| { | |
| "best_metric": 0.7993046045303345, | |
| "best_model_checkpoint": "./colab20240326ryan/checkpoint-2100", | |
| "epoch": 1.160092807424594, | |
| "eval_steps": 100, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 1.9214508533477783, | |
| "learning_rate": 0.0001997679814385151, | |
| "loss": 1.7231, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.908971905708313, | |
| "learning_rate": 0.00019953596287703018, | |
| "loss": 1.5448, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 1.5993173122406006, | |
| "learning_rate": 0.00019930394431554523, | |
| "loss": 1.3519, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.219737529754639, | |
| "learning_rate": 0.00019907192575406032, | |
| "loss": 1.4004, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.771422863006592, | |
| "learning_rate": 0.00019883990719257543, | |
| "loss": 1.2249, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.5087692737579346, | |
| "learning_rate": 0.0001986078886310905, | |
| "loss": 1.3035, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.224257707595825, | |
| "learning_rate": 0.0001983758700696056, | |
| "loss": 1.2159, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.9843618869781494, | |
| "learning_rate": 0.00019814385150812065, | |
| "loss": 1.0327, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.905066967010498, | |
| "learning_rate": 0.00019791183294663573, | |
| "loss": 1.0393, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.216238021850586, | |
| "learning_rate": 0.00019767981438515082, | |
| "loss": 1.1654, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_accuracy": 0.5813018346318171, | |
| "eval_loss": 1.081552505493164, | |
| "eval_runtime": 142.0814, | |
| "eval_samples_per_second": 28.005, | |
| "eval_steps_per_second": 3.505, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.5341720581054688, | |
| "learning_rate": 0.0001974477958236659, | |
| "loss": 1.162, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 3.2532262802124023, | |
| "learning_rate": 0.00019721577726218098, | |
| "loss": 1.04, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.41017746925354, | |
| "learning_rate": 0.00019698375870069607, | |
| "loss": 1.0566, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.7879230976104736, | |
| "learning_rate": 0.00019675174013921115, | |
| "loss": 1.0105, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.4111428260803223, | |
| "learning_rate": 0.00019651972157772623, | |
| "loss": 1.0504, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 3.4261136054992676, | |
| "learning_rate": 0.00019628770301624132, | |
| "loss": 1.0128, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.207861423492432, | |
| "learning_rate": 0.0001960556844547564, | |
| "loss": 0.9814, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.604964256286621, | |
| "learning_rate": 0.00019582366589327148, | |
| "loss": 1.17, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.710505485534668, | |
| "learning_rate": 0.00019559164733178654, | |
| "loss": 1.0255, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.121289253234863, | |
| "learning_rate": 0.00019535962877030162, | |
| "loss": 1.1321, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_accuracy": 0.5996481528022116, | |
| "eval_loss": 0.9905579686164856, | |
| "eval_runtime": 135.9688, | |
| "eval_samples_per_second": 29.264, | |
| "eval_steps_per_second": 3.663, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.4283132553100586, | |
| "learning_rate": 0.0001951276102088167, | |
| "loss": 1.0087, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.141366958618164, | |
| "learning_rate": 0.0001948955916473318, | |
| "loss": 0.8722, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.6067655086517334, | |
| "learning_rate": 0.00019466357308584687, | |
| "loss": 1.0173, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.5523955821990967, | |
| "learning_rate": 0.00019445475638051046, | |
| "loss": 0.9303, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.983736753463745, | |
| "learning_rate": 0.00019422273781902555, | |
| "loss": 0.9035, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 3.1925017833709717, | |
| "learning_rate": 0.00019399071925754063, | |
| "loss": 0.9329, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.603178977966309, | |
| "learning_rate": 0.00019375870069605569, | |
| "loss": 0.9351, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.129456043243408, | |
| "learning_rate": 0.00019352668213457077, | |
| "loss": 1.0367, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.650508403778076, | |
| "learning_rate": 0.00019329466357308585, | |
| "loss": 0.9677, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.0717406272888184, | |
| "learning_rate": 0.00019306264501160094, | |
| "loss": 0.9389, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "eval_accuracy": 0.625031414928374, | |
| "eval_loss": 0.9222464561462402, | |
| "eval_runtime": 133.436, | |
| "eval_samples_per_second": 29.82, | |
| "eval_steps_per_second": 3.732, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 3.079808473587036, | |
| "learning_rate": 0.00019283062645011602, | |
