run-4/checkpoint-1384/trainer_state.json

{
  "best_metric": 0.6359803825635763,
  "best_model_checkpoint": "test-klue/ynat/run-4/checkpoint-692",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_f1": 0.6349053963006231,
      "eval_loss": 0.8050844669342041,
      "eval_runtime": 9.4523,
      "eval_samples_per_second": 250.626,
      "eval_steps_per_second": 2.01,
      "step": 346
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.5458921984247287e-05,
      "loss": 0.7983,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_f1": 0.6359803825635763,
      "eval_loss": 0.820810079574585,
      "eval_runtime": 9.5072,
      "eval_samples_per_second": 249.181,
      "eval_steps_per_second": 1.998,
      "step": 692
    },
    {
      "epoch": 2.89,
      "learning_rate": 6.7151878302612646e-06,
      "loss": 0.5122,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_f1": 0.6358373590219538,
      "eval_loss": 0.9183709025382996,
      "eval_runtime": 9.5449,
      "eval_samples_per_second": 248.195,
      "eval_steps_per_second": 1.991,
      "step": 1038
    },
    {
      "epoch": 4.0,
      "eval_f1": 0.6342325467636821,
      "eval_loss": 1.1191256046295166,
      "eval_runtime": 9.5524,
      "eval_samples_per_second": 248.002,
      "eval_steps_per_second": 1.989,
      "step": 1384
    }
  ],
  "logging_steps": 500,
  "max_steps": 1384,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 673157179094904.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 2.402778145515359e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "seed": 12
  }
}