trainer_state.json

{
  "best_metric": 0.6942489743232727,
  "best_model_checkpoint": "add_BERT_no_pretrain_rte/checkpoint-20",
  "epoch": 6.0,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.7731,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.6942489743232727,
      "eval_runtime": 0.5221,
      "eval_samples_per_second": 530.559,
      "eval_steps_per_second": 5.746,
      "step": 20
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.709,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.71894770860672,
      "eval_runtime": 0.5216,
      "eval_samples_per_second": 531.047,
      "eval_steps_per_second": 5.751,
      "step": 40
    },
    {
      "epoch": 3.0,
      "learning_rate": 3.76e-05,
      "loss": 0.7188,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.6948172450065613,
      "eval_runtime": 0.5214,
      "eval_samples_per_second": 531.271,
      "eval_steps_per_second": 5.754,
      "step": 60
    },
    {
      "epoch": 4.0,
      "learning_rate": 3.680000000000001e-05,
      "loss": 0.7007,
      "step": 80
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.6980084180831909,
      "eval_runtime": 0.5256,
      "eval_samples_per_second": 527.056,
      "eval_steps_per_second": 5.708,
      "step": 80
    },
    {
      "epoch": 5.0,
      "learning_rate": 3.6e-05,
      "loss": 0.7048,
      "step": 100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.7017844319343567,
      "eval_runtime": 0.5223,
      "eval_samples_per_second": 530.325,
      "eval_steps_per_second": 5.744,
      "step": 100
    },
    {
      "epoch": 6.0,
      "learning_rate": 3.52e-05,
      "loss": 0.7065,
      "step": 120
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.726894199848175,
      "eval_runtime": 0.5221,
      "eval_samples_per_second": 530.593,
      "eval_steps_per_second": 5.746,
      "step": 120
    },
    {
      "epoch": 6.0,
      "step": 120,
      "total_flos": 2133790084300800.0,
      "train_loss": 0.7188253561655681,
      "train_runtime": 120.124,
      "train_samples_per_second": 1036.429,
      "train_steps_per_second": 8.325
    }
  ],
  "max_steps": 1000,
  "num_train_epochs": 50,
  "total_flos": 2133790084300800.0,
  "trial_name": null,
  "trial_params": null
}