| "loss": 0.9204, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.8033320903778076, | |
| "learning_rate": 0.0001925986078886311, | |
| "loss": 0.9505, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.029008150100708, | |
| "learning_rate": 0.0001923665893271462, | |
| "loss": 0.9715, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.543546199798584, | |
| "learning_rate": 0.00019213457076566127, | |
| "loss": 0.9928, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.7682580947875977, | |
| "learning_rate": 0.00019190255220417635, | |
| "loss": 1.0367, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.006638526916504, | |
| "learning_rate": 0.00019167053364269144, | |
| "loss": 0.932, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 2.4894397258758545, | |
| "learning_rate": 0.00019143851508120652, | |
| "loss": 0.8321, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 3.1492834091186523, | |
| "learning_rate": 0.00019120649651972158, | |
| "loss": 1.0013, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 4.911842346191406, | |
| "learning_rate": 0.00019097447795823666, | |
| "loss": 0.987, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.6950294971466064, | |
| "learning_rate": 0.00019074245939675174, | |
| "loss": 0.816, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_accuracy": 0.5740135712490575, | |
| "eval_loss": 1.0586621761322021, | |
| "eval_runtime": 131.2597, | |
| "eval_samples_per_second": 30.314, | |
| "eval_steps_per_second": 3.794, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.0810515880584717, | |
| "learning_rate": 0.00019051044083526683, | |
| "loss": 1.0748, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 3.0597758293151855, | |
| "learning_rate": 0.0001902784222737819, | |
| "loss": 0.851, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.7524383068084717, | |
| "learning_rate": 0.000190046403712297, | |
| "loss": 0.9123, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.255000591278076, | |
| "learning_rate": 0.00018981438515081208, | |
| "loss": 1.0082, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.834663152694702, | |
| "learning_rate": 0.00018958236658932716, | |
| "loss": 1.0721, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 2.59566593170166, | |
| "learning_rate": 0.00018935034802784224, | |
| "loss": 0.9448, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.0671868324279785, | |
| "learning_rate": 0.00018911832946635733, | |
| "loss": 0.823, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.022857189178467, | |
| "learning_rate": 0.00018888631090487238, | |
| "loss": 0.9282, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1.6197035312652588, | |
| "learning_rate": 0.00018865429234338747, | |
| "loss": 0.8492, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.2592787742614746, | |
| "learning_rate": 0.00018842227378190255, | |
| "loss": 0.7273, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "eval_accuracy": 0.6267906509173159, | |
| "eval_loss": 0.918483555316925, | |
| "eval_runtime": 131.2522, | |
| "eval_samples_per_second": 30.316, | |
| "eval_steps_per_second": 3.794, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.200505256652832, | |
| "learning_rate": 0.00018819025522041763, | |
| "loss": 0.9205, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.9970719814300537, | |
| "learning_rate": 0.00018795823665893272, | |
| "loss": 0.8762, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.8891489505767822, | |
| "learning_rate": 0.00018772621809744783, | |
| "loss": 0.9051, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.4764907360076904, | |
| "learning_rate": 0.00018749419953596288, | |
| "loss": 0.9494, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.9991800785064697, | |
| "learning_rate": 0.00018726218097447797, | |
| "loss": 0.9639, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.954806327819824, | |
| "learning_rate": 0.00018703016241299305, | |
| "loss": 0.9069, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 3.1720399856567383, | |
| "learning_rate": 0.00018679814385150813, | |
| "loss": 0.84, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 1.5639662742614746, | |
| "learning_rate": 0.00018656612529002322, | |
| "loss": 0.8589, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 3.4077460765838623, | |
| "learning_rate": 0.00018633410672853827, | |
| "loss": 0.8529, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 3.089357852935791, | |
| "learning_rate": 0.00018610208816705336, | |
| "loss": 0.8282, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_accuracy": 0.6293038451872329, | |
| "eval_loss": 0.9175940155982971, | |
| "eval_runtime": 132.213, | |
| "eval_samples_per_second": 30.095, | |
| "eval_steps_per_second": 3.767, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.7197611331939697, | |
| "learning_rate": 0.00018587006960556844, | |
| "loss": 0.8377, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.838669538497925, | |
| "learning_rate": 0.00018563805104408355, | |
| "loss": 0.88, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 4.2069902420043945, | |
| "learning_rate": 0.00018540603248259864, | |
