93 lines
2.3 KiB
JSON
93 lines
2.3 KiB
JSON
{
|
|
"best_global_step": 1902,
|
|
"best_metric": 42.797774942292016,
|
|
"best_model_checkpoint": "mbert-loraxs-qa-vanilla/run-1/checkpoint-1902",
|
|
"epoch": 3.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1902,
|
|
"is_hyper_param_search": true,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 1.0,
|
|
"grad_norm": 50.64515686035156,
|
|
"learning_rate": 0.00010486219815639711,
|
|
"loss": 4.513181668350749,
|
|
"step": 634
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_exact_match": 17.76,
|
|
"eval_f1": 28.945909302213973,
|
|
"eval_loss": 3.11417818069458,
|
|
"eval_runtime": 7.6977,
|
|
"eval_samples_per_second": 164.336,
|
|
"eval_steps_per_second": 10.393,
|
|
"step": 634
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"grad_norm": 69.94309997558594,
|
|
"learning_rate": 5.247241594114434e-05,
|
|
"loss": 3.1071321024127565,
|
|
"step": 1268
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_exact_match": 29.52,
|
|
"eval_f1": 40.87090707056966,
|
|
"eval_loss": 2.5117993354797363,
|
|
"eval_runtime": 7.7422,
|
|
"eval_samples_per_second": 163.389,
|
|
"eval_steps_per_second": 10.333,
|
|
"step": 1268
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"grad_norm": 85.58645629882812,
|
|
"learning_rate": 8.263372589156588e-08,
|
|
"loss": 2.8506941449378944,
|
|
"step": 1902
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_exact_match": 30.88,
|
|
"eval_f1": 42.797774942292016,
|
|
"eval_loss": 2.423711061477661,
|
|
"eval_runtime": 7.6134,
|
|
"eval_samples_per_second": 166.154,
|
|
"eval_steps_per_second": 10.508,
|
|
"step": 1902
|
|
}
|
|
],
|
|
"logging_steps": 500,
|
|
"max_steps": 1902,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"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": 6056721303671808.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": {
|
|
"learning_rate": 0.00014138630500046923,
|
|
"lora_alpha": 128,
|
|
"lora_dropout": 0.10300849900123686,
|
|
"lora_r": 8,
|
|
"num_train_epochs": 3,
|
|
"weight_decay": 0.04651389479910262
|
|
}
|
|
}
|