| { | |
| "best_metric": 0.031263191252946854, | |
| "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold1/checkpoint-1850", | |
| "epoch": 4.994228549442093, | |
| "eval_steps": 50, | |
| "global_step": 3245, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015390534821085032, | |
| "grad_norm": 3.7110273838043213, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 2.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 4.1000189781188965, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.9308, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.957352876663208, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.7153, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 3.3803999423980713, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 1.3784, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 2.1044628620147705, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.8721, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.6930733919143677, | |
| "eval_runtime": 161.9707, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 0.8994517922401428, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5895, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.2064399719238281, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.4074, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 0.6040802597999573, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.2138, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 0.9704834222793579, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.2181, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.6877935528755188, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1563, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.10821972787380219, | |
| "eval_runtime": 161.8566, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8795369267463684, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.1099, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.8044566512107849, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.1087, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.2978191375732422, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0778, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.9771407246589661, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0835, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.7040006518363953, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0879, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.07276571542024612, | |
| "eval_runtime": 161.8499, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.761782705783844, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0827, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 1.3636269569396973, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.0748, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.7119629979133606, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0593, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.6472892165184021, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0522, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.5895541906356812, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0734, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.05593929439783096, | |
| "eval_runtime": 161.9367, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.7122710347175598, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0646, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.7554808259010315, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0611, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3233736455440521, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0462, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.6468804478645325, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0588, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.5243495106697083, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0548, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.05322062969207764, | |
| "eval_runtime": 161.8323, | |
| "eval_samples_per_second": 7.143, | |
| "eval_steps_per_second": 7.143, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.3603438436985016, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0524, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.7787327170372009, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.052, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.4925541281700134, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0584, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.3450389802455902, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0617, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.43563273549079895, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0519, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.049991585314273834, | |
| "eval_runtime": 161.7621, | |
| "eval_samples_per_second": 7.146, | |
| "eval_steps_per_second": 7.146, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.46105748414993286, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.0575, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 0.7581703066825867, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0621, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.3288971781730652, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.0563, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.4313550591468811, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0521, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.40282249450683594, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0463, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.0476096011698246, | |
| "eval_runtime": 161.8196, | |
| "eval_samples_per_second": 7.144, | |
| "eval_steps_per_second": 7.144, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.2729192078113556, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0585, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.4730323851108551, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0449, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.4265844225883484, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.0452, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.5771760940551758, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0429, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.22883038222789764, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0622, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.0440874807536602, | |
| "eval_runtime": 161.8862, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.18309859931468964, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0421, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.4488441050052643, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0543, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.5602842569351196, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.0477, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.3388902544975281, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0614, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.2830776572227478, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0345, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.04345938190817833, | |
| "eval_runtime": 161.823, | |
| "eval_samples_per_second": 7.144, | |
| "eval_steps_per_second": 7.144, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.32943403720855713, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0356, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.4647481143474579, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.0297, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5281929969787598, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0356, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.25071898102760315, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.041, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.25152504444122314, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0317, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.04137541353702545, | |
| "eval_runtime": 161.8013, | |
| "eval_samples_per_second": 7.145, | |
| "eval_steps_per_second": 7.145, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.4894426167011261, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0336, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.15945211052894592, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.0495, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.4114932119846344, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0648, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.37862494587898254, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0418, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.26095160841941833, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.048, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.03805786743760109, | |
| "eval_runtime": 161.9493, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.2188524603843689, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.0322, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.3318057358264923, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.0322, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.33760371804237366, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0391, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.3217584490776062, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0549, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.3070793151855469, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0373, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.037635400891304016, | |
| "eval_runtime": 161.8338, | |
| "eval_samples_per_second": 7.143, | |
| "eval_steps_per_second": 7.143, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.2580590546131134, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0368, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.1826627105474472, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.0393, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.4323957860469818, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.0442, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.878084659576416, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0367, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.7204973101615906, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0312, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.036002594977617264, | |
| "eval_runtime": 161.8847, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.48999980092048645, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0382, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.4334917962551117, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0369, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.2723782956600189, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0325, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.4379688501358032, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.0474, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.23990271985530853, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0362, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.03780118748545647, | |
| "eval_runtime": 161.9299, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.23954886198043823, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.0312, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.18448102474212646, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0267, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.2539280652999878, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.0321, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.5217154622077942, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0376, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.2805447280406952, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.03, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.03487715125083923, | |
| "eval_runtime": 161.9273, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.2448904663324356, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.0377, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.28172314167022705, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.0407, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.2009604424238205, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0325, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.2888275682926178, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0342, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.28310900926589966, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.03, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.034933045506477356, | |