| "loss": 0.8175, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.0792770385742188, | |
| "learning_rate": 0.0001851740139211137, | |
| "loss": 0.8921, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.4577174186706543, | |
| "learning_rate": 0.00018494199535962877, | |
| "loss": 0.8612, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 3.424455165863037, | |
| "learning_rate": 0.00018470997679814386, | |
| "loss": 0.9775, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2.300741672515869, | |
| "learning_rate": 0.00018447795823665894, | |
| "loss": 1.0909, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.8668731451034546, | |
| "learning_rate": 0.00018424593967517403, | |
| "loss": 0.8205, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.170844793319702, | |
| "learning_rate": 0.0001840139211136891, | |
| "loss": 0.8701, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.8682425022125244, | |
| "learning_rate": 0.00018378190255220417, | |
| "loss": 0.8, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.6272932897712993, | |
| "eval_loss": 0.9006840586662292, | |
| "eval_runtime": 129.3938, | |
| "eval_samples_per_second": 30.751, | |
| "eval_steps_per_second": 3.849, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 3.769883871078491, | |
| "learning_rate": 0.00018354988399071928, | |
| "loss": 0.9765, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 3.9122543334960938, | |
| "learning_rate": 0.00018331786542923436, | |
| "loss": 0.9744, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 3.644559860229492, | |
| "learning_rate": 0.00018308584686774944, | |
| "loss": 0.8877, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 3.476562976837158, | |
| "learning_rate": 0.00018285382830626453, | |
| "loss": 0.867, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 3.0982306003570557, | |
| "learning_rate": 0.00018262180974477958, | |
| "loss": 0.9965, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.395843505859375, | |
| "learning_rate": 0.00018238979118329467, | |
| "loss": 0.8635, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 3.6630685329437256, | |
| "learning_rate": 0.00018215777262180975, | |
| "loss": 0.8562, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.064682960510254, | |
| "learning_rate": 0.00018192575406032483, | |
| "loss": 0.7875, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 3.3986759185791016, | |
| "learning_rate": 0.00018169373549883992, | |
| "loss": 0.8583, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 3.0697574615478516, | |
| "learning_rate": 0.000181461716937355, | |
| "loss": 0.8777, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.6202563458155316, | |
| "eval_loss": 0.9337747693061829, | |
| "eval_runtime": 131.2508, | |
| "eval_samples_per_second": 30.316, | |
| "eval_steps_per_second": 3.794, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.8958779573440552, | |
| "learning_rate": 0.00018122969837587008, | |
| "loss": 0.9388, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.3388631343841553, | |
| "learning_rate": 0.00018099767981438517, | |
| "loss": 0.8248, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 4.1161298751831055, | |
| "learning_rate": 0.00018076566125290025, | |
| "loss": 0.8008, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.5049020051956177, | |
| "learning_rate": 0.00018053364269141533, | |
| "loss": 0.8932, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 2.5027551651000977, | |
| "learning_rate": 0.00018030162412993042, | |
| "loss": 0.7656, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.0704867839813232, | |
| "learning_rate": 0.00018006960556844547, | |
| "loss": 0.8206, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.6864800453186035, | |
| "learning_rate": 0.00017983758700696056, | |
| "loss": 0.8875, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 3.732292890548706, | |
| "learning_rate": 0.00017960556844547564, | |
| "loss": 0.9411, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 3.7439608573913574, | |
| "learning_rate": 0.00017937354988399072, | |
| "loss": 0.904, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.159684181213379, | |
| "learning_rate": 0.0001791415313225058, | |
| "loss": 0.7142, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_accuracy": 0.614727318421714, | |
| "eval_loss": 0.9442586302757263, | |
| "eval_runtime": 130.2393, | |
| "eval_samples_per_second": 30.551, | |
| "eval_steps_per_second": 3.824, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.202846050262451, | |
| "learning_rate": 0.0001789095127610209, | |
| "loss": 0.8798, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.4513931274414062, | |
| "learning_rate": 0.00017867749419953597, | |
| "loss": 0.8558, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.3939168453216553, | |
| "learning_rate": 0.00017844547563805106, | |
| "loss": 0.917, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 2.8509578704833984, | |
| "learning_rate": 0.00017821345707656614, | |
| "loss": 0.8373, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.6446682214736938, | |
| "learning_rate": 0.00017798143851508122, | |
