bart-base-detox / checkpoint-1080 /trainer_state.json
ChrisZeng's picture
Training in progress, epoch 7
5489eef
raw
history blame
2.91 kB
{
"best_metric": 0.1867838054895401,
"best_model_checkpoint": "outputs/bart-base-detox/checkpoint-1080",
"epoch": 7.9953917050691246,
"global_step": 1080,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"learning_rate": 9e-06,
"loss": 0.5633,
"step": 135
},
{
"epoch": 1.0,
"eval_loss": 0.25235414505004883,
"eval_runtime": 12.2006,
"eval_samples_per_second": 254.085,
"eval_steps_per_second": 31.802,
"step": 135
},
{
"epoch": 2.0,
"learning_rate": 8.000000000000001e-06,
"loss": 0.2589,
"step": 270
},
{
"epoch": 2.0,
"eval_loss": 0.21927106380462646,
"eval_runtime": 11.2792,
"eval_samples_per_second": 274.842,
"eval_steps_per_second": 34.4,
"step": 270
},
{
"epoch": 3.0,
"learning_rate": 7e-06,
"loss": 0.2307,
"step": 405
},
{
"epoch": 3.0,
"eval_loss": 0.19929151237010956,
"eval_runtime": 11.2313,
"eval_samples_per_second": 276.016,
"eval_steps_per_second": 34.546,
"step": 405
},
{
"epoch": 4.0,
"learning_rate": 6e-06,
"loss": 0.2171,
"step": 540
},
{
"epoch": 4.0,
"eval_loss": 0.20016591250896454,
"eval_runtime": 11.0923,
"eval_samples_per_second": 279.473,
"eval_steps_per_second": 34.979,
"step": 540
},
{
"epoch": 5.0,
"learning_rate": 5e-06,
"loss": 0.2027,
"step": 675
},
{
"epoch": 5.0,
"eval_loss": 0.19366399943828583,
"eval_runtime": 11.2232,
"eval_samples_per_second": 276.213,
"eval_steps_per_second": 34.571,
"step": 675
},
{
"epoch": 6.0,
"learning_rate": 4.000000000000001e-06,
"loss": 0.1946,
"step": 810
},
{
"epoch": 6.0,
"eval_loss": 0.19715721905231476,
"eval_runtime": 11.2739,
"eval_samples_per_second": 274.973,
"eval_steps_per_second": 34.416,
"step": 810
},
{
"epoch": 7.0,
"learning_rate": 3e-06,
"loss": 0.1874,
"step": 945
},
{
"epoch": 7.0,
"eval_loss": 0.19172348082065582,
"eval_runtime": 11.1804,
"eval_samples_per_second": 277.27,
"eval_steps_per_second": 34.704,
"step": 945
},
{
"epoch": 8.0,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.1853,
"step": 1080
},
{
"epoch": 8.0,
"eval_loss": 0.1867838054895401,
"eval_runtime": 11.2543,
"eval_samples_per_second": 275.451,
"eval_steps_per_second": 34.476,
"step": 1080
}
],
"max_steps": 1350,
"num_train_epochs": 10,
"total_flos": 4299589980241920.0,
"trial_name": null,
"trial_params": null
}