|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 10.0, |
|
"eval_steps": 500, |
|
"global_step": 900, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 0.00019777777777777778, |
|
"loss": 0.0793, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"learning_rate": 0.00019555555555555556, |
|
"loss": 0.104, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"learning_rate": 0.00019333333333333333, |
|
"loss": 0.065, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"learning_rate": 0.00019111111111111114, |
|
"loss": 0.0836, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 0.00018888888888888888, |
|
"loss": 0.1303, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"learning_rate": 0.0001866666666666667, |
|
"loss": 0.0363, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"learning_rate": 0.00018444444444444446, |
|
"loss": 0.0567, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 0.00018222222222222224, |
|
"loss": 0.059, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 0.00018, |
|
"loss": 0.1205, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"learning_rate": 0.00017777777777777779, |
|
"loss": 0.0774, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"learning_rate": 0.00017555555555555556, |
|
"loss": 0.0807, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 1.33, |
|
"learning_rate": 0.00017333333333333334, |
|
"loss": 0.0377, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 1.44, |
|
"learning_rate": 0.0001711111111111111, |
|
"loss": 0.0621, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"learning_rate": 0.00016888888888888889, |
|
"loss": 0.0076, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 1.67, |
|
"learning_rate": 0.0001666666666666667, |
|
"loss": 0.0098, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 1.78, |
|
"learning_rate": 0.00016444444444444444, |
|
"loss": 0.0595, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 1.89, |
|
"learning_rate": 0.00016222222222222224, |
|
"loss": 0.0179, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 0.00016, |
|
"loss": 0.0269, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"learning_rate": 0.0001577777777777778, |
|
"loss": 0.0074, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"learning_rate": 0.00015555555555555556, |
|
"loss": 0.0246, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 2.33, |
|
"learning_rate": 0.00015333333333333334, |
|
"loss": 0.0481, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"learning_rate": 0.0001511111111111111, |
|
"loss": 0.0281, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 2.56, |
|
"learning_rate": 0.0001488888888888889, |
|
"loss": 0.0037, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"learning_rate": 0.00014666666666666666, |
|
"loss": 0.0099, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 2.78, |
|
"learning_rate": 0.00014444444444444444, |
|
"loss": 0.0202, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 2.89, |
|
"learning_rate": 0.00014222222222222224, |
|
"loss": 0.0082, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"learning_rate": 0.00014, |
|
"loss": 0.0221, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 3.11, |
|
"learning_rate": 0.0001377777777777778, |
|
"loss": 0.008, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 3.22, |
|
"learning_rate": 0.00013555555555555556, |
|
"loss": 0.0314, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 3.33, |
|
"learning_rate": 0.00013333333333333334, |
|
"loss": 0.0227, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 3.44, |
|
"learning_rate": 0.00013111111111111111, |
|
"loss": 0.0465, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 3.56, |
|
"learning_rate": 0.00012888888888888892, |
|
"loss": 0.026, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 3.67, |
|
"learning_rate": 0.00012666666666666666, |
|
"loss": 0.0058, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 3.78, |
|
"learning_rate": 0.00012444444444444444, |
|
"loss": 0.0071, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 3.89, |
|
"learning_rate": 0.00012222222222222224, |
|
"loss": 0.0256, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 0.00012, |
|
"loss": 0.0099, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 4.11, |
|
"learning_rate": 0.00011777777777777779, |
|
"loss": 0.0152, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 4.22, |
|
"learning_rate": 0.00011555555555555555, |
|
"loss": 0.0057, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 4.33, |
|
"learning_rate": 0.00011333333333333334, |
|
"loss": 0.0009, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 4.44, |
|
"learning_rate": 0.00011111111111111112, |
|
"loss": 0.0122, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 4.56, |
|
"learning_rate": 0.00010888888888888889, |
|
"loss": 0.0014, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 4.67, |
|
"learning_rate": 0.00010666666666666667, |
|
"loss": 0.0035, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 4.78, |
|
"learning_rate": 0.00010444444444444445, |
|
"loss": 0.0054, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 4.89, |
|
"learning_rate": 0.00010222222222222222, |
|
"loss": 0.0036, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.0009, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 5.11, |
|
"learning_rate": 9.777777777777778e-05, |
|
"loss": 0.0012, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 5.22, |
|
"learning_rate": 9.555555555555557e-05, |
|
"loss": 0.0006, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 5.33, |
|
"learning_rate": 9.333333333333334e-05, |
|
"loss": 0.0041, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 5.44, |
|
"learning_rate": 9.111111111111112e-05, |
|
"loss": 0.0019, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 5.56, |
|
"learning_rate": 8.888888888888889e-05, |
|
"loss": 0.0017, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 5.56, |
|
"eval_accuracy": 0.9101851851851852, |
|
"eval_loss": 0.5352890491485596, |
|
"eval_runtime": 12.7384, |
|
"eval_samples_per_second": 84.783, |
|
"eval_steps_per_second": 10.598, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 5.67, |
|
"learning_rate": 8.666666666666667e-05, |
|
"loss": 0.0091, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 5.78, |
|
"learning_rate": 8.444444444444444e-05, |
|
"loss": 0.0015, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 5.89, |
|
"learning_rate": 8.222222222222222e-05, |
|
"loss": 0.0059, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 8e-05, |
|
"loss": 0.0028, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 6.11, |
|
"learning_rate": 7.777777777777778e-05, |
|
"loss": 0.0004, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 6.22, |
|
"learning_rate": 7.555555555555556e-05, |
|
"loss": 0.0024, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 6.33, |
|
"learning_rate": 7.333333333333333e-05, |
|
"loss": 0.0004, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 6.44, |
|
"learning_rate": 7.111111111111112e-05, |
|
"loss": 0.0024, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 6.56, |
|
"learning_rate": 6.88888888888889e-05, |
|
"loss": 0.0004, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 6.67, |
|
"learning_rate": 6.666666666666667e-05, |
|
"loss": 0.0019, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 6.78, |
|
"learning_rate": 6.444444444444446e-05, |
|
"loss": 0.0005, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 6.89, |
|
"learning_rate": 6.222222222222222e-05, |
|
"loss": 0.004, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"learning_rate": 6e-05, |
|
"loss": 0.0047, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 7.11, |
|
"learning_rate": 5.7777777777777776e-05, |
|
"loss": 0.0004, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 7.22, |
|
"learning_rate": 5.555555555555556e-05, |
|
"loss": 0.003, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 7.33, |
|
"learning_rate": 5.333333333333333e-05, |
|
"loss": 0.0027, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 7.44, |
|
"learning_rate": 5.111111111111111e-05, |
|
"loss": 0.0022, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 7.56, |
|
"learning_rate": 4.888888888888889e-05, |
|
"loss": 0.0004, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 7.67, |
|
"learning_rate": 4.666666666666667e-05, |
|
"loss": 0.0031, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 7.78, |
|
"learning_rate": 4.4444444444444447e-05, |
|
"loss": 0.002, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 7.89, |
|
"learning_rate": 4.222222222222222e-05, |
|
"loss": 0.0004, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"learning_rate": 4e-05, |
|
"loss": 0.0021, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 8.11, |
|
"learning_rate": 3.777777777777778e-05, |
|
"loss": 0.0019, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 8.22, |
|
"learning_rate": 3.555555555555556e-05, |
|
"loss": 0.0033, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 8.33, |
|
"learning_rate": 3.3333333333333335e-05, |
|
"loss": 0.0022, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 8.44, |
|
"learning_rate": 3.111111111111111e-05, |
|
"loss": 0.0018, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 8.56, |
|
"learning_rate": 2.8888888888888888e-05, |
|
"loss": 0.0017, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 8.67, |
|
"learning_rate": 2.6666666666666667e-05, |
|
"loss": 0.002, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 8.78, |
|
"learning_rate": 2.4444444444444445e-05, |
|
"loss": 0.0004, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 8.89, |
|
"learning_rate": 2.2222222222222223e-05, |
|
"loss": 0.0003, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0022, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 9.11, |
|
"learning_rate": 1.777777777777778e-05, |
|
"loss": 0.0017, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 9.22, |
|
"learning_rate": 1.5555555555555555e-05, |
|
"loss": 0.0018, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 9.33, |
|
"learning_rate": 1.3333333333333333e-05, |
|
"loss": 0.0032, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 9.44, |
|
"learning_rate": 1.1111111111111112e-05, |
|
"loss": 0.0032, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 9.56, |
|
"learning_rate": 8.88888888888889e-06, |
|
"loss": 0.0003, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 9.67, |
|
"learning_rate": 6.666666666666667e-06, |
|
"loss": 0.0021, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 9.78, |
|
"learning_rate": 4.444444444444445e-06, |
|
"loss": 0.0003, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 9.89, |
|
"learning_rate": 2.2222222222222225e-06, |
|
"loss": 0.0003, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"learning_rate": 0.0, |
|
"loss": 0.002, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"step": 900, |
|
"total_flos": 3.347863979094835e+18, |
|
"train_loss": 0.017901433989819554, |
|
"train_runtime": 786.561, |
|
"train_samples_per_second": 54.923, |
|
"train_steps_per_second": 1.144 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 900, |
|
"num_train_epochs": 10, |
|
"save_steps": 1000, |
|
"total_flos": 3.347863979094835e+18, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|