| "loss": 0.7641, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.431823968887329, | |
| "learning_rate": 0.00017774941995359628, | |
| "loss": 0.8709, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.1793901920318604, | |
| "learning_rate": 0.00017751740139211136, | |
| "loss": 0.7838, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.03481125831604, | |
| "learning_rate": 0.00017728538283062645, | |
| "loss": 0.844, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.9181933403015137, | |
| "learning_rate": 0.00017705336426914153, | |
| "loss": 0.9402, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.2899041175842285, | |
| "learning_rate": 0.00017682134570765661, | |
| "loss": 0.8452, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_accuracy": 0.6282985674792662, | |
| "eval_loss": 0.8846696615219116, | |
| "eval_runtime": 130.7794, | |
| "eval_samples_per_second": 30.425, | |
| "eval_steps_per_second": 3.808, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 3.7156896591186523, | |
| "learning_rate": 0.00017658932714617172, | |
| "loss": 0.8338, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.9189355373382568, | |
| "learning_rate": 0.00017635730858468678, | |
| "loss": 0.9293, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.5769336223602295, | |
| "learning_rate": 0.00017612529002320186, | |
| "loss": 0.9528, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.103059768676758, | |
| "learning_rate": 0.00017589327146171695, | |
| "loss": 0.8122, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 1.972256064414978, | |
| "learning_rate": 0.00017566125290023203, | |
| "loss": 0.9213, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 2.265113592147827, | |
| "learning_rate": 0.00017542923433874711, | |
| "loss": 0.9075, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.6354522705078125, | |
| "learning_rate": 0.00017519721577726217, | |
| "loss": 0.8548, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.182709217071533, | |
| "learning_rate": 0.00017496519721577725, | |
| "loss": 0.7877, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.5550811290740967, | |
| "learning_rate": 0.00017473317865429236, | |
| "loss": 0.9916, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.5702245235443115, | |
| "learning_rate": 0.00017450116009280745, | |
| "loss": 0.845, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.6622266901231465, | |
| "eval_loss": 0.8412047624588013, | |
| "eval_runtime": 129.4336, | |
| "eval_samples_per_second": 30.742, | |
| "eval_steps_per_second": 3.848, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 3.416830539703369, | |
| "learning_rate": 0.00017426914153132253, | |
| "loss": 0.6856, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.638490915298462, | |
| "learning_rate": 0.0001740371229698376, | |
| "loss": 0.7501, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 4.172976016998291, | |
| "learning_rate": 0.00017380510440835267, | |
| "loss": 0.8201, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 2.498607873916626, | |
| "learning_rate": 0.00017357308584686775, | |
| "loss": 0.7571, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 5.480504035949707, | |
| "learning_rate": 0.00017334106728538284, | |
| "loss": 0.7706, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.2535948753356934, | |
| "learning_rate": 0.00017310904872389792, | |
| "loss": 0.8646, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 4.1205878257751465, | |
| "learning_rate": 0.000172877030162413, | |
| "loss": 0.9275, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 3.1862285137176514, | |
| "learning_rate": 0.0001726450116009281, | |
| "loss": 0.7335, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 2.7202231884002686, | |
| "learning_rate": 0.00017241299303944317, | |
| "loss": 0.7428, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.3965518474578857, | |
| "learning_rate": 0.00017218097447795826, | |
| "loss": 0.9167, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_accuracy": 0.6526765518974617, | |
| "eval_loss": 0.87410569190979, | |
| "eval_runtime": 130.959, | |
| "eval_samples_per_second": 30.384, | |
| "eval_steps_per_second": 3.803, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.0782244205474854, | |
| "learning_rate": 0.00017194895591647334, | |
| "loss": 0.8603, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 2.4333736896514893, | |
| "learning_rate": 0.00017171693735498842, | |
| "loss": 0.779, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 3.9308993816375732, | |
| "learning_rate": 0.00017148491879350348, | |
| "loss": 0.7695, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.4168589115142822, | |
| "learning_rate": 0.00017125290023201856, | |
| "loss": 0.8655, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.680983304977417, | |
| "learning_rate": 0.00017102088167053365, | |
| "loss": 0.7862, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 3.8315815925598145, | |
| "learning_rate": 0.00017078886310904873, | |
| "loss": 0.7972, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.910196304321289, | |
| "learning_rate": 0.0001705568445475638, | |
