run-45/checkpoint-856/trainer_state.json

{
  "best_metric": 0.7357937902753369,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-45/checkpoint-856",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 856,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.950324058532715,
      "learning_rate": 0.00010478179948708485,
      "loss": 0.5816,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7041593438781488,
      "eval_f1": 0.5063538611925709,
      "eval_loss": 0.5396628975868225,
      "eval_mcc": 0.30281965656473075,
      "eval_precision": 0.5704845814977973,
      "eval_recall": 0.45518453427065025,
      "eval_runtime": 3.1674,
      "eval_samples_per_second": 538.919,
      "eval_steps_per_second": 17.048,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.071056604385376,
      "learning_rate": 9.168407455119925e-05,
      "loss": 0.5316,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.705916813122437,
      "eval_f1": 0.4434589800443459,
      "eval_loss": 0.5229035019874573,
      "eval_mcc": 0.27911345997194636,
      "eval_precision": 0.6006006006006006,
      "eval_recall": 0.351493848857645,
      "eval_runtime": 3.1873,
      "eval_samples_per_second": 535.562,
      "eval_steps_per_second": 16.942,
      "step": 428
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.054250478744507,
      "learning_rate": 7.858634961531364e-05,
      "loss": 0.509,
      "step": 642
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7217340363210311,
      "eval_f1": 0.43384982121573307,
      "eval_loss": 0.509781539440155,
      "eval_mcc": 0.31331689795359174,
      "eval_precision": 0.674074074074074,
      "eval_recall": 0.31985940246045697,
      "eval_runtime": 3.1209,
      "eval_samples_per_second": 546.962,
      "eval_steps_per_second": 17.303,
      "step": 642
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.881683349609375,
      "learning_rate": 6.548862467942804e-05,
      "loss": 0.496,
      "step": 856
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7357937902753369,
      "eval_f1": 0.507103825136612,
      "eval_loss": 0.49531644582748413,
      "eval_mcc": 0.3606499983342828,
      "eval_precision": 0.6705202312138728,
      "eval_recall": 0.4077328646748682,
      "eval_runtime": 3.9891,
      "eval_samples_per_second": 427.912,
      "eval_steps_per_second": 13.537,
      "step": 856
    }
  ],
  "logging_steps": 500,
  "max_steps": 1926,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 2099102657760.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9457260837583084,
    "learning_rate": 0.00011787952442297046,
    "num_train_epochs": 9,
    "temperature": 19
  }
}