| { | |
| "best_metric": 0.03152759373188019, | |
| "best_model_checkpoint": "saves/psy-course/Ministral-8B-Instruct-2410/train/fold4/checkpoint-1300", | |
| "epoch": 4.994228549442093, | |
| "eval_steps": 50, | |
| "global_step": 3245, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015390534821085032, | |
| "grad_norm": 2.723722457885742, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.3359, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030781069642170065, | |
| "grad_norm": 2.8121957778930664, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 1.2709, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0461716044632551, | |
| "grad_norm": 3.163454294204712, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0348, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06156213928434013, | |
| "grad_norm": 2.2602243423461914, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.5543, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "grad_norm": 1.1449053287506104, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2583, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07695267410542517, | |
| "eval_loss": 0.24188649654388428, | |
| "eval_runtime": 207.1605, | |
| "eval_samples_per_second": 5.58, | |
| "eval_steps_per_second": 5.58, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0923432089265102, | |
| "grad_norm": 1.4981391429901123, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.1592, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10773374374759523, | |
| "grad_norm": 1.1884814500808716, | |
| "learning_rate": 2.1538461538461542e-05, | |
| "loss": 0.1276, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12312427856868026, | |
| "grad_norm": 1.2275854349136353, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.1018, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1385148133897653, | |
| "grad_norm": 1.481279730796814, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.1182, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "grad_norm": 0.7386165261268616, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0849, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15390534821085033, | |
| "eval_loss": 0.06938305497169495, | |
| "eval_runtime": 203.9254, | |
| "eval_samples_per_second": 5.669, | |
| "eval_steps_per_second": 5.669, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16929588303193535, | |
| "grad_norm": 0.8244947791099548, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.0646, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1846864178530204, | |
| "grad_norm": 0.8134088516235352, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.0723, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2000769526741054, | |
| "grad_norm": 1.113705039024353, | |
| "learning_rate": 4e-05, | |
| "loss": 0.069, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21546748749519046, | |
| "grad_norm": 0.7777001261711121, | |
| "learning_rate": 4.3076923076923084e-05, | |
| "loss": 0.0783, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "grad_norm": 0.9160193800926208, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.061, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2308580223162755, | |
| "eval_loss": 0.05820224806666374, | |
| "eval_runtime": 204.2622, | |
| "eval_samples_per_second": 5.659, | |
| "eval_steps_per_second": 5.659, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24624855713736052, | |
| "grad_norm": 0.7862887382507324, | |
| "learning_rate": 4.923076923076924e-05, | |
| "loss": 0.0617, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26163909195844554, | |
| "grad_norm": 0.7871883511543274, | |
| "learning_rate": 5.230769230769231e-05, | |
| "loss": 0.0612, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2770296267795306, | |
| "grad_norm": 0.7988012433052063, | |
| "learning_rate": 5.538461538461539e-05, | |
| "loss": 0.0527, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2924201616006156, | |
| "grad_norm": 0.3650059401988983, | |
| "learning_rate": 5.846153846153847e-05, | |
| "loss": 0.0441, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "grad_norm": 0.5361356735229492, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0578, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30781069642170067, | |
| "eval_loss": 0.05375838652253151, | |
| "eval_runtime": 204.3876, | |
| "eval_samples_per_second": 5.656, | |
| "eval_steps_per_second": 5.656, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3232012312427857, | |
| "grad_norm": 0.7090181112289429, | |
| "learning_rate": 6.461538461538462e-05, | |
| "loss": 0.0557, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3385917660638707, | |
| "grad_norm": 0.829365074634552, | |
| "learning_rate": 6.76923076923077e-05, | |
| "loss": 0.0499, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3452916443347931, | |
| "learning_rate": 7.076923076923078e-05, | |
| "loss": 0.0381, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3693728357060408, | |
| "grad_norm": 0.4547441899776459, | |
| "learning_rate": 7.384615384615386e-05, | |
| "loss": 0.0461, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 0.45340439677238464, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0434, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.042893242090940475, | |
| "eval_runtime": 203.9664, | |
| "eval_samples_per_second": 5.668, | |
| "eval_steps_per_second": 5.668, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4001539053482108, | |
| "grad_norm": 0.6070353388786316, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0406, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41554444016929587, | |
| "grad_norm": 0.4273599088191986, | |
| "learning_rate": 8.307692307692309e-05, | |
| "loss": 0.0416, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4309349749903809, | |
| "grad_norm": 0.565772533416748, | |
| "learning_rate": 8.615384615384617e-05, | |
| "loss": 0.0543, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44632550981146596, | |
| "grad_norm": 0.4499136805534363, | |
| "learning_rate": 8.923076923076924e-05, | |
| "loss": 0.0513, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "grad_norm": 0.37462443113327026, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0405, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.461716044632551, | |
| "eval_loss": 0.0480533242225647, | |
| "eval_runtime": 204.3506, | |
| "eval_samples_per_second": 5.657, | |
| "eval_steps_per_second": 5.657, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.477106579453636, | |
| "grad_norm": 0.4299790561199188, | |
| "learning_rate": 9.53846153846154e-05, | |
| "loss": 0.0518, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49249711427472104, | |
| "grad_norm": 0.3406705856323242, | |
| "learning_rate": 9.846153846153848e-05, | |
| "loss": 0.0473, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5078876490958061, | |
| "grad_norm": 0.4655069410800934, | |
| "learning_rate": 9.999927654251793e-05, | |
| "loss": 0.048, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5232781839168911, | |
| "grad_norm": 0.3722781538963318, | |
| "learning_rate": 9.99934890082745e-05, | |
| "loss": 0.0449, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "grad_norm": 0.4339967966079712, | |
| "learning_rate": 9.99819146097084e-05, | |
| "loss": 0.0438, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5386687187379762, | |
| "eval_loss": 0.045366622507572174, | |
| "eval_runtime": 203.9945, | |
| "eval_samples_per_second": 5.667, | |
| "eval_steps_per_second": 5.667, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5540592535590612, | |
| "grad_norm": 0.38233867287635803, | |
| "learning_rate": 9.996455468658355e-05, | |
| "loss": 0.0575, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5694497883801463, | |
| "grad_norm": 0.4388730823993683, | |
| "learning_rate": 9.994141124835203e-05, | |
| "loss": 0.0372, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5848403232012312, | |
| "grad_norm": 0.5134220123291016, | |
| "learning_rate": 9.991248697392142e-05, | |
| "loss": 0.0385, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6002308580223162, | |
| "grad_norm": 0.5471551418304443, | |
| "learning_rate": 9.987778521134476e-05, | |
| "loss": 0.0363, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "grad_norm": 0.301911324262619, | |
| "learning_rate": 9.983730997743294e-05, | |
| "loss": 0.0486, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6156213928434013, | |
| "eval_loss": 0.044517260044813156, | |
| "eval_runtime": 203.5396, | |
| "eval_samples_per_second": 5.679, | |
| "eval_steps_per_second": 5.679, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6310119276644863, | |
| "grad_norm": 0.2548797130584717, | |
| "learning_rate": 9.979106595728978e-05, | |