| "loss": 0.7454, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.9004689455032349, | |
| "learning_rate": 0.0001703248259860789, | |
| "loss": 0.7709, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.2291324138641357, | |
| "learning_rate": 0.00017009280742459398, | |
| "loss": 0.7085, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.6493847370147705, | |
| "learning_rate": 0.00016986078886310906, | |
| "loss": 0.8226, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_accuracy": 0.6659964815280222, | |
| "eval_loss": 0.8283097743988037, | |
| "eval_runtime": 130.9921, | |
| "eval_samples_per_second": 30.376, | |
| "eval_steps_per_second": 3.802, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 4.123025417327881, | |
| "learning_rate": 0.00016962877030162415, | |
| "loss": 0.9109, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.537853479385376, | |
| "learning_rate": 0.00016939675174013923, | |
| "loss": 0.9124, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.515120506286621, | |
| "learning_rate": 0.0001691647331786543, | |
| "loss": 0.8357, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.7295467853546143, | |
| "learning_rate": 0.00016893271461716937, | |
| "loss": 0.7425, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.3161306381225586, | |
| "learning_rate": 0.00016870069605568445, | |
| "loss": 0.7518, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.593114137649536, | |
| "learning_rate": 0.00016846867749419954, | |
| "loss": 0.816, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.4234368801116943, | |
| "learning_rate": 0.00016823665893271462, | |
| "loss": 0.8256, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.0647542476654053, | |
| "learning_rate": 0.0001680046403712297, | |
| "loss": 0.9176, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.5590307712554932, | |
| "learning_rate": 0.00016777262180974479, | |
| "loss": 0.7476, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 2.5730812549591064, | |
| "learning_rate": 0.00016754060324825987, | |
| "loss": 0.7738, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_accuracy": 0.6401105805478764, | |
| "eval_loss": 0.8641374111175537, | |
| "eval_runtime": 131.4185, | |
| "eval_samples_per_second": 30.277, | |
| "eval_steps_per_second": 3.789, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.080822467803955, | |
| "learning_rate": 0.00016730858468677495, | |
| "loss": 0.8131, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.1145131587982178, | |
| "learning_rate": 0.00016707656612529004, | |
| "loss": 0.8048, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.4306788444519043, | |
| "learning_rate": 0.00016684454756380512, | |
| "loss": 0.7241, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 3.0480475425720215, | |
| "learning_rate": 0.00016661252900232018, | |
| "loss": 0.6803, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 2.5454821586608887, | |
| "learning_rate": 0.00016638051044083526, | |
| "loss": 0.773, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 2.2483272552490234, | |
| "learning_rate": 0.00016614849187935034, | |
| "loss": 0.8333, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.9373365640640259, | |
| "learning_rate": 0.00016591647331786543, | |
| "loss": 0.7289, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.8379623889923096, | |
| "learning_rate": 0.00016568445475638054, | |
| "loss": 0.7733, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.349510431289673, | |
| "learning_rate": 0.00016545243619489562, | |
| "loss": 0.8449, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.029337406158447, | |
| "learning_rate": 0.00016522041763341068, | |
| "loss": 0.8427, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_accuracy": 0.6725307866298065, | |
| "eval_loss": 0.803027331829071, | |
| "eval_runtime": 131.7174, | |
| "eval_samples_per_second": 30.209, | |
| "eval_steps_per_second": 3.781, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.8301985263824463, | |
| "learning_rate": 0.00016498839907192576, | |
| "loss": 0.7437, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2.7581472396850586, | |
| "learning_rate": 0.00016475638051044084, | |
| "loss": 0.7888, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2.044255256652832, | |
| "learning_rate": 0.00016452436194895593, | |
| "loss": 0.7, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.427280426025391, | |
| "learning_rate": 0.000164292343387471, | |
| "loss": 0.8854, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.044015884399414, | |
| "learning_rate": 0.00016406032482598607, | |
| "loss": 0.677, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.954887628555298, | |
| "learning_rate": 0.00016382830626450115, | |
| "loss": 0.7198, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.2452878952026367, | |
| "learning_rate": 0.00016359628770301626, | |
| "loss": 0.7495, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.0875964164733887, | |
| "learning_rate": 0.00016336426914153134, | |
| "loss": 0.8106, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.8363144397735596, | |
| "learning_rate": 0.00016313225058004643, | |