| "eval_runtime": 161.8833, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.1619165986776352, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.0346, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.2185034453868866, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0268, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.30337387323379517, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.0313, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.43012961745262146, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0296, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.49092897772789, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.0345, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.035284582525491714, | |
| "eval_runtime": 161.8654, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.2028495967388153, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.0334, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.29999735951423645, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0291, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.13542650640010834, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0278, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.1620338261127472, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.036, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.23953960835933685, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0211, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03400290384888649, | |
| "eval_runtime": 161.9151, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.19132652878761292, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0293, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.27870815992355347, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0374, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.24894696474075317, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.0266, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.5503875613212585, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0367, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.07178358733654022, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.0401, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.03488364815711975, | |
| "eval_runtime": 161.9246, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.33762502670288086, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.0299, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.0860748142004013, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0252, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.33839380741119385, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.0356, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.19909736514091492, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0346, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.24192649126052856, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0331, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.0329899899661541, | |
| "eval_runtime": 161.9407, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.20480167865753174, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.0347, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.24220184981822968, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0398, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.15550747513771057, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.031, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.1550067663192749, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.0319, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.24922066926956177, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0328, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.03308863937854767, | |
| "eval_runtime": 161.9501, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.2108086496591568, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.0273, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.1620463877916336, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.0293, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 0.14132855832576752, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.0281, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.47077223658561707, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.0311, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.1639624834060669, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.0337, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.03411152958869934, | |
| "eval_runtime": 161.8694, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.4736654460430145, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0377, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.23130185902118683, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.0272, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.2556862235069275, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.0376, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.15771634876728058, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0476, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.2260507494211197, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0212, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.03414667025208473, | |
| "eval_runtime": 161.8928, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.17024976015090942, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.0304, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.1473970264196396, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.024, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.5394685864448547, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0238, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.3650222718715668, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.039, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.10559602081775665, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0239, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.034826118499040604, | |
| "eval_runtime": 161.8929, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.4457942843437195, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.0295, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.1348329782485962, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0274, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.12303648889064789, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0305, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.2129056453704834, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.0348, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.3558931350708008, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.0309, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.033693600445985794, | |
| "eval_runtime": 161.8726, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.3543870449066162, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.0329, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.19590340554714203, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.0293, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.14328280091285706, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.023, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.20440389215946198, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0334, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.2023562490940094, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.0215, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.03142290189862251, | |
| "eval_runtime": 161.9082, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.27520495653152466, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.0215, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.3740273714065552, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.0225, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.16322128474712372, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0227, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.18465684354305267, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.021, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.24486982822418213, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0164, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.03230370208621025, | |
| "eval_runtime": 161.9101, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.10609235614538193, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0197, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.4242578446865082, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.0183, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.24461661279201508, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.0136, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.2401552051305771, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0163, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.6907597184181213, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0321, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.03365868330001831, | |
| "eval_runtime": 161.9854, | |
| "eval_samples_per_second": 7.136, | |
| "eval_steps_per_second": 7.136, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.06974344700574875, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0159, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.28437483310699463, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.024, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.31415972113609314, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.0145, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.19021229445934296, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0198, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.11239872127771378, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.013, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.03585917875170708, | |
| "eval_runtime": 161.8947, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.23577116429805756, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.0194, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.30587878823280334, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0227, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.25759395956993103, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0236, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.11709993332624435, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.026, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.2631906569004059, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0212, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.033405330032110214, | |
| "eval_runtime": 161.9005, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.2674572169780731, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.0223, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.24573783576488495, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0207, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.1961791217327118, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.018, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.2744806408882141, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0175, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.20697583258152008, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.0164, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.03636433184146881, | |
| "eval_runtime": 161.9372, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.3116421103477478, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.012, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.30894482135772705, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0241, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.15000484883785248, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0206, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.17668107151985168, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.0187, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.30187666416168213, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.016, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.034627050161361694, | |
| "eval_runtime": 161.9111, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.22439341247081757, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.0154, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.17332100868225098, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.0213, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.12855423986911774, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0207, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.31204909086227417, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.0191, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.249205082654953, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.0218, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.03257967159152031, | |
