run-1/checkpoint-384/trainer_state.json

{
  "best_metric": 0.7925636007827789,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-192",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 384,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8947777152061462,
      "learning_rate": 0.0005181908296523575,
      "loss": 0.4457,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7925636007827789,
      "eval_f1": 0.8090090090090091,
      "eval_loss": 0.41997030377388,
      "eval_mcc": 0.59400157079309,
      "eval_precision": 0.7495826377295493,
      "eval_recall": 0.8786692759295499,
      "eval_runtime": 66.1643,
      "eval_samples_per_second": 15.446,
      "eval_steps_per_second": 0.484,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.971238613128662,
      "learning_rate": 0.0004606140708020956,
      "loss": 0.4163,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7915851272015656,
      "eval_f1": 0.7957813998082454,
      "eval_loss": 0.39958736300468445,
      "eval_mcc": 0.5836633290639656,
      "eval_precision": 0.7800751879699248,
      "eval_recall": 0.812133072407045,
      "eval_runtime": 66.5614,
      "eval_samples_per_second": 15.354,
      "eval_steps_per_second": 0.481,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 1920,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.6664545915670365,
    "learning_rate": 0.0005757675885026195,
    "num_train_epochs": 10,
    "per_device_train_batch_size": 16,
    "temperature": 37
  }
}