run-20/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8964715005816207,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-1290",
  "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": 1.6259269714355469,
      "learning_rate": 0.0003431069228198171,
      "loss": 0.3055,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8889104303993796,
      "eval_f1": 0.19409282700421943,
      "eval_loss": 0.2940736413002014,
      "eval_mcc": 0.20172457395808635,
      "eval_precision": 0.4859154929577465,
      "eval_recall": 0.12126537785588752,
      "eval_runtime": 9.9362,
      "eval_samples_per_second": 519.111,
      "eval_steps_per_second": 16.304,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.638690710067749,
      "learning_rate": 0.00028592243568318097,
      "loss": 0.2894,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8964715005816207,
      "eval_f1": 0.18844984802431614,
      "eval_loss": 0.2831971049308777,
      "eval_mcc": 0.24798827118495087,
      "eval_precision": 0.6966292134831461,
      "eval_recall": 0.10896309314586995,
      "eval_runtime": 9.9132,
      "eval_samples_per_second": 520.316,
      "eval_steps_per_second": 16.342,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 4515,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8743805278290786,
    "learning_rate": 0.0004002914099564533,
    "num_train_epochs": 7,
    "temperature": 27
  }
}