| "eval_runtime": 161.8733, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.42175859212875366, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.018, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.37882936000823975, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0185, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.24134980142116547, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0241, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.21430109441280365, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0231, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.19930411875247955, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0173, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.03270987421274185, | |
| "eval_runtime": 161.8719, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.2350929081439972, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.0187, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.2002047598361969, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.0292, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.42562997341156006, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.023, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.16100174188613892, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.0165, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.18212825059890747, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0262, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.03202894330024719, | |
| "eval_runtime": 161.9669, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.20856015384197235, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0142, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.6511904001235962, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0205, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.44803279638290405, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0239, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.10874707996845245, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.0196, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.33031442761421204, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.0246, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.034378521144390106, | |
| "eval_runtime": 161.9309, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.2869391441345215, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.0196, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.41830745339393616, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0245, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.5830449461936951, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.0231, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.2141474187374115, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.028, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.19103091955184937, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.0243, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.031263191252946854, | |
| "eval_runtime": 161.9636, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.2518385350704193, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0145, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.30439579486846924, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0258, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.5297009348869324, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0123, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.2028278261423111, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.0189, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.21385866403579712, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0162, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.032001473009586334, | |
| "eval_runtime": 161.9962, | |
| "eval_samples_per_second": 7.136, | |
| "eval_steps_per_second": 7.136, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 0.2492896467447281, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.0229, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.44225117564201355, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.0199, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.13712511956691742, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0168, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.16013561189174652, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0146, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.2570871114730835, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.029, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.032170217484235764, | |
| "eval_runtime": 161.9704, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.19998405873775482, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.0068, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.1979544758796692, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.0104, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.12661851942539215, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.0097, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.3147395849227905, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.0089, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.14858821034431458, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.0113, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.03542900085449219, | |
| "eval_runtime": 161.9641, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 0.5959571599960327, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.0065, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.144076868891716, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0113, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.10973131656646729, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.0105, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.03880462422966957, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.0087, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.30794036388397217, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0091, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.038146741688251495, | |
| "eval_runtime": 161.961, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.11566942185163498, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.0062, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 0.5766760110855103, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.0092, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.5608533620834351, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0121, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.30213215947151184, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.0094, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.22508233785629272, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.0077, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.039552241563797, | |
| "eval_runtime": 161.8803, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.13414816558361053, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.0099, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.2789548635482788, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0095, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.38610661029815674, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0067, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.2520865797996521, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0053, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.3510969579219818, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0094, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.04009649530053139, | |
| "eval_runtime": 161.8741, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.046489037573337555, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.0086, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.3832823932170868, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.0156, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.04104534536600113, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.0088, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.39937496185302734, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.0081, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.08046502619981766, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0152, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.03876478224992752, | |
| "eval_runtime": 161.8694, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.18804286420345306, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.0097, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.27762722969055176, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0078, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.10057356208562851, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.0111, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.2620813548564911, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0116, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.30753788352012634, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0099, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.03909964859485626, | |
| "eval_runtime": 161.8134, | |
| "eval_samples_per_second": 7.144, | |
| "eval_steps_per_second": 7.144, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.1349666714668274, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0078, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.6018354296684265, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.0059, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.25746870040893555, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0131, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.45929262042045593, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0075, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.14661087095737457, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0111, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.037890732288360596, | |
| "eval_runtime": 161.8515, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.1353587806224823, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0132, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.5209174156188965, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.013, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.15474364161491394, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0086, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.21484799683094025, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0112, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.16505950689315796, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0053, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.03818138688802719, | |
| "eval_runtime": 161.8537, | |
| "eval_samples_per_second": 7.142, | |
| "eval_steps_per_second": 7.142, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.33931872248649597, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0133, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.23094065487384796, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.0081, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.20208130776882172, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.0099, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.2577545940876007, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.0106, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.08922095596790314, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.0115, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.03821124881505966, | |
| "eval_runtime": 161.8354, | |
| "eval_samples_per_second": 7.143, | |
| "eval_steps_per_second": 7.143, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.16086284816265106, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.009, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.3469180762767792, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0075, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.21784304082393646, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.0079, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.4547649025917053, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0071, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.18926726281642914, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0067, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.03874478489160538, | |
