463, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0111, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.5744539499282837, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.0123, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.17394568026065826, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0087, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.165720134973526, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0099, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.07900846749544144, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0048, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.03884553909301758, | |
| "eval_runtime": 182.0791, | |
| "eval_samples_per_second": 6.349, | |
| "eval_steps_per_second": 6.349, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.3270080089569092, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0125, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.325246661901474, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.0084, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.24096044898033142, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.0098, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.35391300916671753, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.0118, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.09201920032501221, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.0136, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.037477023899555206, | |
| "eval_runtime": 182.3323, | |
| "eval_samples_per_second": 6.34, | |
| "eval_steps_per_second": 6.34, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.39145535230636597, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.0116, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.2135828584432602, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0068, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.12653900682926178, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.0083, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.48592713475227356, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0088, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.17783097922801971, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0069, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.03769952431321144, | |
| "eval_runtime": 182.2699, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 6.342, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.08266368508338928, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0077, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.26081985235214233, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0051, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.1756199300289154, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.011, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.32230377197265625, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0096, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.2033834159374237, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.0113, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.0389435850083828, | |
| "eval_runtime": 182.6925, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 6.328, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.030453253537416458, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0086, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.2909805178642273, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0037, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.07845457643270493, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.006, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.3073825240135193, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.0057, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.5263724327087402, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.0117, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.039375029504299164, | |
| "eval_runtime": 182.6693, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 6.328, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.3800037205219269, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0085, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.3800065517425537, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0065, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 0.061865080147981644, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0145, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.44012024998664856, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.0081, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.20893503725528717, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.0086, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.039268363267183304, | |
| "eval_runtime": 182.4686, | |
| "eval_samples_per_second": 6.335, | |
| "eval_steps_per_second": 6.335, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.07197685539722443, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0053, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.16688397526741028, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0017, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.09238763898611069, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.002, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.04073997586965561, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0056, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.034568727016448975, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.0016, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.040927786380052567, | |
| "eval_runtime": 182.3483, | |
| "eval_samples_per_second": 6.34, | |
| "eval_steps_per_second": 6.34, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.18634392321109772, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0077, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.3908490538597107, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.0023, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.0462653823196888, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0021, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.05564751476049423, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0044, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.08183032274246216, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0049, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.04324441775679588, | |
| "eval_runtime": 181.9827, | |
| "eval_samples_per_second": 6.352, | |
| "eval_steps_per_second": 6.352, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.10928888618946075, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.0037, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.1508701741695404, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0042, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.2720945477485657, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0024, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 0.4039311707019806, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.0051, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.09777352958917618, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0027, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.04485039412975311, | |
| "eval_runtime": 182.3893, | |
| "eval_samples_per_second": 6.338, | |
| "eval_steps_per_second": 6.338, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.1132354810833931, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0021, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.022593403235077858, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.0019, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.3615880310535431, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.003, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.09757770597934723, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0037, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.023183895274996758, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.0025, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.0455702468752861, | |
| "eval_runtime": 182.6081, | |
| "eval_samples_per_second": 6.33, | |
| "eval_steps_per_second": 6.33, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.05011435225605965, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0017, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.042359739542007446, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0032, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.11338278651237488, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.0025, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.44294944405555725, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0034, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.055634595453739166, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0022, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.04680494964122772, | |
| "eval_runtime": 182.6686, | |
| "eval_samples_per_second": 6.328, | |
| "eval_steps_per_second": 6.328, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.23079335689544678, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.002, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.033984847366809845, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.0016, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.06768519431352615, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0089, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.04040565341711044, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0018, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.25804442167282104, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0032, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.04704459384083748, | |
| "eval_runtime": 182.4137, | |
| "eval_samples_per_second": 6.337, | |
| "eval_steps_per_second": 6.337, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.25351253151893616, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.0068, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.06695833802223206, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0045, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.02735094167292118, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.0024, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.5262628793716431, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0037, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.018443401902914047, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0021, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.0470660924911499, | |
| "eval_runtime": 182.0841, | |
| "eval_samples_per_second": 6.349, | |
| "eval_steps_per_second": 6.349, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.05941687524318695, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0006, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 0.12561437487602234, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.0022, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.03257729485630989, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.0024, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.354690819978714, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.0044, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 0.18320393562316895, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.002, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.04752160608768463, | |
| "eval_runtime": 182.2015, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 6.345, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.08117330819368362, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0014, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.0543917678296566, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0023, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.36946216225624084, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.0051, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.44694221019744873, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0028, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.012448400259017944, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0017, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.04781460389494896, | |
| "eval_runtime": 182.5209, | |
| "eval_samples_per_second": 6.334, | |
| "eval_steps_per_second": 6.334, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.08003149181604385, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.0014, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.42508721351623535, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0024, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.3715968728065491, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0047, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.04861222580075264, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.0034, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.06290697306394577, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.0017, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.04796063154935837, | |
| "eval_runtime": 182.7703, | |
| "eval_samples_per_second": 6.325, | |
| "eval_steps_per_second": 6.325, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.04672136530280113, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.0052, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.16760879755020142, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.0014, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.008177582174539566, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0034, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.04469135031104088, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0018, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.27092278003692627, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0066, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.04783838242292404, | |
| "eval_runtime": 182.74, | |
| "eval_samples_per_second": 6.326, | |
| "eval_steps_per_second": 6.326, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.03831211477518082, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0045, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.3491027355194092, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.0024, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.07816429436206818, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.0023, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.07785133272409439, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0023, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.05292694643139839, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.0017, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.04795035347342491, | |
| "eval_runtime": 182.5596, | |
| "eval_samples_per_second": 6.332, | |
| "eval_steps_per_second": 6.332, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.03667265176773071, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0042, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.043597154319286346, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.0081, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.4396083354949951, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0045, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.09275593608617783, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0024, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 7.696760527880847e+17, | |
| "train_loss": 0.05112033618057293, | |
| "train_runtime": 42506.2119, | |
| "train_samples_per_second": 1.223, | |
| "train_steps_per_second": 0.076 | |
| } | |
| ], | |
| "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 | |
| } | |