run-4/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8900736719658783,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-645",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1290,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.9960685968399048,
      "learning_rate": 0.0005204222579442675,
      "loss": 0.3335,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8900736719658783,
      "eval_f1": 0.0070052539404553416,
      "eval_loss": 0.3329962193965912,
      "eval_mcc": 0.05593214040002362,
      "eval_precision": 1.0,
      "eval_recall": 0.0035149384885764497,
      "eval_runtime": 9.9674,
      "eval_samples_per_second": 517.487,
      "eval_steps_per_second": 16.253,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6790906190872192,
      "learning_rate": 0.0004625975626171267,
      "loss": 0.3255,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8883288096161303,
      "eval_f1": 0.16034985422740525,
      "eval_loss": 0.3358159363269806,
      "eval_mcc": 0.17495541464580086,
      "eval_precision": 0.4700854700854701,
      "eval_recall": 0.09666080843585237,
      "eval_runtime": 9.9229,
      "eval_samples_per_second": 519.808,
      "eval_steps_per_second": 16.326,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 6450,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.976928179899354,
    "learning_rate": 0.0005782469532714083,
    "num_train_epochs": 10,
    "temperature": 24
  }
}