run-20/checkpoint-1054/trainer_state.json

{
  "best_metric": 0.9036697247706422,
  "best_model_checkpoint": "electra-distilled-sst\\run-20\\checkpoint-1054",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1054,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 44.26585006713867,
      "learning_rate": 6.064402229159703e-05,
      "loss": 1.6431,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8944954128440367,
      "eval_loss": 1.0662537813186646,
      "eval_runtime": 1.4489,
      "eval_samples_per_second": 601.833,
      "eval_steps_per_second": 4.831,
      "step": 527
    },
    {
      "epoch": 2.0,
      "grad_norm": 44.72014617919922,
      "learning_rate": 5.054307420446946e-05,
      "loss": 0.778,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9036697247706422,
      "eval_loss": 1.0714185237884521,
      "eval_runtime": 1.4519,
      "eval_samples_per_second": 600.589,
      "eval_steps_per_second": 4.821,
      "step": 1054
    }
  ],
  "logging_steps": 500,
  "max_steps": 3689,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 373776249469380.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.5844023105776549,
    "learning_rate": 7.0706636609893e-05,
    "num_train_epochs": 7,
    "temperature": 25
  }
}