| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002, | |
| "eval_loss": 2.1960251331329346, | |
| "eval_runtime": 2.1983, | |
| "eval_samples_per_second": 70.964, | |
| "eval_steps_per_second": 3.639, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 9.552270889282227, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.3747, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_loss": 0.7324444651603699, | |
| "eval_runtime": 1.9429, | |
| "eval_samples_per_second": 80.293, | |
| "eval_steps_per_second": 4.118, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 12.921321868896484, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6642, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 0.7477038502693176, | |
| "eval_runtime": 1.9488, | |
| "eval_samples_per_second": 80.05, | |
| "eval_steps_per_second": 4.105, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 6.498777389526367, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.6259, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "eval_loss": 0.6763167977333069, | |
| "eval_runtime": 1.9355, | |
| "eval_samples_per_second": 80.601, | |
| "eval_steps_per_second": 4.133, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 13.172846794128418, | |
| "learning_rate": 1e-05, | |
| "loss": 0.665, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.679334819316864, | |
| "eval_runtime": 1.9711, | |
| "eval_samples_per_second": 79.144, | |
| "eval_steps_per_second": 4.059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 5.069002628326416, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.6586, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.6933216452598572, | |
| "eval_runtime": 1.9422, | |
| "eval_samples_per_second": 80.321, | |
| "eval_steps_per_second": 4.119, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 4.540435791015625, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.6939, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_loss": 0.7091314792633057, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 78.787, | |
| "eval_steps_per_second": 4.04, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.8698132038116455, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 0.6783, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_loss": 0.7234737873077393, | |
| "eval_runtime": 1.9392, | |
| "eval_samples_per_second": 80.445, | |
| "eval_steps_per_second": 4.125, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.8550472259521484, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7386, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.7281543016433716, | |
| "eval_runtime": 1.9771, | |
| "eval_samples_per_second": 78.905, | |
| "eval_steps_per_second": 4.046, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.354882001876831, | |
| "learning_rate": 1.999048221581858e-05, | |
| "loss": 0.7279, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "eval_loss": 0.744288444519043, | |
| "eval_runtime": 1.9696, | |
| "eval_samples_per_second": 79.203, | |
| "eval_steps_per_second": 4.062, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.7983996868133545, | |
| "learning_rate": 1.9961946980917457e-05, | |
| "loss": 0.7591, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.7486099600791931, | |
| "eval_runtime": 1.9436, | |
| "eval_samples_per_second": 80.264, | |
| "eval_steps_per_second": 4.116, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.0514655113220215, | |
| "learning_rate": 1.9914448613738107e-05, | |
| "loss": 0.7497, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_loss": 0.7431505918502808, | |
| "eval_runtime": 1.9501, | |
| "eval_samples_per_second": 79.994, | |
| "eval_steps_per_second": 4.102, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.5434212684631348, | |
| "learning_rate": 1.9848077530122083e-05, | |
| "loss": 0.7297, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.7465248107910156, | |
| "eval_runtime": 1.9802, | |
| "eval_samples_per_second": 78.781, | |
| "eval_steps_per_second": 4.04, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 3.6651933193206787, | |
| "learning_rate": 1.9762960071199334e-05, | |
| "loss": 0.7435, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_loss": 0.7544175982475281, | |
| "eval_runtime": 1.9459, | |
| "eval_samples_per_second": 80.167, | |
| "eval_steps_per_second": 4.111, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.424015522003174, | |
| "learning_rate": 1.9659258262890683e-05, | |
| "loss": 0.7611, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "eval_loss": 0.7523372173309326, | |
| "eval_runtime": 2.0693, | |
| "eval_samples_per_second": 75.388, | |
| "eval_steps_per_second": 3.866, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 2.423607587814331, | |
| "learning_rate": 1.953716950748227e-05, | |
| "loss": 0.7622, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 0.7510794401168823, | |
| "eval_runtime": 1.9388, | |