| "eval_runtime": 161.8861, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.13456791639328003, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0107, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.33499956130981445, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0045, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.24915644526481628, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.0112, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.32178714871406555, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0103, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.1828913390636444, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.0084, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.038912083953619, | |
| "eval_runtime": 161.957, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.08263581991195679, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0088, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.4691775441169739, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0036, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.1671994924545288, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.0062, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.21955837309360504, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.0051, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.7991011142730713, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.0123, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.04052657261490822, | |
| "eval_runtime": 161.9569, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.37465089559555054, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0059, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.5578173398971558, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0065, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 0.02930700220167637, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0119, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.4806767404079437, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.0076, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.053415652364492416, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.0091, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.04062394052743912, | |
| "eval_runtime": 161.9501, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.12807564437389374, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0049, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.16146157681941986, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0016, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.09077801555395126, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.0018, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.03578656539320946, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0053, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.04418487846851349, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.0018, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.04137811437249184, | |
| "eval_runtime": 161.9659, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.3220917284488678, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0041, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.21681171655654907, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.002, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.049075912684202194, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0023, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.10186077654361725, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0041, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.031536947935819626, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0056, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.04437172785401344, | |
| "eval_runtime": 161.8853, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.03274516761302948, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.0035, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.13729196786880493, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0033, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.06894241273403168, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0024, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 1.0029683113098145, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.005, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.10748273134231567, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0025, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.0455368235707283, | |
| "eval_runtime": 161.8777, | |
| "eval_samples_per_second": 7.141, | |
| "eval_steps_per_second": 7.141, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.056326281279325485, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0021, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.04037575051188469, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.0014, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.2542364001274109, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.0025, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.153919517993927, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0033, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.04472075775265694, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.0022, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.04717316851019859, | |
| "eval_runtime": 161.9719, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.024519631639122963, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0016, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.032540712505578995, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0022, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.14218932390213013, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.003, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.6652355790138245, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0036, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.07552742958068848, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0026, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.048119738698005676, | |
| "eval_runtime": 161.8284, | |
| "eval_samples_per_second": 7.143, | |
| "eval_steps_per_second": 7.143, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.10875464975833893, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.0016, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.05334506183862686, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.0012, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.07590065896511078, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0082, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.046069368720054626, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0018, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.2599368095397949, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0028, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.048302020877599716, | |
| "eval_runtime": 161.9161, | |
| "eval_samples_per_second": 7.139, | |
| "eval_steps_per_second": 7.139, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.0736982598900795, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.0056, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.04838383570313454, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0036, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.03644100949168205, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.0024, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.4585660398006439, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0031, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.022391265258193016, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0016, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.04836786165833473, | |
| "eval_runtime": 161.9039, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.1084195226430893, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0006, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 0.06898686289787292, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.0024, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.01959504745900631, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.0023, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.07250258326530457, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.004, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 0.12506908178329468, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.0021, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.04873073846101761, | |
| "eval_runtime": 161.9, | |
| "eval_samples_per_second": 7.14, | |
| "eval_steps_per_second": 7.14, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.15250593423843384, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0014, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.1494806557893753, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0033, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.2010517716407776, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.004, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.9992215037345886, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0043, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.02085724286735058, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0021, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.04876425862312317, | |
| "eval_runtime": 161.9981, | |
| "eval_samples_per_second": 7.136, | |
| "eval_steps_per_second": 7.136, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.03375128284096718, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.0029, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.5665181875228882, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0022, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.6858422756195068, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0053, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.04665651544928551, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.0049, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.11667704582214355, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.003, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.048775553703308105, | |
| "eval_runtime": 161.9762, | |
| "eval_samples_per_second": 7.137, | |
| "eval_steps_per_second": 7.137, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.07030779868364334, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.005, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.2987179756164551, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.002, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.007006101310253143, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0037, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.010003631934523582, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0019, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.34444305300712585, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0047, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.04887409508228302, | |
| "eval_runtime": 161.959, | |
| "eval_samples_per_second": 7.138, | |
| "eval_steps_per_second": 7.138, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.06820440292358398, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0033, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.4628816843032837, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.0026, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.0736798495054245, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.0027, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.14411404728889465, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0019, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.059557218104600906, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.0024, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.04892992228269577, | |
| "eval_runtime": 162.0065, | |
| "eval_samples_per_second": 7.136, | |
| "eval_steps_per_second": 7.136, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.5008711814880371, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0035, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.04182853922247887, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.003, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.4152490198612213, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0056, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.17047202587127686, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0032, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 7.696760527880847e+17, | |
| "train_loss": 0.05118226883445044, | |
| "train_runtime": 38291.9901, | |
| "train_samples_per_second": 1.357, | |
| "train_steps_per_second": 0.085 | |
| } | |
| ], | |
| "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 | |
| } | |