|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 1.0, |
|
"global_step": 10000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 9.900000000000001e-05, |
|
"loss": 7.687, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 9.8e-05, |
|
"loss": 0.4531, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 9.7e-05, |
|
"loss": 0.3587, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 9.6e-05, |
|
"loss": 0.3042, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 9.5e-05, |
|
"loss": 0.2718, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 9.4e-05, |
|
"loss": 0.2143, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 9.300000000000001e-05, |
|
"loss": 0.1839, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 9.200000000000001e-05, |
|
"loss": 0.1717, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 9.1e-05, |
|
"loss": 0.1688, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 9e-05, |
|
"loss": 0.1566, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 8.900000000000001e-05, |
|
"loss": 0.1528, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 8.800000000000001e-05, |
|
"loss": 0.1461, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 8.7e-05, |
|
"loss": 0.146, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 8.6e-05, |
|
"loss": 0.1388, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 8.5e-05, |
|
"loss": 0.1343, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 8.4e-05, |
|
"loss": 0.1318, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 8.3e-05, |
|
"loss": 0.1289, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 8.2e-05, |
|
"loss": 0.1285, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 8.1e-05, |
|
"loss": 0.1289, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 8e-05, |
|
"loss": 0.1286, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"learning_rate": 7.900000000000001e-05, |
|
"loss": 0.1259, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"learning_rate": 7.800000000000001e-05, |
|
"loss": 0.1232, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"learning_rate": 7.7e-05, |
|
"loss": 0.1199, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"learning_rate": 7.6e-05, |
|
"loss": 0.1227, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"learning_rate": 7.500000000000001e-05, |
|
"loss": 0.1165, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"learning_rate": 7.4e-05, |
|
"loss": 0.1214, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"learning_rate": 7.3e-05, |
|
"loss": 0.1191, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"learning_rate": 7.2e-05, |
|
"loss": 0.12, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"learning_rate": 7.1e-05, |
|
"loss": 0.1213, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"learning_rate": 7e-05, |
|
"loss": 0.116, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"learning_rate": 6.9e-05, |
|
"loss": 0.117, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"learning_rate": 6.800000000000001e-05, |
|
"loss": 0.1138, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"learning_rate": 6.7e-05, |
|
"loss": 0.113, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"learning_rate": 6.6e-05, |
|
"loss": 0.1107, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"learning_rate": 6.500000000000001e-05, |
|
"loss": 0.1127, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"learning_rate": 6.400000000000001e-05, |
|
"loss": 0.1075, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"learning_rate": 6.3e-05, |
|
"loss": 0.1109, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"learning_rate": 6.2e-05, |
|
"loss": 0.1076, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"learning_rate": 6.1e-05, |
|
"loss": 0.1068, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 6e-05, |
|
"loss": 0.1064, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"learning_rate": 5.9e-05, |
|
"loss": 0.108, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"learning_rate": 5.8e-05, |
|
"loss": 0.1079, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"learning_rate": 5.6999999999999996e-05, |
|
"loss": 0.1035, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"learning_rate": 5.6000000000000006e-05, |
|
"loss": 0.1039, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"learning_rate": 5.500000000000001e-05, |
|
"loss": 0.1068, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"learning_rate": 5.4000000000000005e-05, |
|
"loss": 0.1062, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"learning_rate": 5.300000000000001e-05, |
|
"loss": 0.1054, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"learning_rate": 5.2000000000000004e-05, |
|
"loss": 0.1074, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"learning_rate": 5.1000000000000006e-05, |
|
"loss": 0.106, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"learning_rate": 5e-05, |
|
"loss": 0.1045, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"learning_rate": 4.9e-05, |
|
"loss": 0.1057, |
|
"step": 5100 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"learning_rate": 4.8e-05, |
|
"loss": 0.1053, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"learning_rate": 4.7e-05, |
|
"loss": 0.105, |
|
"step": 5300 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"learning_rate": 4.600000000000001e-05, |
|
"loss": 0.1021, |
|
"step": 5400 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"learning_rate": 4.5e-05, |
|
"loss": 0.105, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 4.4000000000000006e-05, |
|
"loss": 0.1044, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"learning_rate": 4.3e-05, |
|
"loss": 0.0998, |
|
"step": 5700 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"learning_rate": 4.2e-05, |
|
"loss": 0.1043, |
|
"step": 5800 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"learning_rate": 4.1e-05, |
|
"loss": 0.1029, |
|
"step": 5900 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"learning_rate": 4e-05, |
|
"loss": 0.0997, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"learning_rate": 3.9000000000000006e-05, |
|
"loss": 0.1032, |
|
"step": 6100 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"learning_rate": 3.8e-05, |
|
"loss": 0.1015, |
|
"step": 6200 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"learning_rate": 3.7e-05, |
|
"loss": 0.1011, |
|
"step": 6300 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"learning_rate": 3.6e-05, |
|
"loss": 0.1021, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"learning_rate": 3.5e-05, |
|
"loss": 0.0997, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"learning_rate": 3.4000000000000007e-05, |
|
"loss": 0.1005, |
|
"step": 6600 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"learning_rate": 3.3e-05, |
|
"loss": 0.0984, |
|
"step": 6700 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"learning_rate": 3.2000000000000005e-05, |
|
"loss": 0.1033, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"learning_rate": 3.1e-05, |
|
"loss": 0.0966, |
|
"step": 6900 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"learning_rate": 3e-05, |
|
"loss": 0.1013, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"learning_rate": 2.9e-05, |
|
"loss": 0.0994, |
|
"step": 7100 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"learning_rate": 2.8000000000000003e-05, |
|
"loss": 0.1001, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"learning_rate": 2.7000000000000002e-05, |
|
"loss": 0.0977, |
|
"step": 7300 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"learning_rate": 2.6000000000000002e-05, |
|
"loss": 0.0983, |
|
"step": 7400 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"learning_rate": 2.5e-05, |
|
"loss": 0.0996, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"learning_rate": 2.4e-05, |
|
"loss": 0.0946, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"learning_rate": 2.3000000000000003e-05, |
|
"loss": 0.0967, |
|
"step": 7700 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"learning_rate": 2.2000000000000003e-05, |
|
"loss": 0.0978, |
|
"step": 7800 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"learning_rate": 2.1e-05, |
|
"loss": 0.0962, |
|
"step": 7900 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0995, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 1.9e-05, |
|
"loss": 0.1004, |
|
"step": 8100 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"learning_rate": 1.8e-05, |
|
"loss": 0.101, |
|
"step": 8200 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"learning_rate": 1.7000000000000003e-05, |
|
"loss": 0.0972, |
|
"step": 8300 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"learning_rate": 1.6000000000000003e-05, |
|
"loss": 0.0965, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"learning_rate": 1.5e-05, |
|
"loss": 0.0979, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"learning_rate": 1.4000000000000001e-05, |
|
"loss": 0.0929, |
|
"step": 8600 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"learning_rate": 1.3000000000000001e-05, |
|
"loss": 0.0979, |
|
"step": 8700 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"learning_rate": 1.2e-05, |
|
"loss": 0.0979, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 1.1000000000000001e-05, |
|
"loss": 0.0983, |
|
"step": 8900 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"learning_rate": 1e-05, |
|
"loss": 0.0978, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"learning_rate": 9e-06, |
|
"loss": 0.0949, |
|
"step": 9100 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 0.0954, |
|
"step": 9200 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"learning_rate": 7.000000000000001e-06, |
|
"loss": 0.0976, |
|
"step": 9300 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"learning_rate": 6e-06, |
|
"loss": 0.0973, |
|
"step": 9400 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"learning_rate": 5e-06, |
|
"loss": 0.095, |
|
"step": 9500 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 0.0952, |
|
"step": 9600 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"learning_rate": 3e-06, |
|
"loss": 0.0955, |
|
"step": 9700 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 0.0983, |
|
"step": 9800 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 0.0955, |
|
"step": 9900 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 0.0, |
|
"loss": 0.0953, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"step": 10000, |
|
"total_flos": 0, |
|
"train_runtime": 46877.0363, |
|
"train_samples_per_second": 0.213 |
|
} |
|
], |
|
"max_steps": 10000, |
|
"num_train_epochs": 1, |
|
"total_flos": 0, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|