| "loss": 0.6737, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 3.34063982963562, | |
| "learning_rate": 0.00016290023201856148, | |
| "loss": 0.6783, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.6564463433023373, | |
| "eval_loss": 0.8367487192153931, | |
| "eval_runtime": 129.8582, | |
| "eval_samples_per_second": 30.641, | |
| "eval_steps_per_second": 3.835, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.197628974914551, | |
| "learning_rate": 0.00016266821345707657, | |
| "loss": 0.7794, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.9976580142974854, | |
| "learning_rate": 0.00016243619489559165, | |
| "loss": 0.832, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.8508596420288086, | |
| "learning_rate": 0.00016220417633410673, | |
| "loss": 0.86, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.7021024227142334, | |
| "learning_rate": 0.00016197215777262182, | |
| "loss": 0.7531, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 2.3222107887268066, | |
| "learning_rate": 0.0001617401392111369, | |
| "loss": 0.7338, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 2.1219635009765625, | |
| "learning_rate": 0.00016150812064965198, | |
| "loss": 0.8965, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 11.041630744934082, | |
| "learning_rate": 0.00016127610208816707, | |
| "loss": 0.7939, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.4006307125091553, | |
| "learning_rate": 0.00016104408352668215, | |
| "loss": 0.736, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 2.535405158996582, | |
| "learning_rate": 0.00016081206496519723, | |
| "loss": 0.7982, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 2.8077518939971924, | |
| "learning_rate": 0.00016058004640371232, | |
| "loss": 0.7856, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_accuracy": 0.6051771801960292, | |
| "eval_loss": 0.9696215391159058, | |
| "eval_runtime": 130.9284, | |
| "eval_samples_per_second": 30.391, | |
| "eval_steps_per_second": 3.804, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 2.17937970161438, | |
| "learning_rate": 0.00016034802784222737, | |
| "loss": 0.8074, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.2899444103240967, | |
| "learning_rate": 0.00016011600928074246, | |
| "loss": 0.8416, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.247441530227661, | |
| "learning_rate": 0.00015988399071925754, | |
| "loss": 0.8302, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 2.508978843688965, | |
| "learning_rate": 0.00015965197215777262, | |
| "loss": 0.7192, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.634054183959961, | |
| "learning_rate": 0.0001594199535962877, | |
| "loss": 0.7919, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 2.7715981006622314, | |
| "learning_rate": 0.0001591879350348028, | |
| "loss": 0.703, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.9510867595672607, | |
| "learning_rate": 0.00015895591647331787, | |
| "loss": 0.7246, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.5826807022094727, | |
| "learning_rate": 0.00015872389791183296, | |
| "loss": 0.8291, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.8682587146759033, | |
| "learning_rate": 0.00015849187935034804, | |
| "loss": 0.7284, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.7725648880004883, | |
| "learning_rate": 0.00015825986078886313, | |
| "loss": 0.7356, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_accuracy": 0.6516712741894949, | |
| "eval_loss": 0.857125461101532, | |
| "eval_runtime": 130.6056, | |
| "eval_samples_per_second": 30.466, | |
| "eval_steps_per_second": 3.813, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.756324291229248, | |
| "learning_rate": 0.0001580278422273782, | |
| "loss": 0.8463, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.613060712814331, | |
| "learning_rate": 0.00015779582366589326, | |
| "loss": 0.7517, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.3475732803344727, | |
| "learning_rate": 0.00015756380510440835, | |
| "loss": 0.835, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.118978977203369, | |
| "learning_rate": 0.00015733178654292343, | |
| "loss": 0.8143, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.3323171138763428, | |
| "learning_rate": 0.00015709976798143852, | |
| "loss": 0.8139, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 2.580026865005493, | |
| "learning_rate": 0.00015686774941995363, | |
| "loss": 0.8665, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 2.8367908000946045, | |
| "learning_rate": 0.00015663573085846868, | |
| "loss": 0.7187, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 3.431257724761963, | |
| "learning_rate": 0.00015640371229698377, | |
| "loss": 0.8616, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 3.8367366790771484, | |
| "learning_rate": 0.00015617169373549885, | |
| "loss": 0.7892, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.648777723312378, | |
| "learning_rate": 0.00015593967517401393, | |
| "loss": 0.9186, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_accuracy": 0.6675043980899723, | |
| "eval_loss": 0.8260459899902344, | |
