bert-base-uncased-stsb / trainer_state.json
Jeremiah Zhou
End of training
a64de90
raw
history blame
1.78 kB
{
"best_metric": 0.8843169724798454,
"best_model_checkpoint": "./fine-tune/bert-base-uncased/stsb/checkpoint-540",
"epoch": 3.0,
"global_step": 540,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_combined_score": 0.8770512336585177,
"eval_loss": 0.517871081829071,
"eval_pearson": 0.8805737478242902,
"eval_runtime": 6.6982,
"eval_samples_per_second": 223.94,
"eval_spearmanr": 0.873528719492745,
"eval_steps_per_second": 28.067,
"step": 180
},
{
"epoch": 2.0,
"eval_combined_score": 0.8834893774364951,
"eval_loss": 0.5145099759101868,
"eval_pearson": 0.8850209566170038,
"eval_runtime": 6.9391,
"eval_samples_per_second": 216.167,
"eval_spearmanr": 0.8819577982559864,
"eval_steps_per_second": 27.093,
"step": 360
},
{
"epoch": 2.78,
"learning_rate": 1.4814814814814815e-06,
"loss": 0.7868,
"step": 500
},
{
"epoch": 3.0,
"eval_combined_score": 0.8859105179324802,
"eval_loss": 0.5143781900405884,
"eval_pearson": 0.887504063385115,
"eval_runtime": 7.1241,
"eval_samples_per_second": 210.554,
"eval_spearmanr": 0.8843169724798454,
"eval_steps_per_second": 26.389,
"step": 540
},
{
"epoch": 3.0,
"step": 540,
"total_flos": 1134458907008256.0,
"train_loss": 0.7539703951941596,
"train_runtime": 279.0486,
"train_samples_per_second": 61.806,
"train_steps_per_second": 1.935
}
],
"max_steps": 540,
"num_train_epochs": 3,
"total_flos": 1134458907008256.0,
"trial_name": null,
"trial_params": null
}