| "eval_samples_per_second": 80.462, | |
| "eval_steps_per_second": 4.126, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.693723678588867, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.7068, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.7511183619499207, | |
| "eval_runtime": 1.9737, | |
| "eval_samples_per_second": 79.04, | |
| "eval_steps_per_second": 4.053, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.4280779361724854, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 0.7297, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_loss": 0.7428472638130188, | |
| "eval_runtime": 1.9385, | |
| "eval_samples_per_second": 80.476, | |
| "eval_steps_per_second": 4.127, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 2.3453116416931152, | |
| "learning_rate": 1.9063077870366504e-05, | |
| "loss": 0.7415, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_loss": 0.7462921142578125, | |
| "eval_runtime": 1.9445, | |
| "eval_samples_per_second": 80.225, | |
| "eval_steps_per_second": 4.114, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 2.466562271118164, | |
| "learning_rate": 1.887010833178222e-05, | |
| "loss": 0.7309, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 0.7448516488075256, | |
| "eval_runtime": 1.9649, | |
| "eval_samples_per_second": 79.392, | |
| "eval_steps_per_second": 4.071, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.566894292831421, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.7362, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.7468461394309998, | |
| "eval_runtime": 1.9409, | |
| "eval_samples_per_second": 80.377, | |
| "eval_steps_per_second": 4.122, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.3160593509674072, | |
| "learning_rate": 1.843391445812886e-05, | |
| "loss": 0.3678, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_loss": 0.7951092720031738, | |
| "eval_runtime": 1.9636, | |
| "eval_samples_per_second": 79.445, | |
| "eval_steps_per_second": 4.074, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 2.5666403770446777, | |
| "learning_rate": 1.819152044288992e-05, | |
| "loss": 0.3964, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_loss": 0.78642338514328, | |
| "eval_runtime": 1.9832, | |
| "eval_samples_per_second": 78.66, | |
| "eval_steps_per_second": 4.034, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.0465054512023926, | |
| "learning_rate": 1.7933533402912354e-05, | |
| "loss": 0.3784, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_loss": 0.7795366644859314, | |
| "eval_runtime": 1.9712, | |
| "eval_samples_per_second": 79.139, | |
| "eval_steps_per_second": 4.058, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.862433910369873, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.3834, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.7931370139122009, | |
| "eval_runtime": 1.9437, | |
| "eval_samples_per_second": 80.26, | |
| "eval_steps_per_second": 4.116, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.331970691680908, | |
| "learning_rate": 1.737277336810124e-05, | |
| "loss": 0.3578, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.7878620624542236, | |
| "eval_runtime": 1.9394, | |
| "eval_samples_per_second": 80.438, | |
| "eval_steps_per_second": 4.125, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.595731496810913, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.3719, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_loss": 0.7958313226699829, | |
| "eval_runtime": 1.9743, | |
| "eval_samples_per_second": 79.014, | |
| "eval_steps_per_second": 4.052, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 2.6935150623321533, | |
| "learning_rate": 1.6755902076156606e-05, | |
| "loss": 0.3733, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "eval_loss": 0.786521852016449, | |
| "eval_runtime": 1.9731, | |
| "eval_samples_per_second": 79.063, | |
| "eval_steps_per_second": 4.055, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.2424511909484863, | |
| "learning_rate": 1.6427876096865394e-05, | |
| "loss": 0.3759, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_loss": 0.7788485884666443, | |
| "eval_runtime": 1.9496, | |
| "eval_samples_per_second": 80.016, | |
| "eval_steps_per_second": 4.103, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 1.8085654973983765, | |
| "learning_rate": 1.608761429008721e-05, | |
| "loss": 0.3821, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_loss": 0.791174054145813, | |
| "eval_runtime": 1.968, | |
| "eval_samples_per_second": 79.27, | |
| "eval_steps_per_second": 4.065, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.0668280124664307, | |
| "learning_rate": 1.573576436351046e-05, | |
| "loss": 0.3863, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_loss": 0.7752686738967896, | |
| "eval_runtime": 1.9725, | |
| "eval_samples_per_second": 79.087, | |
| "eval_steps_per_second": 4.056, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.206247091293335, | |
