110 lines
2.7 KiB
JSON
110 lines
2.7 KiB
JSON
|
{
|
||
|
|
"best_global_step": 2536,
|
||
|
|
"best_metric": 76.5481177817324,
|
||
|
|
"best_model_checkpoint": "mbert-loraxs-qa-vanilla/run-24/checkpoint-2536",
|
||
|
|
"epoch": 4.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 2536,
|
||
|
|
"is_hyper_param_search": true,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"grad_norm": 5.415954113006592,
|
||
|
|
"learning_rate": 0.00031164479502205044,
|
||
|
|
"loss": 2.9678479504510054,
|
||
|
|
"step": 634
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_exact_match": 56.16,
|
||
|
|
"eval_f1": 71.44176954968724,
|
||
|
|
"eval_loss": 1.397819995880127,
|
||
|
|
"eval_runtime": 7.8224,
|
||
|
|
"eval_samples_per_second": 161.715,
|
||
|
|
"eval_steps_per_second": 10.227,
|
||
|
|
"step": 634
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"grad_norm": 5.330461502075195,
|
||
|
|
"learning_rate": 0.00023376430623845565,
|
||
|
|
"loss": 1.3548693732132295,
|
||
|
|
"step": 1268
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_exact_match": 59.92,
|
||
|
|
"eval_f1": 74.39548380172305,
|
||
|
|
"eval_loss": 1.3065598011016846,
|
||
|
|
"eval_runtime": 7.8066,
|
||
|
|
"eval_samples_per_second": 162.043,
|
||
|
|
"eval_steps_per_second": 10.248,
|
||
|
|
"step": 1268
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"grad_norm": 3.078212261199951,
|
||
|
|
"learning_rate": 0.00015588381745486088,
|
||
|
|
"loss": 1.0759374576412164,
|
||
|
|
"step": 1902
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"eval_exact_match": 62.24,
|
||
|
|
"eval_f1": 76.12391951339926,
|
||
|
|
"eval_loss": 1.238013505935669,
|
||
|
|
"eval_runtime": 7.8257,
|
||
|
|
"eval_samples_per_second": 161.646,
|
||
|
|
"eval_steps_per_second": 10.223,
|
||
|
|
"step": 1902
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"grad_norm": 5.6887898445129395,
|
||
|
|
"learning_rate": 7.800332867126607e-05,
|
||
|
|
"loss": 0.8924907118740142,
|
||
|
|
"step": 2536
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"eval_exact_match": 62.24,
|
||
|
|
"eval_f1": 76.5481177817324,
|
||
|
|
"eval_loss": 1.2567627429962158,
|
||
|
|
"eval_runtime": 7.8214,
|
||
|
|
"eval_samples_per_second": 161.735,
|
||
|
|
"eval_steps_per_second": 10.228,
|
||
|
|
"step": 2536
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"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": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1.2361982924648448e+16,
|
||
|
|
"train_batch_size": 16,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": {
|
||
|
|
"learning_rate": 0.00035046219952617657,
|
||
|
|
"lora_alpha": 64,
|
||
|
|
"lora_dropout": 0.12144321544438236,
|
||
|
|
"lora_r": 256,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"weight_decay": 0.022438856746700774
|
||
|
|
}
|
||
|
|
}
|