run-20/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7346221441124781,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-428",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 428,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.5011236667633057,
      "learning_rate": 0.00029874954647137283,
      "loss": 0.4553,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6848271821909784,
      "eval_f1": 0.1567398119122257,
      "eval_loss": 0.4299508333206177,
      "eval_mcc": 0.17036819686661997,
      "eval_precision": 0.7246376811594203,
      "eval_recall": 0.08787346221441125,
      "eval_runtime": 3.1578,
      "eval_samples_per_second": 540.559,
      "eval_steps_per_second": 17.1,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.671453595161438,
      "learning_rate": 0.0002489579553928107,
      "loss": 0.4196,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7346221441124781,
      "eval_f1": 0.4372670807453416,
      "eval_loss": 0.4135269522666931,
      "eval_mcc": 0.3504334189243189,
      "eval_precision": 0.7457627118644068,
      "eval_recall": 0.3093145869947276,
      "eval_runtime": 3.2195,
      "eval_samples_per_second": 530.204,
      "eval_steps_per_second": 16.773,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 1498,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7240883100721838,
    "learning_rate": 0.000348541137549935,
    "num_train_epochs": 7,
    "temperature": 12
  }
}