| "eval_runtime": 131.0193, | |
| "eval_samples_per_second": 30.37, | |
| "eval_steps_per_second": 3.801, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.5578160285949707, | |
| "learning_rate": 0.00015570765661252902, | |
| "loss": 0.6849, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 2.5033838748931885, | |
| "learning_rate": 0.00015547563805104407, | |
| "loss": 0.6708, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.074505090713501, | |
| "learning_rate": 0.00015524361948955916, | |
| "loss": 0.717, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.335425853729248, | |
| "learning_rate": 0.00015501160092807424, | |
| "loss": 0.9028, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.3634660243988037, | |
| "learning_rate": 0.00015477958236658935, | |
| "loss": 0.6975, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.022599697113037, | |
| "learning_rate": 0.00015454756380510443, | |
| "loss": 0.6303, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.197246551513672, | |
| "learning_rate": 0.0001543155452436195, | |
| "loss": 0.7147, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.748758554458618, | |
| "learning_rate": 0.00015408352668213457, | |
| "loss": 0.7944, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.029123067855835, | |
| "learning_rate": 0.00015385150812064966, | |
| "loss": 0.7363, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.3515162467956543, | |
| "learning_rate": 0.00015361948955916474, | |
| "loss": 0.8218, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_accuracy": 0.654938426740387, | |
| "eval_loss": 0.8351722359657288, | |
| "eval_runtime": 129.1873, | |
| "eval_samples_per_second": 30.8, | |
| "eval_steps_per_second": 3.855, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 3.3627982139587402, | |
| "learning_rate": 0.00015338747099767982, | |
| "loss": 0.7137, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.0946731567382812, | |
| "learning_rate": 0.0001531554524361949, | |
| "loss": 0.8327, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.9171329736709595, | |
| "learning_rate": 0.00015292343387470996, | |
| "loss": 0.7805, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.749093532562256, | |
| "learning_rate": 0.00015269141531322507, | |
| "loss": 0.8531, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.960636615753174, | |
| "learning_rate": 0.00015245939675174016, | |
| "loss": 0.7838, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.5994982719421387, | |
| "learning_rate": 0.00015222737819025524, | |
| "loss": 0.6932, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.6657791137695312, | |
| "learning_rate": 0.00015199535962877032, | |
| "loss": 0.731, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.7149091958999634, | |
| "learning_rate": 0.00015176334106728538, | |
| "loss": 0.8234, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.5878645181655884, | |
| "learning_rate": 0.00015153132250580046, | |
| "loss": 0.7098, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.8897655010223389, | |
| "learning_rate": 0.00015129930394431555, | |
| "loss": 0.6245, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_accuracy": 0.6763005780346821, | |
| "eval_loss": 0.7993046045303345, | |
| "eval_runtime": 133.106, | |
| "eval_samples_per_second": 29.893, | |
| "eval_steps_per_second": 3.741, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 3.2624616622924805, | |
| "learning_rate": 0.00015106728538283063, | |
| "loss": 0.8209, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 2.469926595687866, | |
| "learning_rate": 0.00015083526682134571, | |
| "loss": 0.7127, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.8582072257995605, | |
| "learning_rate": 0.0001506032482598608, | |
| "loss": 0.655, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.1348557472229004, | |
| "learning_rate": 0.00015037122969837588, | |
| "loss": 0.8291, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.1626625061035156, | |
| "learning_rate": 0.00015013921113689096, | |
| "loss": 0.8373, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.9470633268356323, | |
| "learning_rate": 0.00014990719257540605, | |
| "loss": 0.6225, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 2.336871862411499, | |
| "learning_rate": 0.00014967517401392113, | |
| "loss": 0.5857, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.737004280090332, | |
| "learning_rate": 0.00014944315545243621, | |
| "loss": 0.5935, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 2.336336612701416, | |
| "learning_rate": 0.00014921113689095127, | |
| "loss": 0.5824, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 2.338193655014038, | |
| "learning_rate": 0.00014897911832946635, | |
| "loss": 0.4945, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.6589595375722543, | |
| "eval_loss": 0.8315911889076233, | |
| "eval_runtime": 132.4435, | |
| "eval_samples_per_second": 30.043, | |
| "eval_steps_per_second": 3.76, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.667480707168579, | |
| "learning_rate": 0.00014874709976798144, | |
| "loss": 0.6186, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.019312858581543, | |