| "learning_rate": 1.5372996083468242e-05, | |
| "loss": 0.388, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "eval_loss": 0.7804670929908752, | |
| "eval_runtime": 1.9824, | |
| "eval_samples_per_second": 78.692, | |
| "eval_steps_per_second": 4.035, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.0638809204101562, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3747, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.7704665660858154, | |
| "eval_runtime": 1.9518, | |
| "eval_samples_per_second": 79.928, | |
| "eval_steps_per_second": 4.099, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.225670337677002, | |
| "learning_rate": 1.4617486132350343e-05, | |
| "loss": 0.3843, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "eval_loss": 0.7650146484375, | |
| "eval_runtime": 1.9688, | |
| "eval_samples_per_second": 79.235, | |
| "eval_steps_per_second": 4.063, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.266324043273926, | |
| "learning_rate": 1.4226182617406996e-05, | |
| "loss": 0.382, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_loss": 0.7695351839065552, | |
| "eval_runtime": 1.9796, | |
| "eval_samples_per_second": 78.804, | |
| "eval_steps_per_second": 4.041, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 3.5260887145996094, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 0.3728, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_loss": 0.7620097398757935, | |
| "eval_runtime": 1.94, | |
| "eval_samples_per_second": 80.411, | |
| "eval_steps_per_second": 4.124, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.2748708724975586, | |
| "learning_rate": 1.342020143325669e-05, | |
| "loss": 0.3633, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 0.7597982287406921, | |
| "eval_runtime": 1.9492, | |
| "eval_samples_per_second": 80.035, | |
| "eval_steps_per_second": 4.104, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 2.0982842445373535, | |
| "learning_rate": 1.300705799504273e-05, | |
| "loss": 0.365, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_loss": 0.7562901973724365, | |
| "eval_runtime": 1.9752, | |
| "eval_samples_per_second": 78.981, | |
| "eval_steps_per_second": 4.05, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 2.122321367263794, | |
| "learning_rate": 1.2588190451025209e-05, | |
| "loss": 0.3941, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_loss": 0.7595393061637878, | |
| "eval_runtime": 3.6677, | |
| "eval_samples_per_second": 42.533, | |
| "eval_steps_per_second": 2.181, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.7221115827560425, | |
| "learning_rate": 1.2164396139381029e-05, | |
| "loss": 0.3929, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_loss": 0.7426199316978455, | |
| "eval_runtime": 1.9437, | |
| "eval_samples_per_second": 80.258, | |
| "eval_steps_per_second": 4.116, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.285029888153076, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.377, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.7527384161949158, | |
| "eval_runtime": 1.9813, | |
| "eval_samples_per_second": 78.738, | |
| "eval_steps_per_second": 4.038, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 1.6015441417694092, | |
| "learning_rate": 1.130526192220052e-05, | |
| "loss": 0.1718, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_loss": 0.8362773060798645, | |
| "eval_runtime": 1.951, | |
| "eval_samples_per_second": 79.961, | |
| "eval_steps_per_second": 4.101, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.6940577030181885, | |
| "learning_rate": 1.0871557427476585e-05, | |
| "loss": 0.1756, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "eval_loss": 0.8487461805343628, | |
| "eval_runtime": 1.9498, | |
| "eval_samples_per_second": 80.006, | |
| "eval_steps_per_second": 4.103, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 2.2040927410125732, | |
| "learning_rate": 1.0436193873653362e-05, | |
| "loss": 0.1732, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "eval_loss": 0.8493539094924927, | |
| "eval_runtime": 1.9411, | |
| "eval_samples_per_second": 80.366, | |
| "eval_steps_per_second": 4.121, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.932246446609497, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1697, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_loss": 0.8528260588645935, | |
| "eval_runtime": 1.943, | |
| "eval_samples_per_second": 80.289, | |
| "eval_steps_per_second": 4.117, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.6120601892471313, | |
| "learning_rate": 9.563806126346643e-06, | |
| "loss": 0.1751, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "eval_loss": 0.8474267721176147, | |
| "eval_runtime": 1.9441, | |
| "eval_samples_per_second": 80.244, | |
| "eval_steps_per_second": 4.115, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.7986987829208374, | |
| "learning_rate": 9.128442572523418e-06, | |
| "loss": 0.1823, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "eval_loss": 0.8338098526000977, | |
| "eval_runtime": 1.9366, | |
| "eval_samples_per_second": 80.555, | |
| "eval_steps_per_second": 4.131, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 1.5987346172332764, | |
| "learning_rate": 8.694738077799487e-06, | |
| "loss": 0.1766, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_loss": 0.8440003991127014, | |
| "eval_runtime": 1.9724, | |
| "eval_samples_per_second": 79.09, | |
| "eval_steps_per_second": 4.056, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.1890445947647095, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.1726, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.8430134654045105, | |
| "eval_runtime": 1.9429, | |
| "eval_samples_per_second": 80.292, | |
| "eval_steps_per_second": 4.118, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.1193804740905762, | |
| "learning_rate": 7.835603860618973e-06, | |
| "loss": 0.1801, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_loss": 0.8422533869743347, | |
| "eval_runtime": 1.9378, | |
| "eval_samples_per_second": 80.504, | |
| "eval_steps_per_second": 4.128, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.293116807937622, | |
| "learning_rate": 7.411809548974792e-06, | |
| "loss": 0.1738, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.8328295946121216, | |
| "eval_runtime": 1.9499, | |
| "eval_samples_per_second": 80.005, | |
| "eval_steps_per_second": 4.103, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 1.0970814228057861, | |
| "learning_rate": 6.992942004957271e-06, | |
| "loss": 0.1721, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_loss": 0.8437429666519165, | |
| "eval_runtime": 1.9535, | |
| "eval_samples_per_second": 79.856, | |
| "eval_steps_per_second": 4.095, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.5809910297393799, | |
| "learning_rate": 6.579798566743314e-06, | |
| "loss": 0.1773, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "eval_loss": 0.8351316452026367, | |
| "eval_runtime": 1.9853, | |
| "eval_samples_per_second": 78.576, | |
| "eval_steps_per_second": 4.03, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 1.7653720378875732, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 0.1716, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "eval_loss": 0.8409612774848938, | |
| "eval_runtime": 1.9672, | |
| "eval_samples_per_second": 79.301, | |
| "eval_steps_per_second": 4.067, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.1679255962371826, | |
| "learning_rate": 5.773817382593008e-06, | |
| "loss": 0.1677, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "eval_loss": 0.8521665930747986, | |
| "eval_runtime": 1.9722, | |
| "eval_samples_per_second": 79.098, | |
| "eval_steps_per_second": 4.056, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.1089847087860107, | |
| "learning_rate": 5.382513867649663e-06, | |
| "loss": 0.173, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_loss": 0.8496531248092651, | |
| "eval_runtime": 1.9402, | |
| "eval_samples_per_second": 80.405, | |
| "eval_steps_per_second": 4.123, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.194023609161377, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.1759, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 0.8479322791099548, | |
| "eval_runtime": 1.9504, | |
| "eval_samples_per_second": 79.982, | |
| "eval_steps_per_second": 4.102, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 1.9135856628417969, | |
| "learning_rate": 4.627003916531761e-06, | |
| "loss": 0.1727, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "eval_loss": 0.8385621309280396, | |
| "eval_runtime": 1.9425, | |
| "eval_samples_per_second": 80.31, | |
| "eval_steps_per_second": 4.118, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.567000389099121, | |
| "learning_rate": 4.264235636489542e-06, | |
| "loss": 0.1797, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "eval_loss": 0.8441168069839478, | |
| "eval_runtime": 1.9451, | |
| "eval_samples_per_second": 80.2, | |
| "eval_steps_per_second": 4.113, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.9476852416992188, | |
| "learning_rate": 3.912385709912794e-06, | |
| "loss": 0.1692, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "eval_loss": 0.8405532836914062, | |
| "eval_runtime": 1.9399, | |
| "eval_samples_per_second": 80.414, | |
| "eval_steps_per_second": 4.124, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.281925082206726, | |
| "learning_rate": 3.5721239031346067e-06, | |
| "loss": 0.1657, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.8360433578491211, | |
| "eval_runtime": 1.9471, | |
| "eval_samples_per_second": 80.118, | |
| "eval_steps_per_second": 4.109, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 1.1708404112293888e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |