50 lines
1.0 KiB
JSON
50 lines
1.0 KiB
JSON
|
{
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.0,
|
||
|
|
"global_step": 2000,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.25,
|
||
|
|
"learning_rate": 3.7500000000000003e-05,
|
||
|
|
"loss": 0.8478,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5,
|
||
|
|
"learning_rate": 2.5e-05,
|
||
|
|
"loss": 0.5567,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.75,
|
||
|
|
"learning_rate": 1.25e-05,
|
||
|
|
"loss": 0.5195,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 0.5205,
|
||
|
|
"step": 2000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"step": 2000,
|
||
|
|
"total_flos": 289235175800832.0,
|
||
|
|
"train_loss": 0.6111455688476563,
|
||
|
|
"train_runtime": 3912.895,
|
||
|
|
"train_samples_per_second": 4.089,
|
||
|
|
"train_steps_per_second": 0.511
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"max_steps": 2000,
|
||
|
|
"num_train_epochs": 1,
|
||
|
|
"total_flos": 289235175800832.0,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|