|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 2.0, |
|
"global_step": 2484, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 4.0160642570281125e-05, |
|
"loss": 3.66, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 8.032128514056225e-05, |
|
"loss": 1.57, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"learning_rate": 9.771812080536913e-05, |
|
"loss": 1.36, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"learning_rate": 9.324384787472036e-05, |
|
"loss": 1.26, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 8.876957494407159e-05, |
|
"loss": 1.19, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"learning_rate": 8.429530201342283e-05, |
|
"loss": 1.14, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 7.982102908277405e-05, |
|
"loss": 1.12, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"learning_rate": 7.534675615212528e-05, |
|
"loss": 1.1, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"learning_rate": 7.087248322147651e-05, |
|
"loss": 1.04, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 6.639821029082774e-05, |
|
"loss": 1.035, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 6.192393736017897e-05, |
|
"loss": 0.985, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"learning_rate": 5.7449664429530206e-05, |
|
"loss": 0.97, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_HasAns_exact": 67.8608261412365, |
|
"eval_HasAns_f1": 74.52685937888027, |
|
"eval_HasAns_total": 18423, |
|
"eval_NoAns_exact": 67.53011054281471, |
|
"eval_NoAns_f1": 67.53011054281471, |
|
"eval_NoAns_total": 6061, |
|
"eval_best_exact": 67.7789576866525, |
|
"eval_best_exact_thresh": 0.0, |
|
"eval_best_f1": 72.79481826242022, |
|
"eval_best_f1_thresh": 0.0, |
|
"eval_exact": 67.7789576866525, |
|
"eval_f1": 72.79481826242029, |
|
"eval_total": 24484, |
|
"step": 1242 |
|
}, |
|
{ |
|
"epoch": 1.05, |
|
"learning_rate": 5.297539149888143e-05, |
|
"loss": 0.885, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 1.13, |
|
"learning_rate": 4.850111856823266e-05, |
|
"loss": 0.82, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"learning_rate": 4.40268456375839e-05, |
|
"loss": 0.83, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.29, |
|
"learning_rate": 3.955257270693513e-05, |
|
"loss": 0.835, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.37, |
|
"learning_rate": 3.507829977628635e-05, |
|
"loss": 0.82, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.45, |
|
"learning_rate": 3.0604026845637584e-05, |
|
"loss": 0.805, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"learning_rate": 2.6129753914988815e-05, |
|
"loss": 0.8, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 1.61, |
|
"learning_rate": 2.1655480984340046e-05, |
|
"loss": 0.795, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.69, |
|
"learning_rate": 1.7181208053691277e-05, |
|
"loss": 0.795, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 1.77, |
|
"learning_rate": 1.2706935123042507e-05, |
|
"loss": 0.76, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 1.85, |
|
"learning_rate": 8.232662192393737e-06, |
|
"loss": 0.76, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 1.93, |
|
"learning_rate": 3.7583892617449665e-06, |
|
"loss": 0.76, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_HasAns_exact": 69.19611355371003, |
|
"eval_HasAns_f1": 75.75383941256567, |
|
"eval_HasAns_total": 18423, |
|
"eval_NoAns_exact": 77.6109552879063, |
|
"eval_NoAns_f1": 77.6109552879063, |
|
"eval_NoAns_total": 6061, |
|
"eval_best_exact": 71.27920274464957, |
|
"eval_best_exact_thresh": 0.0, |
|
"eval_best_f1": 76.21356737043324, |
|
"eval_best_f1_thresh": 0.0, |
|
"eval_exact": 71.27920274464957, |
|
"eval_f1": 76.21356737043327, |
|
"eval_total": 24484, |
|
"step": 2484 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"step": 2484, |
|
"total_flos": 1.2455075452551168e+17, |
|
"train_loss": 1.0762882447665056, |
|
"train_runtime": 1067.9056, |
|
"train_samples_per_second": 616.543, |
|
"train_steps_per_second": 2.41 |
|
} |
|
], |
|
"max_steps": 2484, |
|
"num_train_epochs": 2, |
|
"total_flos": 1.2455075452551168e+17, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|