| "loss": 0.0427, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6464024624855714, | |
| "grad_norm": 0.49584707617759705, | |
| "learning_rate": 9.973905850376977e-05, | |
| "loss": 0.0449, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6617929973066564, | |
| "grad_norm": 0.4419512152671814, | |
| "learning_rate": 9.968129363685833e-05, | |
| "loss": 0.0429, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6771835321277414, | |
| "grad_norm": 0.46319007873535156, | |
| "learning_rate": 9.961777804297514e-05, | |
| "loss": 0.0581, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "grad_norm": 0.3866153061389923, | |
| "learning_rate": 9.954851907420009e-05, | |
| "loss": 0.0286, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6925740669488265, | |
| "eval_loss": 0.03892648592591286, | |
| "eval_runtime": 204.167, | |
| "eval_samples_per_second": 5.662, | |
| "eval_steps_per_second": 5.662, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.25133654475212097, | |
| "learning_rate": 9.947352474742222e-05, | |
| "loss": 0.0351, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7233551365909965, | |
| "grad_norm": 0.40146341919898987, | |
| "learning_rate": 9.93928037434118e-05, | |
| "loss": 0.0318, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7387456714120816, | |
| "grad_norm": 0.5690613985061646, | |
| "learning_rate": 9.930636540581556e-05, | |
| "loss": 0.0395, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7541362062331666, | |
| "grad_norm": 0.22376570105552673, | |
| "learning_rate": 9.921421974007506e-05, | |
| "loss": 0.0443, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": 0.36554858088493347, | |
| "learning_rate": 9.911637741226849e-05, | |
| "loss": 0.0307, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": 0.03748767450451851, | |
| "eval_runtime": 203.4971, | |
| "eval_samples_per_second": 5.681, | |
| "eval_steps_per_second": 5.681, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7849172758753367, | |
| "grad_norm": 0.4653914272785187, | |
| "learning_rate": 9.90128497478762e-05, | |
| "loss": 0.0312, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8003078106964217, | |
| "grad_norm": 0.32648298144340515, | |
| "learning_rate": 9.890364873046965e-05, | |
| "loss": 0.0428, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8156983455175068, | |
| "grad_norm": 0.38220086693763733, | |
| "learning_rate": 9.878878700032427e-05, | |
| "loss": 0.0545, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8310888803385917, | |
| "grad_norm": 0.29954004287719727, | |
| "learning_rate": 9.866827785295638e-05, | |
| "loss": 0.0428, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "grad_norm": 0.2096661925315857, | |
| "learning_rate": 9.854213523758413e-05, | |
| "loss": 0.0424, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8464794151596768, | |
| "eval_loss": 0.03465096279978752, | |
| "eval_runtime": 204.2401, | |
| "eval_samples_per_second": 5.66, | |
| "eval_steps_per_second": 5.66, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8618699499807618, | |
| "grad_norm": 0.23920032382011414, | |
| "learning_rate": 9.841037375551294e-05, | |
| "loss": 0.0355, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8772604848018468, | |
| "grad_norm": 0.21398098766803741, | |
| "learning_rate": 9.827300865844527e-05, | |
| "loss": 0.0248, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8926510196229319, | |
| "grad_norm": 0.3663936257362366, | |
| "learning_rate": 9.813005584671522e-05, | |
| "loss": 0.0332, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9080415544440169, | |
| "grad_norm": 0.2479315549135208, | |
| "learning_rate": 9.79815318674481e-05, | |
| "loss": 0.0504, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "grad_norm": 0.3093143403530121, | |
| "learning_rate": 9.782745391264498e-05, | |
| "loss": 0.0372, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.923432089265102, | |
| "eval_loss": 0.03546147793531418, | |
| "eval_runtime": 203.9304, | |
| "eval_samples_per_second": 5.669, | |
| "eval_steps_per_second": 5.669, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.938822624086187, | |
| "grad_norm": 0.3326164186000824, | |
| "learning_rate": 9.76678398171927e-05, | |
| "loss": 0.0313, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.954213158907272, | |
| "grad_norm": 0.24715834856033325, | |
| "learning_rate": 9.750270805679945e-05, | |
| "loss": 0.0439, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9696036937283571, | |
| "grad_norm": 0.4079567790031433, | |
| "learning_rate": 9.733207774585618e-05, | |
| "loss": 0.0429, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9849942285494421, | |
| "grad_norm": 0.2544328272342682, | |
| "learning_rate": 9.715596863522398e-05, | |
| "loss": 0.0341, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "grad_norm": 0.29755645990371704, | |
| "learning_rate": 9.697440110994801e-05, | |
| "loss": 0.0293, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0003847633705272, | |
| "eval_loss": 0.03536020964384079, | |
| "eval_runtime": 204.0914, | |
| "eval_samples_per_second": 5.664, | |
| "eval_steps_per_second": 5.664, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0157752981916122, | |
| "grad_norm": 0.3858153522014618, | |
| "learning_rate": 9.678739618689775e-05, | |
| "loss": 0.0343, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0311658330126972, | |
| "grad_norm": 0.24061763286590576, | |
| "learning_rate": 9.659497551233431e-05, | |
| "loss": 0.0304, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0465563678337821, | |
| "grad_norm": 0.20914499461650848, | |
| "learning_rate": 9.639716135940485e-05, | |
| "loss": 0.0323, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0619469026548674, | |
| "grad_norm": 0.5369499325752258, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.045, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "grad_norm": 0.3364550471305847, | |
| "learning_rate": 9.598544482992517e-05, | |
| "loss": 0.0349, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0773374374759523, | |
| "eval_loss": 0.03680739179253578, | |
| "eval_runtime": 204.5265, | |
| "eval_samples_per_second": 5.652, | |
| "eval_steps_per_second": 5.652, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0927279722970373, | |
| "grad_norm": 0.3737899661064148, | |
| "learning_rate": 9.57715901105348e-05, | |
| "loss": 0.0317, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1081185071181223, | |
| "grad_norm": 0.24602745473384857, | |
| "learning_rate": 9.555243722158158e-05, | |
| "loss": 0.0282, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1235090419392073, | |
| "grad_norm": 0.23905214667320251, | |
| "learning_rate": 9.532801153052958e-05, | |
| "loss": 0.0293, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1388995767602923, | |
| "grad_norm": 0.5344342589378357, | |
| "learning_rate": 9.509833901518209e-05, | |
| "loss": 0.0363, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "grad_norm": 0.3220731019973755, | |
| "learning_rate": 9.486344626067475e-05, | |
| "loss": 0.0273, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1542901115813775, | |
| "eval_loss": 0.03574880212545395, | |
| "eval_runtime": 204.2631, | |
| "eval_samples_per_second": 5.659, | |
| "eval_steps_per_second": 5.659, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1696806464024625, | |
| "grad_norm": 0.31345313787460327, | |
| "learning_rate": 9.462336045639815e-05, | |
| "loss": 0.035, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1850711812235475, | |
| "grad_norm": 0.22023189067840576, | |
| "learning_rate": 9.437810939285068e-05, | |
| "loss": 0.0351, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.2004617160446325, | |
| "grad_norm": 0.3565050959587097, | |
| "learning_rate": 9.412772145842165e-05, | |
| "loss": 0.0341, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2158522508657175, | |
| "grad_norm": 0.351315975189209, | |
| "learning_rate": 9.387222563610535e-05, | |
| "loss": 0.0333, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "grad_norm": 0.4306800961494446, | |
| "learning_rate": 9.361165150014606e-05, | |
| "loss": 0.0284, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2312427856868027, | |
| "eval_loss": 0.033947862684726715, | |
| "eval_runtime": 203.8218, | |
| "eval_samples_per_second": 5.672, | |
| "eval_steps_per_second": 5.672, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2466333205078877, | |
| "grad_norm": 0.1783798635005951, | |
| "learning_rate": 9.334602921261492e-05, | |
| "loss": 0.029, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2620238553289727, | |
| "grad_norm": 0.382519006729126, | |
| "learning_rate": 9.307538951991852e-05, | |
| "loss": 0.0268, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2774143901500576, | |
