run-0/checkpoint-594/trainer_state.json

{
  "best_metric": 0.5108910891089109,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-297",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 594,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.3744499087333679,
      "learning_rate": 1.7212091585793122e-05,
      "loss": 0.1576,
      "step": 297
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5108910891089109,
      "eval_f1": 0.127208480565371,
      "eval_loss": 0.13107798993587494,
      "eval_mcc": 0.04175421326407947,
      "eval_precision": 0.5806451612903226,
      "eval_recall": 0.07142857142857142,
      "eval_runtime": 0.9746,
      "eval_samples_per_second": 518.155,
      "eval_steps_per_second": 16.417,
      "step": 297
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2569155693054199,
      "learning_rate": 1.5060580137568982e-05,
      "loss": 0.1312,
      "step": 594
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5089108910891089,
      "eval_f1": 0.1205673758865248,
      "eval_loss": 0.12648388743400574,
      "eval_mcc": 0.03400602754322123,
      "eval_precision": 0.5666666666666667,
      "eval_recall": 0.06746031746031746,
      "eval_runtime": 0.9581,
      "eval_samples_per_second": 527.09,
      "eval_steps_per_second": 16.7,
      "step": 594
    }
  ],
  "logging_steps": 500,
  "max_steps": 2673,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 1461402531960.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.14400624333141754,
    "learning_rate": 1.9363603034017264e-05,
    "num_train_epochs": 9,
    "temperature": 24
  }
}