|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 3.0, |
|
"eval_steps": 500, |
|
"global_step": 900, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 0.625, |
|
"learning_rate": 7.4074074074074075e-06, |
|
"loss": 0.4546, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 0.41796875, |
|
"learning_rate": 9.994529604032901e-06, |
|
"loss": 0.2781, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 0.38671875, |
|
"learning_rate": 9.964784918620284e-06, |
|
"loss": 0.2492, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 0.3984375, |
|
"learning_rate": 9.909333658705933e-06, |
|
"loss": 0.2379, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 0.400390625, |
|
"learning_rate": 9.82846293883108e-06, |
|
"loss": 0.2301, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 0.37109375, |
|
"learning_rate": 9.722591489961829e-06, |
|
"loss": 0.2255, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 0.388671875, |
|
"learning_rate": 9.59226749139145e-06, |
|
"loss": 0.2223, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 0.76171875, |
|
"learning_rate": 9.438165732387471e-06, |
|
"loss": 0.2171, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 0.39453125, |
|
"learning_rate": 9.261084118279846e-06, |
|
"loss": 0.2154, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 0.41015625, |
|
"learning_rate": 9.06193953908105e-06, |
|
"loss": 0.2126, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 0.37109375, |
|
"learning_rate": 8.84176312202936e-06, |
|
"loss": 0.2143, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 0.37890625, |
|
"learning_rate": 8.601694892636701e-06, |
|
"loss": 0.2082, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 0.369140625, |
|
"learning_rate": 8.34297787188496e-06, |
|
"loss": 0.2072, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 8.066951640134183e-06, |
|
"loss": 0.2069, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 7.77504540106735e-06, |
|
"loss": 0.2067, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 0.431640625, |
|
"learning_rate": 7.468770581585147e-06, |
|
"loss": 0.1909, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 1.13, |
|
"grad_norm": 0.39453125, |
|
"learning_rate": 7.149713005966784e-06, |
|
"loss": 0.1909, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 0.400390625, |
|
"learning_rate": 6.819524684817439e-06, |
|
"loss": 0.1877, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"grad_norm": 0.396484375, |
|
"learning_rate": 6.479915261317299e-06, |
|
"loss": 0.1901, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 1.33, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 6.132643159061707e-06, |
|
"loss": 0.1886, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"grad_norm": 0.376953125, |
|
"learning_rate": 5.779506477326933e-06, |
|
"loss": 0.1887, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 1.47, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 5.4223336809039205e-06, |
|
"loss": 0.1866, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"grad_norm": 0.419921875, |
|
"learning_rate": 5.062974132706017e-06, |
|
"loss": 0.1887, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"grad_norm": 0.390625, |
|
"learning_rate": 4.703288518170774e-06, |
|
"loss": 0.1849, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 1.67, |
|
"grad_norm": 0.390625, |
|
"learning_rate": 4.345139211036192e-06, |
|
"loss": 0.187, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.73, |
|
"grad_norm": 0.421875, |
|
"learning_rate": 3.9903806303753036e-06, |
|
"loss": 0.1839, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 1.8, |
|
"grad_norm": 0.40234375, |
|
"learning_rate": 3.6408496388182857e-06, |
|
"loss": 0.187, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 1.87, |
|
"grad_norm": 0.390625, |
|
"learning_rate": 3.2983560316780104e-06, |
|
"loss": 0.1849, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 1.93, |
|
"grad_norm": 0.412109375, |
|
"learning_rate": 2.9646731662242554e-06, |
|
"loss": 0.184, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 0.4140625, |
|
"learning_rate": 2.6415287796261707e-06, |
|
"loss": 0.1844, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.07, |
|
"grad_norm": 0.412109375, |
|
"learning_rate": 2.330596043105683e-06, |
|
"loss": 0.1795, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 2.13, |
|
"grad_norm": 0.4453125, |
|
"learning_rate": 2.03348489862149e-06, |
|
"loss": 0.1787, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 2.2, |
|
"grad_norm": 0.39453125, |
|
"learning_rate": 1.7517337229403946e-06, |
|
"loss": 0.1781, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 2.27, |
|
"grad_norm": 0.39453125, |
|
"learning_rate": 1.4868013622576138e-06, |
|
"loss": 0.1759, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 2.33, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 1.240059578609054e-06, |
|
"loss": 0.178, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"grad_norm": 0.416015625, |
|
"learning_rate": 1.012785947186397e-06, |
|
"loss": 0.1785, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 2.47, |
|
"grad_norm": 0.40234375, |
|
"learning_rate": 8.061572413311253e-07, |
|
"loss": 0.1778, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"grad_norm": 0.404296875, |
|
"learning_rate": 6.212433394585865e-07, |
|
"loss": 0.1771, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"grad_norm": 0.41796875, |
|
"learning_rate": 4.590016854606727e-07, |
|
"loss": 0.1756, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"grad_norm": 0.392578125, |
|
"learning_rate": 3.2027233126997405e-07, |
|
"loss": 0.1776, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 2.73, |
|
"grad_norm": 0.400390625, |
|
"learning_rate": 2.057735872539157e-07, |
|
"loss": 0.1764, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"grad_norm": 0.412109375, |
|
"learning_rate": 1.1609830296019142e-07, |
|
"loss": 0.1815, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 2.87, |
|
"grad_norm": 0.490234375, |
|
"learning_rate": 5.1710797470987393e-08, |
|
"loss": 0.1763, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 2.93, |
|
"grad_norm": 0.400390625, |
|
"learning_rate": 1.2944455259944476e-08, |
|
"loss": 0.1788, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 0.41015625, |
|
"learning_rate": 0.0, |
|
"loss": 0.1766, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"step": 900, |
|
"total_flos": 4.549145968838181e+18, |
|
"train_loss": 0.20135066880120173, |
|
"train_runtime": 71504.7155, |
|
"train_samples_per_second": 2.411, |
|
"train_steps_per_second": 0.013 |
|
} |
|
], |
|
"logging_steps": 20, |
|
"max_steps": 900, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 10000, |
|
"total_flos": 4.549145968838181e+18, |
|
"train_batch_size": 24, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|