| "grad_norm": 0.30088889598846436, | |
| "learning_rate": 9.279976374923988e-05, | |
| "loss": 0.0323, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2928049249711426, | |
| "grad_norm": 0.37487244606018066, | |
| "learning_rate": 9.251918380491238e-05, | |
| "loss": 0.0259, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "grad_norm": 0.32325953245162964, | |
| "learning_rate": 9.223368216472668e-05, | |
| "loss": 0.0299, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3081954597922278, | |
| "eval_loss": 0.03348197042942047, | |
| "eval_runtime": 204.3814, | |
| "eval_samples_per_second": 5.656, | |
| "eval_steps_per_second": 5.656, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.3235859946133128, | |
| "grad_norm": 0.2362068146467209, | |
| "learning_rate": 9.194329187617127e-05, | |
| "loss": 0.027, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3389765294343978, | |
| "grad_norm": 0.275078684091568, | |
| "learning_rate": 9.164804655260735e-05, | |
| "loss": 0.0255, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3543670642554828, | |
| "grad_norm": 0.17166070640087128, | |
| "learning_rate": 9.134798036937778e-05, | |
| "loss": 0.0306, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.3697575990765678, | |
| "grad_norm": 0.270987331867218, | |
| "learning_rate": 9.104312805985134e-05, | |
| "loss": 0.0368, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "grad_norm": 0.2537315785884857, | |
| "learning_rate": 9.073352491140221e-05, | |
| "loss": 0.0218, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.385148133897653, | |
| "eval_loss": 0.03290697559714317, | |
| "eval_runtime": 204.8882, | |
| "eval_samples_per_second": 5.642, | |
| "eval_steps_per_second": 5.642, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.400538668718738, | |
| "grad_norm": 0.3661804497241974, | |
| "learning_rate": 9.041920676132543e-05, | |
| "loss": 0.0299, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.415929203539823, | |
| "grad_norm": 0.255628377199173, | |
| "learning_rate": 9.010020999268854e-05, | |
| "loss": 0.0348, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.431319738360908, | |
| "grad_norm": 0.24899516999721527, | |
| "learning_rate": 8.977657153012027e-05, | |
| "loss": 0.0274, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.446710273181993, | |
| "grad_norm": 0.5146085619926453, | |
| "learning_rate": 8.944832883553633e-05, | |
| "loss": 0.0284, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "grad_norm": 0.14211301505565643, | |
| "learning_rate": 8.91155199038032e-05, | |
| "loss": 0.0368, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4621008080030782, | |
| "eval_loss": 0.03461240977048874, | |
| "eval_runtime": 204.423, | |
| "eval_samples_per_second": 5.655, | |
| "eval_steps_per_second": 5.655, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4774913428241632, | |
| "grad_norm": 0.2752813696861267, | |
| "learning_rate": 8.877818325834006e-05, | |
| "loss": 0.0291, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4928818776452482, | |
| "grad_norm": 0.11674728989601135, | |
| "learning_rate": 8.843635794665961e-05, | |
| "loss": 0.0227, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.5082724124663334, | |
| "grad_norm": 0.3706522285938263, | |
| "learning_rate": 8.809008353584829e-05, | |
| "loss": 0.0296, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5236629472874181, | |
| "grad_norm": 0.2965226471424103, | |
| "learning_rate": 8.773940010798627e-05, | |
| "loss": 0.0293, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "grad_norm": 0.3654739260673523, | |
| "learning_rate": 8.738434825550785e-05, | |
| "loss": 0.0314, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5390534821085033, | |
| "eval_loss": 0.03569772094488144, | |
| "eval_runtime": 204.8803, | |
| "eval_samples_per_second": 5.642, | |
| "eval_steps_per_second": 5.642, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5544440169295883, | |
| "grad_norm": 0.2630115747451782, | |
| "learning_rate": 8.702496907650274e-05, | |
| "loss": 0.0339, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.5698345517506733, | |
| "grad_norm": 0.3185119926929474, | |
| "learning_rate": 8.666130416995897e-05, | |
| "loss": 0.0342, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5852250865717585, | |
| "grad_norm": 0.2420577108860016, | |
| "learning_rate": 8.629339563094758e-05, | |
| "loss": 0.0282, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6006156213928433, | |
| "grad_norm": 0.38299399614334106, | |
| "learning_rate": 8.592128604575014e-05, | |
| "loss": 0.0274, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "grad_norm": 0.576649010181427, | |
| "learning_rate": 8.554501848692921e-05, | |
| "loss": 0.0323, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6160061562139285, | |
| "eval_loss": 0.032495949417352676, | |
| "eval_runtime": 204.9427, | |
| "eval_samples_per_second": 5.641, | |
| "eval_steps_per_second": 5.641, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6313966910350135, | |
| "grad_norm": 0.1790565550327301, | |
| "learning_rate": 8.51646365083426e-05, | |
| "loss": 0.0227, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6467872258560985, | |
| "grad_norm": 0.2587273418903351, | |
| "learning_rate": 8.478018414010191e-05, | |
| "loss": 0.0288, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6621777606771835, | |
| "grad_norm": 0.7771003246307373, | |
| "learning_rate": 8.439170588347598e-05, | |
| "loss": 0.0247, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6775682954982685, | |
| "grad_norm": 0.6521177291870117, | |
| "learning_rate": 8.399924670573969e-05, | |
| "loss": 0.0289, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "grad_norm": 0.1374584585428238, | |
| "learning_rate": 8.360285203496894e-05, | |
| "loss": 0.0337, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6929588303193537, | |
| "eval_loss": 0.03624284267425537, | |
| "eval_runtime": 204.8407, | |
| "eval_samples_per_second": 5.643, | |
| "eval_steps_per_second": 5.643, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.7083493651404387, | |
| "grad_norm": 0.4112257659435272, | |
| "learning_rate": 8.320256775478228e-05, | |
| "loss": 0.0315, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.7237398999615237, | |
| "grad_norm": 0.17391926050186157, | |
| "learning_rate": 8.279844019902968e-05, | |
| "loss": 0.0236, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.45332232117652893, | |
| "learning_rate": 8.239051614642934e-05, | |
| "loss": 0.034, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.7545209696036936, | |
| "grad_norm": 0.23857010900974274, | |
| "learning_rate": 8.197884281515292e-05, | |
| "loss": 0.0387, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "grad_norm": 0.31684473156929016, | |
| "learning_rate": 8.156346785735997e-05, | |
| "loss": 0.0232, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7699115044247788, | |
| "eval_loss": 0.033525850623846054, | |
| "eval_runtime": 205.0596, | |
| "eval_samples_per_second": 5.637, | |
| "eval_steps_per_second": 5.637, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7853020392458638, | |
| "grad_norm": 0.1595790982246399, | |
| "learning_rate": 8.114443935368198e-05, | |
| "loss": 0.0322, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8006925740669488, | |
| "grad_norm": 0.3000686466693878, | |
| "learning_rate": 8.072180580765709e-05, | |
| "loss": 0.025, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.8160831088880338, | |
| "grad_norm": 0.23324336111545563, | |
| "learning_rate": 8.02956161401155e-05, | |
| "loss": 0.0249, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8314736437091188, | |
| "grad_norm": 0.27352774143218994, | |
| "learning_rate": 7.98659196835169e-05, | |
| "loss": 0.0297, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "grad_norm": 0.1584983468055725, | |
| "learning_rate": 7.943276617624014e-05, | |
| "loss": 0.0209, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.846864178530204, | |
| "eval_loss": 0.03702833503484726, | |
| "eval_runtime": 205.1099, | |
| "eval_samples_per_second": 5.636, | |
| "eval_steps_per_second": 5.636, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.862254713351289, | |
| "grad_norm": 0.39062175154685974, | |
| "learning_rate": 7.899620575682578e-05, | |
| "loss": 0.0295, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.877645248172374, | |
| "grad_norm": 0.16856245696544647, | |
| "learning_rate": 7.85562889581725e-05, | |
| "loss": 0.0225, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.893035782993459, | |
| "grad_norm": 0.12971876561641693, | |
| "learning_rate": 7.811306670168788e-05, | |
| "loss": 0.0277, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.908426317814544, | |
| "grad_norm": 0.31323835253715515, | |
| "learning_rate": 7.766659029139393e-05, | |
| "loss": 0.0338, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "grad_norm": 0.41592398285865784, | |
| "learning_rate": 7.72169114079887e-05, | |
| "loss": 0.0284, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9238168526356292, | |
| "eval_loss": 0.03319057822227478, | |
| "eval_runtime": 205.5318, | |
| "eval_samples_per_second": 5.624, | |
| "eval_steps_per_second": 5.624, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9392073874567142, | |
| "grad_norm": 0.7021281719207764, | |
| "learning_rate": 7.676408210286407e-05, | |
| "loss": 0.0299, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.9545979222777992, | |
| "grad_norm": 0.23839502036571503, | |
| "learning_rate": 7.630815479208066e-05, | |
| "loss": 0.029, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9699884570988841, | |
| "grad_norm": 0.24497456848621368, | |
| "learning_rate": 7.584918225030053e-05, | |
| "loss": 0.0234, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9853789919199691, | |
| "grad_norm": 0.5139015316963196, | |
| "learning_rate": 7.538721760467845e-05, | |
| "loss": 0.0306, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "grad_norm": 0.13878822326660156, | |
| "learning_rate": 7.49223143287122e-05, | |
| "loss": 0.0218, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.0007695267410543, | |
| "eval_loss": 0.03152759373188019, | |
| "eval_runtime": 205.3179, | |
| "eval_samples_per_second": 5.63, | |
| "eval_steps_per_second": 5.63, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.016160061562139, | |
| "grad_norm": 0.2881993353366852, | |
| "learning_rate": 7.445452623605307e-05, | |
| "loss": 0.0183, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0315505963832243, | |
| "grad_norm": 0.31737855076789856, | |
| "learning_rate": 7.398390747427662e-05, | |
| "loss": 0.0179, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.0469411312043095, | |
| "grad_norm": 0.191533625125885, | |
| "learning_rate": 7.351051251861501e-05, | |
| "loss": 0.0177, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.0623316660253943, | |
| "grad_norm": 0.07588755339384079, | |
| "learning_rate": 7.303439616565146e-05, | |
| "loss": 0.0213, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "grad_norm": 0.16579431295394897, | |
| "learning_rate": 7.255561352697732e-05, | |
| "loss": 0.0168, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0777222008464795, | |
| "eval_loss": 0.03168359026312828, | |
| "eval_runtime": 205.4243, | |
| "eval_samples_per_second": 5.627, | |
| "eval_steps_per_second": 5.627, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0931127356675643, | |
| "grad_norm": 0.13742847740650177, | |
| "learning_rate": 7.207422002281281e-05, | |
| "loss": 0.0185, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.1085032704886495, | |
| "grad_norm": 0.21488317847251892, | |
| "learning_rate": 7.159027137559197e-05, | |
| "loss": 0.0154, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.1238938053097347, | |
| "grad_norm": 0.2521539628505707, | |
| "learning_rate": 7.110382360351267e-05, | |
| "loss": 0.0129, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1392843401308195, | |
| "grad_norm": 0.17760948836803436, | |
| "learning_rate": 7.061493301405244e-05, | |
| "loss": 0.0151, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "grad_norm": 0.4586429297924042, | |
| "learning_rate": 7.012365619745064e-05, | |
| "loss": 0.0227, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1546748749519047, | |
| "eval_loss": 0.03400878235697746, | |
| "eval_runtime": 205.5666, | |
| "eval_samples_per_second": 5.623, | |
| "eval_steps_per_second": 5.623, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1700654097729895, | |
| "grad_norm": 0.078793466091156, | |
| "learning_rate": 6.963005002015808e-05, | |
| "loss": 0.0134, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1854559445940747, | |
| "grad_norm": 0.25502467155456543, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0205, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.20084647941516, | |
| "grad_norm": 0.22734858095645905, | |
| "learning_rate": 6.863607839083507e-05, | |
| "loss": 0.0119, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.2162370142362446, | |
| "grad_norm": 0.3715822696685791, | |
| "learning_rate": 6.813582799336624e-05, | |
| "loss": 0.0196, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "grad_norm": 0.11712564527988434, | |
| "learning_rate": 6.763347833101192e-05, | |
| "loss": 0.0105, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.23162754905733, | |
| "eval_loss": 0.03768837824463844, | |
| "eval_runtime": 205.7631, | |
| "eval_samples_per_second": 5.618, | |
| "eval_steps_per_second": 5.618, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2470180838784146, | |
| "grad_norm": 0.4459506571292877, | |
| "learning_rate": 6.712908755193094e-05, | |
| "loss": 0.0184, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.2624086186995, | |
| "grad_norm": 0.10984374582767487, | |
| "learning_rate": 6.662271404054621e-05, | |
| "loss": 0.0222, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2777991535205846, | |
| "grad_norm": 0.30882224440574646, | |
| "learning_rate": 6.611441641078657e-05, | |
| "loss": 0.0182, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.29318968834167, | |
| "grad_norm": 0.24357309937477112, | |
| "learning_rate": 6.560425349930208e-05, | |
| "loss": 0.0199, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "grad_norm": 0.7325841188430786, | |
| "learning_rate": 6.509228435865357e-05, | |
| "loss": 0.0195, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.308580223162755, | |
| "eval_loss": 0.03477541357278824, | |
| "eval_runtime": 205.7086, | |
| "eval_samples_per_second": 5.62, | |
| "eval_steps_per_second": 5.62, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.32397075798384, | |
| "grad_norm": 0.16104528307914734, | |
| "learning_rate": 6.45785682504772e-05, | |
| "loss": 0.0213, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.339361292804925, | |
| "grad_norm": 0.24331380426883698, | |
| "learning_rate": 6.406316463862467e-05, | |
| "loss": 0.0176, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3547518276260098, | |
| "grad_norm": 0.07721444219350815, | |
| "learning_rate": 6.35461331822803e-05, | |
| "loss": 0.016, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.370142362447095, | |
| "grad_norm": 0.2456425130367279, | |
| "learning_rate": 6.302753372905515e-05, | |
| "loss": 0.0182, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "grad_norm": 0.3821743428707123, | |
| "learning_rate": 6.250742630805971e-05, | |
| "loss": 0.0154, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.38553289726818, | |
| "eval_loss": 0.036726828664541245, | |
| "eval_runtime": 205.7487, | |
| "eval_samples_per_second": 5.619, | |
| "eval_steps_per_second": 5.619, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.400923432089265, | |
| "grad_norm": 0.24071243405342102, | |
| "learning_rate": 6.198587112295526e-05, | |
| "loss": 0.0138, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.41631396691035, | |
| "grad_norm": 0.46907514333724976, | |
| "learning_rate": 6.14629285449852e-05, | |
| "loss": 0.0175, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.431704501731435, | |
| "grad_norm": 0.14040012657642365, | |
| "learning_rate": 6.093865910598688e-05, | |
| "loss": 0.0203, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.44709503655252, | |
| "grad_norm": 0.2829310894012451, | |
| "learning_rate": 6.0413123491385025e-05, | |
| "loss": 0.0181, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "grad_norm": 0.24676255881786346, | |
| "learning_rate": 5.988638253316713e-05, | |
| "loss": 0.0125, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4624855713736054, | |
| "eval_loss": 0.03492435812950134, | |
| "eval_runtime": 205.471, | |
| "eval_samples_per_second": 5.626, | |
| "eval_steps_per_second": 5.626, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.47787610619469, | |
| "grad_norm": 0.23506391048431396, | |
| "learning_rate": 5.935849720284206e-05, | |
| "loss": 0.0156, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.4932666410157753, | |
| "grad_norm": 0.23212280869483948, | |
| "learning_rate": 5.8829528604382436e-05, | |
| "loss": 0.0212, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.50865717583686, | |
| "grad_norm": 0.17145459353923798, | |
| "learning_rate": 5.829953796715176e-05, | |
| "loss": 0.0165, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.5240477106579453, | |
| "grad_norm": 0.44674772024154663, | |
| "learning_rate": 5.7768586638816924e-05, | |
| "loss": 0.0209, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "grad_norm": 0.18063971400260925, | |
| "learning_rate": 5.723673607824705e-05, | |
| "loss": 0.0189, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5394382454790305, | |
| "eval_loss": 0.03426335006952286, | |
| "eval_runtime": 205.7619, | |
| "eval_samples_per_second": 5.618, | |
| "eval_steps_per_second": 5.618, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5548287803001153, | |
| "grad_norm": 0.2853492796421051, | |
| "learning_rate": 5.670404784839953e-05, | |
| "loss": 0.0142, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5702193151212005, | |
| "grad_norm": 0.35591042041778564, | |
| "learning_rate": 5.6170583609193984e-05, | |
| "loss": 0.0174, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.5856098499422853, | |
| "grad_norm": 0.4493762254714966, | |
| "learning_rate": 5.56364051103749e-05, | |
| "loss": 0.0203, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.6010003847633705, | |
| "grad_norm": 0.14954905211925507, | |
| "learning_rate": 5.5101574184364056e-05, | |
| "loss": 0.0192, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "grad_norm": 0.41614750027656555, | |
| "learning_rate": 5.4566152739103116e-05, | |
| "loss": 0.0211, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6163909195844557, | |
| "eval_loss": 0.03510381281375885, | |
| "eval_runtime": 205.6758, | |
| "eval_samples_per_second": 5.62, | |
| "eval_steps_per_second": 5.62, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.6317814544055405, | |
| "grad_norm": 0.35076820850372314, | |
| "learning_rate": 5.403020275088786e-05, | |
| "loss": 0.0167, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.6471719892266257, | |
| "grad_norm": 0.4516291618347168, | |
| "learning_rate": 5.349378625719408e-05, | |
| "loss": 0.0237, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6625625240477104, | |
| "grad_norm": 0.5218458771705627, | |
| "learning_rate": 5.2956965349496743e-05, | |
| "loss": 0.019, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6779530588687956, | |
| "grad_norm": 0.09230326861143112, | |
| "learning_rate": 5.241980216608261e-05, | |
| "loss": 0.0123, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "grad_norm": 0.07761550694704056, | |
| "learning_rate": 5.188235888485773e-05, | |
| "loss": 0.0223, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.693343593689881, | |
| "eval_loss": 0.03282194584608078, | |
| "eval_runtime": 205.5202, | |
| "eval_samples_per_second": 5.625, | |
| "eval_steps_per_second": 5.625, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.7087341285109656, | |
| "grad_norm": 0.2341829240322113, | |
| "learning_rate": 5.13446977161501e-05, | |
| "loss": 0.0136, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.724124663332051, | |
| "grad_norm": 0.25585412979125977, | |
| "learning_rate": 5.0806880895508766e-05, | |
| "loss": 0.0168, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.7395151981531356, | |
| "grad_norm": 0.6785100102424622, | |
| "learning_rate": 5.026897067649976e-05, | |
| "loss": 0.0239, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.754905732974221, | |
| "grad_norm": 0.21607786417007446, | |
| "learning_rate": 4.973102932350026e-05, | |
| "loss": 0.0174, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "grad_norm": 0.3086630403995514, | |
| "learning_rate": 4.9193119104491245e-05, | |
| "loss": 0.018, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.770296267795306, | |
| "eval_loss": 0.032899610698223114, | |
| "eval_runtime": 205.5601, | |
| "eval_samples_per_second": 5.624, | |
| "eval_steps_per_second": 5.624, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.785686802616391, | |
| "grad_norm": 0.3963531255722046, | |
| "learning_rate": 4.86553022838499e-05, | |
| "loss": 0.0199, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.801077337437476, | |
| "grad_norm": 0.4831998646259308, | |
| "learning_rate": 4.8117641115142285e-05, | |
| "loss": 0.0213, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.8164678722585608, | |
| "grad_norm": 0.46783447265625, | |
| "learning_rate": 4.758019783391741e-05, | |
| "loss": 0.0196, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.831858407079646, | |
| "grad_norm": 0.2343854308128357, | |
| "learning_rate": 4.704303465050327e-05, | |
| "loss": 0.0188, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "grad_norm": 0.13035911321640015, | |
| "learning_rate": 4.6506213742805924e-05, | |
| "loss": 0.0199, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.847248941900731, | |
| "eval_loss": 0.03257604315876961, | |
| "eval_runtime": 206.0681, | |
| "eval_samples_per_second": 5.61, | |
| "eval_steps_per_second": 5.61, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.862639476721816, | |
| "grad_norm": 0.2662080228328705, | |
| "learning_rate": 4.5969797249112146e-05, | |
| "loss": 0.0137, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.878030011542901, | |
| "grad_norm": 0.255823016166687, | |
| "learning_rate": 4.54338472608969e-05, | |
| "loss": 0.0247, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.893420546363986, | |
| "grad_norm": 0.21525438129901886, | |
| "learning_rate": 4.489842581563597e-05, | |
| "loss": 0.0155, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.908811081185071, | |
| "grad_norm": 0.1422373503446579, | |
| "learning_rate": 4.436359488962511e-05, | |
| "loss": 0.0169, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "grad_norm": 0.17372463643550873, | |
| "learning_rate": 4.382941639080602e-05, | |
| "loss": 0.0176, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.9242016160061564, | |
| "eval_loss": 0.03274904936552048, | |
| "eval_runtime": 205.6606, | |
| "eval_samples_per_second": 5.621, | |
| "eval_steps_per_second": 5.621, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.939592150827241, | |
| "grad_norm": 1.1632893085479736, | |
| "learning_rate": 4.3295952151600475e-05, | |
| "loss": 0.0221, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9549826856483263, | |
| "grad_norm": 0.4008699059486389, | |
| "learning_rate": 4.276326392175296e-05, | |
| "loss": 0.0212, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.970373220469411, | |
| "grad_norm": 0.11178206652402878, | |
| "learning_rate": 4.2231413361183095e-05, | |
| "loss": 0.0151, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.9857637552904963, | |
| "grad_norm": 0.3245430588722229, | |
| "learning_rate": 4.1700462032848244e-05, | |
| "loss": 0.0133, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "grad_norm": 0.15760967135429382, | |
| "learning_rate": 4.117047139561758e-05, | |
| "loss": 0.0235, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0011542901115815, | |
| "eval_loss": 0.03375434875488281, | |
| "eval_runtime": 205.5345, | |
| "eval_samples_per_second": 5.624, | |
| "eval_steps_per_second": 5.624, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.0165448249326663, | |
| "grad_norm": 0.06392020732164383, | |
| "learning_rate": 4.0641502797157966e-05, | |
| "loss": 0.0061, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.0319353597537515, | |
| "grad_norm": 0.05534449592232704, | |
| "learning_rate": 4.011361746683288e-05, | |
| "loss": 0.0101, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0473258945748363, | |
| "grad_norm": 0.1328858733177185, | |
| "learning_rate": 3.958687650861499e-05, | |
| "loss": 0.008, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0627164293959215, | |
| "grad_norm": 0.171275794506073, | |
| "learning_rate": 3.906134089401313e-05, | |
| "loss": 0.0059, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "grad_norm": 0.15738222002983093, | |
| "learning_rate": 3.853707145501482e-05, | |
| "loss": 0.008, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0781069642170067, | |
| "eval_loss": 0.03930912911891937, | |
| "eval_runtime": 205.627, | |
| "eval_samples_per_second": 5.622, | |
| "eval_steps_per_second": 5.622, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0934974990380915, | |
| "grad_norm": 0.40708914399147034, | |
| "learning_rate": 3.801412887704475e-05, | |
| "loss": 0.005, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.1088880338591767, | |
| "grad_norm": 0.05073259398341179, | |
| "learning_rate": 3.749257369194029e-05, | |
| "loss": 0.0093, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.1242785686802614, | |
| "grad_norm": 0.21584106981754303, | |
| "learning_rate": 3.697246627094487e-05, | |
| "loss": 0.0099, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.1396691035013466, | |
| "grad_norm": 0.04678943380713463, | |
| "learning_rate": 3.6453866817719726e-05, | |
| "loss": 0.0081, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "grad_norm": 0.40781915187835693, | |
| "learning_rate": 3.593683536137533e-05, | |
| "loss": 0.0096, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.155059638322432, | |
| "eval_loss": 0.040919192135334015, | |
| "eval_runtime": 205.5663, | |
| "eval_samples_per_second": 5.623, | |
| "eval_steps_per_second": 5.623, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1704501731435166, | |
| "grad_norm": 0.03877325356006622, | |
| "learning_rate": 3.542143174952282e-05, | |
| "loss": 0.0058, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.185840707964602, | |
| "grad_norm": 0.5559370517730713, | |
| "learning_rate": 3.490771564134643e-05, | |
| "loss": 0.0086, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.2012312427856866, | |
| "grad_norm": 0.2304099053144455, | |
| "learning_rate": 3.4395746500697925e-05, | |
| "loss": 0.0087, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.216621777606772, | |
| "grad_norm": 0.43068602681159973, | |
| "learning_rate": 3.388558358921343e-05, | |
| "loss": 0.0078, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "grad_norm": 0.3900976777076721, | |
| "learning_rate": 3.337728595945378e-05, | |
| "loss": 0.0077, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.232012312427857, | |
| "eval_loss": 0.03953250125050545, | |
| "eval_runtime": 205.7063, | |
| "eval_samples_per_second": 5.62, | |
| "eval_steps_per_second": 5.62, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.247402847248942, | |
| "grad_norm": 0.07008818536996841, | |
| "learning_rate": 3.287091244806907e-05, | |
| "loss": 0.008, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.262793382070027, | |
| "grad_norm": 0.41572317481040955, | |
| "learning_rate": 3.2366521668988106e-05, | |
| "loss": 0.0088, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.2781839168911118, | |
| "grad_norm": 0.4126039445400238, | |
| "learning_rate": 3.186417200663379e-05, | |
| "loss": 0.0077, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.293574451712197, | |
| "grad_norm": 0.8514243364334106, | |
| "learning_rate": 3.136392160916495e-05, | |
| "loss": 0.0072, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "grad_norm": 0.3819625675678253, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0057, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.308964986533282, | |
| "eval_loss": 0.041769955307245255, | |
| "eval_runtime": 205.7847, | |
| "eval_samples_per_second": 5.618, | |
| "eval_steps_per_second": 5.618, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.324355521354367, | |
| "grad_norm": 0.19759462773799896, | |
| "learning_rate": 3.036994997984194e-05, | |
| "loss": 0.0082, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.339746056175452, | |
| "grad_norm": 0.25018686056137085, | |
| "learning_rate": 2.9876343802549357e-05, | |
| "loss": 0.0107, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.355136590996537, | |
| "grad_norm": 0.2334885597229004, | |
| "learning_rate": 2.938506698594755e-05, | |
| "loss": 0.0086, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.370527125817622, | |
| "grad_norm": 0.3894292712211609, | |
| "learning_rate": 2.8896176396487334e-05, | |
| "loss": 0.0071, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "grad_norm": 0.10237349569797516, | |
| "learning_rate": 2.8409728624408067e-05, | |
| "loss": 0.0115, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3859176606387074, | |
| "eval_loss": 0.038854435086250305, | |
| "eval_runtime": 205.9546, | |
| "eval_samples_per_second": 5.613, | |
| "eval_steps_per_second": 5.613, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.401308195459792, | |
| "grad_norm": 0.1161830723285675, | |
| "learning_rate": 2.7925779977187215e-05, | |
| "loss": 0.007, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.4166987302808773, | |
| "grad_norm": 0.263019859790802, | |
| "learning_rate": 2.7444386473022686e-05, | |
| "loss": 0.0083, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.432089265101962, | |
| "grad_norm": 0.18079723417758942, | |
| "learning_rate": 2.696560383434854e-05, | |
| "loss": 0.0121, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.4474797999230473, | |
| "grad_norm": 0.20434071123600006, | |
| "learning_rate": 2.6489487481384988e-05, | |
| "loss": 0.0124, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "grad_norm": 0.12657655775547028, | |
| "learning_rate": 2.601609252572339e-05, | |
| "loss": 0.0076, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4628703347441325, | |
| "eval_loss": 0.03708665817975998, | |
| "eval_runtime": 205.6181, | |
| "eval_samples_per_second": 5.622, | |
| "eval_steps_per_second": 5.622, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.07753147929906845, | |
| "learning_rate": 2.554547376394692e-05, | |
| "loss": 0.0057, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4936514043863025, | |
| "grad_norm": 0.20280352234840393, | |
| "learning_rate": 2.5077685671287808e-05, | |
| "loss": 0.0052, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.5090419392073873, | |
| "grad_norm": 0.1367408186197281, | |
| "learning_rate": 2.4612782395321593e-05, | |
| "loss": 0.0086, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.5244324740284725, | |
| "grad_norm": 0.2880767285823822, | |
| "learning_rate": 2.4150817749699496e-05, | |
| "loss": 0.0066, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "grad_norm": 0.139032244682312, | |
| "learning_rate": 2.3691845207919357e-05, | |
| "loss": 0.0114, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5398230088495577, | |
| "eval_loss": 0.03903697803616524, | |
| "eval_runtime": 206.0643, | |
| "eval_samples_per_second": 5.61, | |
| "eval_steps_per_second": 5.61, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5552135436706425, | |
| "grad_norm": 0.044838063418865204, | |
| "learning_rate": 2.3235917897135934e-05, | |
| "loss": 0.0089, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.5706040784917277, | |
| "grad_norm": 0.4467596113681793, | |
| "learning_rate": 2.2783088592011303e-05, | |
| "loss": 0.0098, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5859946133128124, | |
| "grad_norm": 0.11289315670728683, | |
| "learning_rate": 2.233340970860608e-05, | |
| "loss": 0.0058, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.6013851481338977, | |
| "grad_norm": 0.6711908578872681, | |
| "learning_rate": 2.1886933298312125e-05, | |
| "loss": 0.0114, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "grad_norm": 0.1004253402352333, | |
| "learning_rate": 2.1443711041827503e-05, | |
| "loss": 0.0058, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.616775682954983, | |
| "eval_loss": 0.0413433276116848, | |
| "eval_runtime": 179.1526, | |
| "eval_samples_per_second": 6.453, | |
| "eval_steps_per_second": 6.453, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.6321662177760676, | |
| "grad_norm": 0.30656421184539795, | |
| "learning_rate": 2.1003794243174252e-05, | |
| "loss": 0.0164, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.647556752597153, | |
| "grad_norm": 0.2508336305618286, | |
| "learning_rate": 2.0567233823759886e-05, | |
| "loss": 0.0093, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.6629472874182376, | |
| "grad_norm": 0.1813203990459442, | |
| "learning_rate": 2.0134080316483113e-05, | |
| "loss": 0.0101, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.678337822239323, | |
| "grad_norm": 0.27697306871414185, | |
| "learning_rate": 1.970438385988452e-05, | |
| "loss": 0.0115, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "grad_norm": 0.12298834323883057, | |
| "learning_rate": 1.927819419234292e-05, | |
| "loss": 0.0118, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.693728357060408, | |
| "eval_loss": 0.03708096966147423, | |
| "eval_runtime": 177.4502, | |
| "eval_samples_per_second": 6.515, | |
| "eval_steps_per_second": 6.515, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.709118891881493, | |
| "grad_norm": 0.1878751665353775, | |
| "learning_rate": 1.885556064631801e-05, | |
| "loss": 0.0087, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.724509426702578, | |
| "grad_norm": 0.11255352199077606, | |
| "learning_rate": 1.843653214264003e-05, | |
| "loss": 0.0068, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.7398999615236628, | |
| "grad_norm": 0.030022628605365753, | |
| "learning_rate": 1.8021157184847087e-05, | |
| "loss": 0.0099, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.755290496344748, | |
| "grad_norm": 0.12099307775497437, | |
| "learning_rate": 1.7609483853570668e-05, | |
| "loss": 0.0067, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "grad_norm": 0.02878260426223278, | |
| "learning_rate": 1.720155980097034e-05, | |
| "loss": 0.0075, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.770681031165833, | |
| "eval_loss": 0.03793875500559807, | |
| "eval_runtime": 177.3943, | |
| "eval_samples_per_second": 6.517, | |
| "eval_steps_per_second": 6.517, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.786071565986918, | |
| "grad_norm": 0.25587835907936096, | |
| "learning_rate": 1.6797432245217736e-05, | |
| "loss": 0.0063, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.801462100808003, | |
| "grad_norm": 0.2432209849357605, | |
| "learning_rate": 1.6397147965031072e-05, | |
| "loss": 0.0054, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.816852635629088, | |
| "grad_norm": 0.27390816807746887, | |
| "learning_rate": 1.600075329426033e-05, | |
| "loss": 0.0134, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.832243170450173, | |
| "grad_norm": 0.09597912430763245, | |
| "learning_rate": 1.560829411652403e-05, | |
| "loss": 0.0082, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "grad_norm": 0.2104773223400116, | |
| "learning_rate": 1.521981585989809e-05, | |
| "loss": 0.0072, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8476337052712584, | |
| "eval_loss": 0.039310846477746964, | |
| "eval_runtime": 199.9173, | |
| "eval_samples_per_second": 5.782, | |
| "eval_steps_per_second": 5.782, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.863024240092343, | |
| "grad_norm": 0.057136524468660355, | |
| "learning_rate": 1.4835363491657412e-05, | |
| "loss": 0.0098, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8784147749134283, | |
| "grad_norm": 0.30410435795783997, | |
| "learning_rate": 1.4454981513070791e-05, | |
| "loss": 0.0052, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.893805309734513, | |
| "grad_norm": 0.057996392250061035, | |
| "learning_rate": 1.4078713954249873e-05, | |
| "loss": 0.0062, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.9091958445555983, | |
| "grad_norm": 0.06961487978696823, | |
| "learning_rate": 1.3706604369052434e-05, | |
| "loss": 0.0051, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "grad_norm": 0.48686274886131287, | |
| "learning_rate": 1.3338695830041047e-05, | |
| "loss": 0.0094, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9245863793766835, | |
| "eval_loss": 0.03958968073129654, | |
| "eval_runtime": 204.9587, | |
| "eval_samples_per_second": 5.64, | |
| "eval_steps_per_second": 5.64, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.9399769141977683, | |
| "grad_norm": 0.3194223642349243, | |
| "learning_rate": 1.2975030923497262e-05, | |
| "loss": 0.0068, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.9553674490188535, | |
| "grad_norm": 0.38718247413635254, | |
| "learning_rate": 1.2615651744492151e-05, | |
| "loss": 0.0057, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9707579838399383, | |
| "grad_norm": 0.03030051290988922, | |
| "learning_rate": 1.2260599892013719e-05, | |
| "loss": 0.0106, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9861485186610235, | |
| "grad_norm": 0.23781616985797882, | |
| "learning_rate": 1.1909916464151715e-05, | |
| "loss": 0.0074, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "grad_norm": 0.054951585829257965, | |
| "learning_rate": 1.1563642053340406e-05, | |
| "loss": 0.0096, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.001539053482109, | |
| "eval_loss": 0.03961732238531113, | |
| "eval_runtime": 204.3777, | |
| "eval_samples_per_second": 5.656, | |
| "eval_steps_per_second": 5.656, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.016929588303194, | |
| "grad_norm": 0.04557346552610397, | |
| "learning_rate": 1.1221816741659964e-05, | |
| "loss": 0.0037, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.032320123124278, | |
| "grad_norm": 0.12311459332704544, | |
| "learning_rate": 1.0884480096196803e-05, | |
| "loss": 0.0027, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.047710657945363, | |
| "grad_norm": 0.13414661586284637, | |
| "learning_rate": 1.0551671164463666e-05, | |
| "loss": 0.0026, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.063101192766449, | |
| "grad_norm": 0.010681160725653172, | |
| "learning_rate": 1.0223428469879736e-05, | |
| "loss": 0.0056, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "grad_norm": 0.03912714123725891, | |
| "learning_rate": 9.899790007311455e-06, | |
| "loss": 0.0019, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.078491727587534, | |
| "eval_loss": 0.040722113102674484, | |
| "eval_runtime": 204.3597, | |
| "eval_samples_per_second": 5.657, | |
| "eval_steps_per_second": 5.657, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.093882262408619, | |
| "grad_norm": 0.1255974918603897, | |
| "learning_rate": 9.580793238674573e-06, | |
| "loss": 0.0052, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.109272797229703, | |
| "grad_norm": 0.07881850004196167, | |
| "learning_rate": 9.266475088597781e-06, | |
| "loss": 0.0025, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.124663332050789, | |
| "grad_norm": 0.08200334012508392, | |
| "learning_rate": 8.956871940148681e-06, | |
| "loss": 0.0029, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.140053866871874, | |
| "grad_norm": 0.017860237509012222, | |
| "learning_rate": 8.65201963062225e-06, | |
| "loss": 0.0064, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "grad_norm": 0.04901472106575966, | |
| "learning_rate": 8.351953447392674e-06, | |
| "loss": 0.0074, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.155444401692959, | |
| "eval_loss": 0.043858833611011505, | |
| "eval_runtime": 204.6799, | |
| "eval_samples_per_second": 5.648, | |
| "eval_steps_per_second": 5.648, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.170834936514044, | |
| "grad_norm": 0.04998721927404404, | |
| "learning_rate": 8.056708123828738e-06, | |
| "loss": 0.004, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.186225471335129, | |
| "grad_norm": 0.11339768022298813, | |
| "learning_rate": 7.76631783527334e-06, | |
| "loss": 0.0079, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.201616006156214, | |
| "grad_norm": 0.04655611515045166, | |
| "learning_rate": 7.4808161950876145e-06, | |
| "loss": 0.0035, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.217006540977299, | |
| "grad_norm": 0.12219782173633575, | |
| "learning_rate": 7.200236250760111e-06, | |
| "loss": 0.004, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "grad_norm": 0.025195302441716194, | |
| "learning_rate": 6.924610480081478e-06, | |
| "loss": 0.0016, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.232397075798384, | |
| "eval_loss": 0.04438121244311333, | |
| "eval_runtime": 205.0733, | |
| "eval_samples_per_second": 5.637, | |
| "eval_steps_per_second": 5.637, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.247787610619469, | |
| "grad_norm": 0.03959754481911659, | |
| "learning_rate": 6.6539707873850774e-06, | |
| "loss": 0.0023, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.263178145440554, | |
| "grad_norm": 0.019419098272919655, | |
| "learning_rate": 6.3883484998539525e-06, | |
| "loss": 0.0035, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.278568680261639, | |
| "grad_norm": 0.12422063201665878, | |
| "learning_rate": 6.127774363894667e-06, | |
| "loss": 0.0024, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.293959215082724, | |
| "grad_norm": 0.1957329660654068, | |
| "learning_rate": 5.872278541578352e-06, | |
| "loss": 0.0037, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "grad_norm": 0.07204370200634003, | |
| "learning_rate": 5.6218906071493275e-06, | |
| "loss": 0.003, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.309349749903809, | |
| "eval_loss": 0.04541156440973282, | |
| "eval_runtime": 204.5478, | |
| "eval_samples_per_second": 5.651, | |
| "eval_steps_per_second": 5.651, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.324740284724895, | |
| "grad_norm": 0.05197502672672272, | |
| "learning_rate": 5.376639543601858e-06, | |
| "loss": 0.0031, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.340130819545979, | |
| "grad_norm": 0.025601493194699287, | |
| "learning_rate": 5.136553739325256e-06, | |
| "loss": 0.0017, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.355521354367064, | |
| "grad_norm": 0.04088883846998215, | |
| "learning_rate": 4.901660984817908e-06, | |
| "loss": 0.0022, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.370911889188149, | |
| "grad_norm": 0.02851008251309395, | |
| "learning_rate": 4.671988469470434e-06, | |
| "loss": 0.0021, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "grad_norm": 0.025700489059090614, | |
| "learning_rate": 4.447562778418435e-06, | |
| "loss": 0.0034, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.3863024240092345, | |
| "eval_loss": 0.046728041023015976, | |
| "eval_runtime": 204.4404, | |
| "eval_samples_per_second": 5.654, | |
| "eval_steps_per_second": 5.654, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.40169295883032, | |
| "grad_norm": 0.2153148651123047, | |
| "learning_rate": 4.228409889465218e-06, | |
| "loss": 0.0025, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.417083493651404, | |
| "grad_norm": 0.016644202172756195, | |
| "learning_rate": 4.014555170074824e-06, | |
| "loss": 0.001, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.432474028472489, | |
| "grad_norm": 0.05260796844959259, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0067, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.4478645632935745, | |
| "grad_norm": 0.030759744346141815, | |
| "learning_rate": 3.602838640595152e-06, | |
| "loss": 0.0024, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "grad_norm": 0.45937487483024597, | |
| "learning_rate": 3.4050244876656866e-06, | |
| "loss": 0.0039, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.46325509811466, | |
| "eval_loss": 0.047509122639894485, | |
| "eval_runtime": 205.0035, | |
| "eval_samples_per_second": 5.639, | |
| "eval_steps_per_second": 5.639, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.478645632935745, | |
| "grad_norm": 0.019434956833720207, | |
| "learning_rate": 3.2126038131022552e-06, | |
| "loss": 0.0071, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.494036167756829, | |
| "grad_norm": 0.1172695979475975, | |
| "learning_rate": 3.025598890051995e-06, | |
| "loss": 0.0041, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.5094267025779144, | |
| "grad_norm": 0.0296728927642107, | |
| "learning_rate": 2.8440313647760288e-06, | |
| "loss": 0.0043, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.524817237399, | |
| "grad_norm": 0.2890836298465729, | |
| "learning_rate": 2.6679222541438397e-06, | |
| "loss": 0.0053, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "grad_norm": 0.02802952751517296, | |
| "learning_rate": 2.4972919432005582e-06, | |
| "loss": 0.0019, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.540207772220085, | |
| "eval_loss": 0.04699544608592987, | |
| "eval_runtime": 204.8451, | |
| "eval_samples_per_second": 5.643, | |
| "eval_steps_per_second": 5.643, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.555598307041169, | |
| "grad_norm": 0.16024689376354218, | |
| "learning_rate": 2.3321601828073135e-06, | |
| "loss": 0.0012, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.570988841862254, | |
| "grad_norm": 0.06289687752723694, | |
| "learning_rate": 2.1725460873550317e-06, | |
| "loss": 0.0066, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.58637937668334, | |
| "grad_norm": 0.022529326379299164, | |
| "learning_rate": 2.0184681325519096e-06, | |
| "loss": 0.0022, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.601769911504425, | |
| "grad_norm": 0.3705087900161743, | |
| "learning_rate": 1.869944153284786e-06, | |
| "loss": 0.0024, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "grad_norm": 0.03192564845085144, | |
| "learning_rate": 1.7269913415547456e-06, | |
| "loss": 0.003, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.61716044632551, | |
| "eval_loss": 0.04718010500073433, | |
| "eval_runtime": 204.2121, | |
| "eval_samples_per_second": 5.661, | |
| "eval_steps_per_second": 5.661, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.632550981146595, | |
| "grad_norm": 0.13988825678825378, | |
| "learning_rate": 1.589626244487069e-06, | |
| "loss": 0.0027, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.64794151596768, | |
| "grad_norm": 0.30462774634361267, | |
| "learning_rate": 1.4578647624158726e-06, | |
| "loss": 0.0029, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.663332050788765, | |
| "grad_norm": 0.14595875144004822, | |
| "learning_rate": 1.3317221470436293e-06, | |
| "loss": 0.0042, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.67872258560985, | |
| "grad_norm": 0.14335300028324127, | |
| "learning_rate": 1.2112129996757315e-06, | |
| "loss": 0.0054, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "grad_norm": 0.008524621836841106, | |
| "learning_rate": 1.0963512695303546e-06, | |
| "loss": 0.0056, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.694113120430935, | |
| "eval_loss": 0.04739828780293465, | |
| "eval_runtime": 203.8435, | |
| "eval_samples_per_second": 5.671, | |
| "eval_steps_per_second": 5.671, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.7095036552520195, | |
| "grad_norm": 0.027586083859205246, | |
| "learning_rate": 9.871502521237975e-07, | |
| "loss": 0.0021, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.724894190073105, | |
| "grad_norm": 0.14117112755775452, | |
| "learning_rate": 8.83622587731514e-07, | |
| "loss": 0.0026, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.74028472489419, | |
| "grad_norm": 0.4257708191871643, | |
| "learning_rate": 7.857802599249564e-07, | |
| "loss": 0.0078, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.755675259715275, | |
| "grad_norm": 0.11569110304117203, | |
| "learning_rate": 6.936345941844335e-07, | |
| "loss": 0.002, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "grad_norm": 0.08163293451070786, | |
| "learning_rate": 6.071962565881939e-07, | |
| "loss": 0.0036, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.77106579453636, | |
| "eval_loss": 0.04735378548502922, | |
| "eval_runtime": 204.0317, | |
| "eval_samples_per_second": 5.666, | |
| "eval_steps_per_second": 5.666, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.786456329357446, | |
| "grad_norm": 0.07471448928117752, | |
| "learning_rate": 5.264752525777961e-07, | |
| "loss": 0.0072, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.80184686417853, | |
| "grad_norm": 0.2608599364757538, | |
| "learning_rate": 4.5148092579991863e-07, | |
| "loss": 0.0045, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.817237398999615, | |
| "grad_norm": 0.012474717572331429, | |
| "learning_rate": 3.822219570248531e-07, | |
| "loss": 0.0048, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.8326279338207, | |
| "grad_norm": 0.025006014853715897, | |
| "learning_rate": 3.187063631416742e-07, | |
| "loss": 0.0018, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "grad_norm": 0.051658909767866135, | |
| "learning_rate": 2.6094149623024365e-07, | |
| "loss": 0.0029, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.8480184686417855, | |
| "eval_loss": 0.047333285212516785, | |
| "eval_runtime": 204.3694, | |
| "eval_samples_per_second": 5.656, | |
| "eval_steps_per_second": 5.656, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.86340900346287, | |
| "grad_norm": 0.019093699753284454, | |
| "learning_rate": 2.0893404271022376e-07, | |
| "loss": 0.0028, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.878799538283955, | |
| "grad_norm": 0.13918374478816986, | |
| "learning_rate": 1.626900225670691e-07, | |
| "loss": 0.0025, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.89419007310504, | |
| "grad_norm": 0.05475996062159538, | |
| "learning_rate": 1.2221478865524493e-07, | |
| "loss": 0.0025, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.9095806079261255, | |
| "grad_norm": 0.013572592288255692, | |
| "learning_rate": 8.751302607857836e-08, | |
| "loss": 0.0045, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "grad_norm": 0.026636293157935143, | |
| "learning_rate": 5.8588751647975506e-08, | |
| "loss": 0.003, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.924971142747211, | |
| "eval_loss": 0.04742961749434471, | |
| "eval_runtime": 204.1387, | |
| "eval_samples_per_second": 5.663, | |
| "eval_steps_per_second": 5.663, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.940361677568296, | |
| "grad_norm": 0.0202015433460474, | |
| "learning_rate": 3.544531341646007e-08, | |
| "loss": 0.0047, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.95575221238938, | |
| "grad_norm": 0.03867825120687485, | |
| "learning_rate": 1.808539029161116e-08, | |
| "loss": 0.0017, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.9711427472104655, | |
| "grad_norm": 0.0632685050368309, | |
| "learning_rate": 6.510991725500182e-09, | |
| "loss": 0.0053, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.986533282031551, | |
| "grad_norm": 0.035415440797805786, | |
| "learning_rate": 7.234574820769169e-10, | |
| "loss": 0.0015, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.994228549442093, | |
| "step": 3245, | |
| "total_flos": 8.169046237996646e+17, | |
| "train_loss": 0.03526924680516652, | |
| "train_runtime": 44618.8073, | |
| "train_samples_per_second": 1.165, | |
| "train_steps_per_second": 0.073 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3245, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.169046237996646e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |