roberta-base-wnli / trainer_state.json
Jeremiah Zhou
End of training
2eea97f
{
"best_metric": 0.5633802816901409,
"best_model_checkpoint": "./fine-tune/roberta-base/wnli/checkpoint-40",
"epoch": 10.0,
"global_step": 400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_accuracy": 0.5633802816901409,
"eval_loss": 0.6849406957626343,
"eval_runtime": 0.2643,
"eval_samples_per_second": 268.645,
"eval_steps_per_second": 34.054,
"step": 40
},
{
"epoch": 2.0,
"eval_accuracy": 0.5633802816901409,
"eval_loss": 0.6912119388580322,
"eval_runtime": 0.259,
"eval_samples_per_second": 274.159,
"eval_steps_per_second": 34.753,
"step": 80
},
{
"epoch": 3.0,
"eval_accuracy": 0.5633802816901409,
"eval_loss": 0.6918124556541443,
"eval_runtime": 0.3165,
"eval_samples_per_second": 224.361,
"eval_steps_per_second": 28.44,
"step": 120
},
{
"epoch": 4.0,
"eval_accuracy": 0.43661971830985913,
"eval_loss": 0.6964250802993774,
"eval_runtime": 0.2736,
"eval_samples_per_second": 259.529,
"eval_steps_per_second": 32.898,
"step": 160
},
{
"epoch": 5.0,
"eval_accuracy": 0.5633802816901409,
"eval_loss": 0.692846953868866,
"eval_runtime": 0.2609,
"eval_samples_per_second": 272.168,
"eval_steps_per_second": 34.5,
"step": 200
},
{
"epoch": 6.0,
"eval_accuracy": 0.43661971830985913,
"eval_loss": 0.7005417346954346,
"eval_runtime": 0.2776,
"eval_samples_per_second": 255.727,
"eval_steps_per_second": 32.416,
"step": 240
},
{
"epoch": 7.0,
"eval_accuracy": 0.30985915492957744,
"eval_loss": 0.6964187622070312,
"eval_runtime": 0.2676,
"eval_samples_per_second": 265.355,
"eval_steps_per_second": 33.637,
"step": 280
},
{
"epoch": 8.0,
"eval_accuracy": 0.352112676056338,
"eval_loss": 0.6986197233200073,
"eval_runtime": 0.2698,
"eval_samples_per_second": 263.149,
"eval_steps_per_second": 33.357,
"step": 320
},
{
"epoch": 9.0,
"eval_accuracy": 0.5492957746478874,
"eval_loss": 0.6969391107559204,
"eval_runtime": 0.263,
"eval_samples_per_second": 270.009,
"eval_steps_per_second": 34.226,
"step": 360
},
{
"epoch": 10.0,
"eval_accuracy": 0.5633802816901409,
"eval_loss": 0.697629451751709,
"eval_runtime": 0.2692,
"eval_samples_per_second": 263.768,
"eval_steps_per_second": 33.435,
"step": 400
},
{
"epoch": 10.0,
"step": 400,
"total_flos": 417688800384000.0,
"train_loss": 0.6946794891357422,
"train_runtime": 122.3783,
"train_samples_per_second": 51.888,
"train_steps_per_second": 3.269
}
],
"max_steps": 400,
"num_train_epochs": 10,
"total_flos": 417688800384000.0,
"trial_name": null,
"trial_params": null
}