|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 20.0, |
|
"eval_steps": 500, |
|
"global_step": 2440, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 5.0601301193237305, |
|
"learning_rate": 4.75e-05, |
|
"loss": 0.5621, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.7218045112781954, |
|
"eval_f1": 0.6526816032372154, |
|
"eval_loss": 0.5100470185279846, |
|
"eval_precision": 0.6593383311603651, |
|
"eval_recall": 0.6481633024186215, |
|
"eval_runtime": 5.1999, |
|
"eval_samples_per_second": 76.732, |
|
"eval_steps_per_second": 9.615, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 3.8218886852264404, |
|
"learning_rate": 4.5e-05, |
|
"loss": 0.5049, |
|
"step": 244 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.7343358395989975, |
|
"eval_f1": 0.7011278195488722, |
|
"eval_loss": 0.48896074295043945, |
|
"eval_precision": 0.6944563758389262, |
|
"eval_recall": 0.7195399163484271, |
|
"eval_runtime": 5.0602, |
|
"eval_samples_per_second": 78.85, |
|
"eval_steps_per_second": 9.881, |
|
"step": 244 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 4.400880813598633, |
|
"learning_rate": 4.25e-05, |
|
"loss": 0.4776, |
|
"step": 366 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.7593984962406015, |
|
"eval_f1": 0.7216117216117216, |
|
"eval_loss": 0.4479529559612274, |
|
"eval_precision": 0.7150252525252525, |
|
"eval_recall": 0.7322695035460993, |
|
"eval_runtime": 5.0795, |
|
"eval_samples_per_second": 78.551, |
|
"eval_steps_per_second": 9.843, |
|
"step": 366 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 2.5400054454803467, |
|
"learning_rate": 4e-05, |
|
"loss": 0.4422, |
|
"step": 488 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.7944862155388471, |
|
"eval_f1": 0.7482456140350877, |
|
"eval_loss": 0.4104432165622711, |
|
"eval_precision": 0.75243993993994, |
|
"eval_recall": 0.7445899254409893, |
|
"eval_runtime": 5.0748, |
|
"eval_samples_per_second": 78.624, |
|
"eval_steps_per_second": 9.853, |
|
"step": 488 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": 1.0803650617599487, |
|
"learning_rate": 3.7500000000000003e-05, |
|
"loss": 0.4146, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.7593984962406015, |
|
"eval_f1": 0.7282941777323799, |
|
"eval_loss": 0.42569735646247864, |
|
"eval_precision": 0.7202380952380952, |
|
"eval_recall": 0.7472722313147845, |
|
"eval_runtime": 5.0669, |
|
"eval_samples_per_second": 78.747, |
|
"eval_steps_per_second": 9.868, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 1.867180585861206, |
|
"learning_rate": 3.5e-05, |
|
"loss": 0.3828, |
|
"step": 732 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.8245614035087719, |
|
"eval_f1": 0.7894173351830629, |
|
"eval_loss": 0.3869302272796631, |
|
"eval_precision": 0.7880252100840336, |
|
"eval_recall": 0.7908710674668122, |
|
"eval_runtime": 5.0936, |
|
"eval_samples_per_second": 78.333, |
|
"eval_steps_per_second": 9.816, |
|
"step": 732 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"grad_norm": 1.7762138843536377, |
|
"learning_rate": 3.2500000000000004e-05, |
|
"loss": 0.3697, |
|
"step": 854 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.8145363408521303, |
|
"eval_f1": 0.7854090354090355, |
|
"eval_loss": 0.3959118127822876, |
|
"eval_precision": 0.7765993265993265, |
|
"eval_recall": 0.7987815966539371, |
|
"eval_runtime": 5.0555, |
|
"eval_samples_per_second": 78.923, |
|
"eval_steps_per_second": 9.89, |
|
"step": 854 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"grad_norm": 5.161904335021973, |
|
"learning_rate": 3e-05, |
|
"loss": 0.3486, |
|
"step": 976 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.8320802005012531, |
|
"eval_f1": 0.801779557335113, |
|
"eval_loss": 0.38075554370880127, |
|
"eval_precision": 0.7960927960927962, |
|
"eval_recall": 0.8086924895435534, |
|
"eval_runtime": 5.0601, |
|
"eval_samples_per_second": 78.852, |
|
"eval_steps_per_second": 9.881, |
|
"step": 976 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"grad_norm": 4.554713249206543, |
|
"learning_rate": 2.7500000000000004e-05, |
|
"loss": 0.3437, |
|
"step": 1098 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.8270676691729323, |
|
"eval_f1": 0.7949075143216848, |
|
"eval_loss": 0.3737644851207733, |
|
"eval_precision": 0.7904105571847508, |
|
"eval_recall": 0.8001454809965449, |
|
"eval_runtime": 5.1356, |
|
"eval_samples_per_second": 77.694, |
|
"eval_steps_per_second": 9.736, |
|
"step": 1098 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 6.5807719230651855, |
|
"learning_rate": 2.5e-05, |
|
"loss": 0.3317, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.8471177944862155, |
|
"eval_f1": 0.8151164115613962, |
|
"eval_loss": 0.3643290102481842, |
|
"eval_precision": 0.8159193371512123, |
|
"eval_recall": 0.8143298781596654, |
|
"eval_runtime": 5.08, |
|
"eval_samples_per_second": 78.544, |
|
"eval_steps_per_second": 9.843, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"grad_norm": 4.961609363555908, |
|
"learning_rate": 2.25e-05, |
|
"loss": 0.3114, |
|
"step": 1342 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.8270676691729323, |
|
"eval_f1": 0.7967966933608887, |
|
"eval_loss": 0.36825552582740784, |
|
"eval_precision": 0.7902444649446494, |
|
"eval_recall": 0.8051463902527732, |
|
"eval_runtime": 5.0688, |
|
"eval_samples_per_second": 78.716, |
|
"eval_steps_per_second": 9.864, |
|
"step": 1342 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 10.724159240722656, |
|
"learning_rate": 2e-05, |
|
"loss": 0.3035, |
|
"step": 1464 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.8345864661654135, |
|
"eval_f1": 0.8060710498409331, |
|
"eval_loss": 0.36597177386283875, |
|
"eval_precision": 0.7988372093023256, |
|
"eval_recall": 0.8154664484451719, |
|
"eval_runtime": 5.0475, |
|
"eval_samples_per_second": 79.05, |
|
"eval_steps_per_second": 9.906, |
|
"step": 1464 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"grad_norm": 1.6210856437683105, |
|
"learning_rate": 1.75e-05, |
|
"loss": 0.3117, |
|
"step": 1586 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.8471177944862155, |
|
"eval_f1": 0.8141734808401475, |
|
"eval_loss": 0.3518490195274353, |
|
"eval_precision": 0.8166666666666667, |
|
"eval_recall": 0.8118294235315512, |
|
"eval_runtime": 5.1528, |
|
"eval_samples_per_second": 77.434, |
|
"eval_steps_per_second": 9.703, |
|
"step": 1586 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"grad_norm": 12.59556770324707, |
|
"learning_rate": 1.5e-05, |
|
"loss": 0.3048, |
|
"step": 1708 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.8446115288220551, |
|
"eval_f1": 0.814390756302521, |
|
"eval_loss": 0.3533446788787842, |
|
"eval_precision": 0.8114780902550687, |
|
"eval_recall": 0.8175577377705037, |
|
"eval_runtime": 5.0807, |
|
"eval_samples_per_second": 78.532, |
|
"eval_steps_per_second": 9.841, |
|
"step": 1708 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"grad_norm": 1.0728732347488403, |
|
"learning_rate": 1.25e-05, |
|
"loss": 0.2916, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.8421052631578947, |
|
"eval_f1": 0.8118502107020158, |
|
"eval_loss": 0.35696789622306824, |
|
"eval_precision": 0.8083091673078061, |
|
"eval_recall": 0.8157846881251136, |
|
"eval_runtime": 5.066, |
|
"eval_samples_per_second": 78.761, |
|
"eval_steps_per_second": 9.87, |
|
"step": 1830 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"grad_norm": 6.875555992126465, |
|
"learning_rate": 1e-05, |
|
"loss": 0.2832, |
|
"step": 1952 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.8471177944862155, |
|
"eval_f1": 0.8186863532409097, |
|
"eval_loss": 0.35785549879074097, |
|
"eval_precision": 0.8138123167155425, |
|
"eval_recall": 0.8243316966721222, |
|
"eval_runtime": 5.0594, |
|
"eval_samples_per_second": 78.863, |
|
"eval_steps_per_second": 9.883, |
|
"step": 1952 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"grad_norm": 4.91969108581543, |
|
"learning_rate": 7.5e-06, |
|
"loss": 0.284, |
|
"step": 2074 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.8471177944862155, |
|
"eval_f1": 0.8160386984618873, |
|
"eval_loss": 0.3496108949184418, |
|
"eval_precision": 0.8152632848784607, |
|
"eval_recall": 0.8168303327877796, |
|
"eval_runtime": 5.1011, |
|
"eval_samples_per_second": 78.218, |
|
"eval_steps_per_second": 9.802, |
|
"step": 2074 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"grad_norm": 3.59561824798584, |
|
"learning_rate": 5e-06, |
|
"loss": 0.2906, |
|
"step": 2196 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.8446115288220551, |
|
"eval_f1": 0.8152777777777778, |
|
"eval_loss": 0.35370829701423645, |
|
"eval_precision": 0.8110639802050195, |
|
"eval_recall": 0.8200581923986179, |
|
"eval_runtime": 5.0539, |
|
"eval_samples_per_second": 78.948, |
|
"eval_steps_per_second": 9.893, |
|
"step": 2196 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"grad_norm": 4.449235916137695, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.2805, |
|
"step": 2318 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_accuracy": 0.849624060150376, |
|
"eval_f1": 0.8186033824331697, |
|
"eval_loss": 0.3505256772041321, |
|
"eval_precision": 0.8186033824331697, |
|
"eval_recall": 0.8186033824331697, |
|
"eval_runtime": 5.085, |
|
"eval_samples_per_second": 78.465, |
|
"eval_steps_per_second": 9.833, |
|
"step": 2318 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"grad_norm": 8.314062118530273, |
|
"learning_rate": 0.0, |
|
"loss": 0.2815, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.8471177944862155, |
|
"eval_f1": 0.8169408154516122, |
|
"eval_loss": 0.35131508111953735, |
|
"eval_precision": 0.814695340501792, |
|
"eval_recall": 0.8193307874158937, |
|
"eval_runtime": 5.0603, |
|
"eval_samples_per_second": 78.849, |
|
"eval_steps_per_second": 9.881, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"step": 2440, |
|
"total_flos": 7597037114448000.0, |
|
"train_loss": 0.3560234429406338, |
|
"train_runtime": 1952.1853, |
|
"train_samples_per_second": 37.271, |
|
"train_steps_per_second": 1.25 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 2440, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 20, |
|
"save_steps": 500, |
|
"total_flos": 7597037114448000.0, |
|
"train_batch_size": 30, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|