| "learning_rate": 0.00014851508120649652, | |
| "loss": 0.5037, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.4240450859069824, | |
| "learning_rate": 0.0001482830626450116, | |
| "loss": 0.5781, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.2333035469055176, | |
| "learning_rate": 0.0001480510440835267, | |
| "loss": 0.4923, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.023408889770508, | |
| "learning_rate": 0.00014781902552204177, | |
| "loss": 0.5855, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.4406158924102783, | |
| "learning_rate": 0.00014758700696055685, | |
| "loss": 0.5579, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 3.5192463397979736, | |
| "learning_rate": 0.00014735498839907194, | |
| "loss": 0.6066, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 4.174234390258789, | |
| "learning_rate": 0.00014712296983758702, | |
| "loss": 0.6238, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 2.916022539138794, | |
| "learning_rate": 0.00014689095127610208, | |
| "loss": 0.5475, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 3.1607933044433594, | |
| "learning_rate": 0.00014665893271461716, | |
| "loss": 0.6064, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.6680070369439558, | |
| "eval_loss": 0.8378371596336365, | |
| "eval_runtime": 132.7481, | |
| "eval_samples_per_second": 29.974, | |
| "eval_steps_per_second": 3.751, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.8557933568954468, | |
| "learning_rate": 0.00014642691415313224, | |
| "loss": 0.6037, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 4.142065048217773, | |
| "learning_rate": 0.00014619489559164733, | |
| "loss": 0.6552, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 3.622699499130249, | |
| "learning_rate": 0.00014596287703016244, | |
| "loss": 0.4755, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 3.9584805965423584, | |
| "learning_rate": 0.00014573085846867752, | |
| "loss": 0.677, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 2.498189926147461, | |
| "learning_rate": 0.00014549883990719258, | |
| "loss": 0.5519, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 5.097834587097168, | |
| "learning_rate": 0.00014526682134570766, | |
| "loss": 0.6068, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.7660293579101562, | |
| "learning_rate": 0.00014503480278422275, | |
| "loss": 0.5356, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.423243999481201, | |
| "learning_rate": 0.00014480278422273783, | |
| "loss": 0.6954, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 3.099900484085083, | |
| "learning_rate": 0.0001445707656612529, | |
| "loss": 0.4953, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 4.438780784606934, | |
| "learning_rate": 0.00014433874709976797, | |
| "loss": 0.638, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.6828348831364665, | |
| "eval_loss": 0.8223534822463989, | |
| "eval_runtime": 130.5252, | |
| "eval_samples_per_second": 30.485, | |
| "eval_steps_per_second": 3.815, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.960069417953491, | |
| "learning_rate": 0.00014410672853828305, | |
| "loss": 0.5635, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.4343048334121704, | |
| "learning_rate": 0.00014387470997679816, | |
| "loss": 0.5937, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.8916206359863281, | |
| "learning_rate": 0.00014364269141531325, | |
| "loss": 0.5543, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 2.315703868865967, | |
| "learning_rate": 0.00014341067285382833, | |
| "loss": 0.4934, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 2.3009326457977295, | |
| "learning_rate": 0.00014317865429234339, | |
| "loss": 0.5487, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 2.4745514392852783, | |
| "learning_rate": 0.00014294663573085847, | |
| "loss": 0.6988, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.8443727493286133, | |
| "learning_rate": 0.00014271461716937355, | |
| "loss": 0.5079, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 3.124251127243042, | |
| "learning_rate": 0.00014248259860788864, | |
| "loss": 0.5499, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 3.0896270275115967, | |
| "learning_rate": 0.00014225058004640372, | |
| "loss": 0.6026, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 2.8856897354125977, | |
| "learning_rate": 0.0001420185614849188, | |
| "loss": 0.6253, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "eval_accuracy": 0.6617240512691631, | |
| "eval_loss": 0.8880072236061096, | |
| "eval_runtime": 129.7205, | |
| "eval_samples_per_second": 30.674, | |
| "eval_steps_per_second": 3.839, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "step": 2500, | |
| "total_flos": 3.099325741767844e+18, | |
| "train_loss": 0.8266718128204346, | |
| "train_runtime": 5452.5845, | |
| "train_samples_per_second": 25.29, | |
| "train_steps_per_second": 1.581 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 8620, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "total_flos": 3.099325741767844e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |