127 lines
3.2 KiB
JSON
127 lines
3.2 KiB
JSON
|
{
|
||
|
|
"best_global_step": 3170,
|
||
|
|
"best_metric": 76.43362369922768,
|
||
|
|
"best_model_checkpoint": "mbert-loraxs-qa-vanilla/run-14/checkpoint-3170",
|
||
|
|
"epoch": 5.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 3170,
|
||
|
|
"is_hyper_param_search": true,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"grad_norm": 4.179904460906982,
|
||
|
|
"learning_rate": 0.00025183820996153036,
|
||
|
|
"loss": 3.0449813698368495,
|
||
|
|
"step": 634
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_exact_match": 55.76,
|
||
|
|
"eval_f1": 70.60123748438602,
|
||
|
|
"eval_loss": 1.4089006185531616,
|
||
|
|
"eval_runtime": 7.7514,
|
||
|
|
"eval_samples_per_second": 163.197,
|
||
|
|
"eval_steps_per_second": 10.321,
|
||
|
|
"step": 634
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"grad_norm": 5.116003036499023,
|
||
|
|
"learning_rate": 0.00018890347400740728,
|
||
|
|
"loss": 1.3749292415775336,
|
||
|
|
"step": 1268
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_exact_match": 59.2,
|
||
|
|
"eval_f1": 73.91639681478482,
|
||
|
|
"eval_loss": 1.304753065109253,
|
||
|
|
"eval_runtime": 7.7166,
|
||
|
|
"eval_samples_per_second": 163.932,
|
||
|
|
"eval_steps_per_second": 10.367,
|
||
|
|
"step": 1268
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"grad_norm": 2.986513376235962,
|
||
|
|
"learning_rate": 0.00012596873805328422,
|
||
|
|
"loss": 1.0913604832598087,
|
||
|
|
"step": 1902
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"eval_exact_match": 61.52,
|
||
|
|
"eval_f1": 75.7146922946509,
|
||
|
|
"eval_loss": 1.2520850896835327,
|
||
|
|
"eval_runtime": 7.7623,
|
||
|
|
"eval_samples_per_second": 162.967,
|
||
|
|
"eval_steps_per_second": 10.306,
|
||
|
|
"step": 1902
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"grad_norm": 5.569353103637695,
|
||
|
|
"learning_rate": 6.303400209916113e-05,
|
||
|
|
"loss": 0.9191935927333892,
|
||
|
|
"step": 2536
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"eval_exact_match": 61.84,
|
||
|
|
"eval_f1": 76.26906549535173,
|
||
|
|
"eval_loss": 1.261074185371399,
|
||
|
|
"eval_runtime": 7.792,
|
||
|
|
"eval_samples_per_second": 162.347,
|
||
|
|
"eval_steps_per_second": 10.267,
|
||
|
|
"step": 2536
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"grad_norm": 6.102704048156738,
|
||
|
|
"learning_rate": 9.926614503804903e-08,
|
||
|
|
"loss": 0.8178048600157729,
|
||
|
|
"step": 3170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"eval_exact_match": 61.76,
|
||
|
|
"eval_f1": 76.43362369922768,
|
||
|
|
"eval_loss": 1.305261254310608,
|
||
|
|
"eval_runtime": 7.7595,
|
||
|
|
"eval_samples_per_second": 163.026,
|
||
|
|
"eval_steps_per_second": 10.31,
|
||
|
|
"step": 3170
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 500,
|
||
|
|
"max_steps": 3170,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1.545247865581056e+16,
|
||
|
|
"train_batch_size": 16,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": {
|
||
|
|
"learning_rate": 0.00028320631179355385,
|
||
|
|
"lora_alpha": 64,
|
||
|
|
"lora_dropout": 0.007091286682386649,
|
||
|
|
"lora_r": 256,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"weight_decay": 0.03343443262877938
|
||
|
|
}
|
||
|
|
}
|