run-19/checkpoint-428/trainer_state.json

{
  "best_metric": 0.7258347978910369,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-19/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": 7.411643028259277,
      "learning_rate": 9.671414990548834e-05,
      "loss": 0.5212,
      "step": 214
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7076742823667252,
      "eval_f1": 0.4651661307609861,
      "eval_loss": 0.48319703340530396,
      "eval_mcc": 0.2902539914884967,
      "eval_precision": 0.5961538461538461,
      "eval_recall": 0.38137082601054484,
      "eval_runtime": 3.1282,
      "eval_samples_per_second": 545.683,
      "eval_steps_per_second": 17.262,
      "step": 214
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.417043924331665,
      "learning_rate": 0.0,
      "loss": 0.4749,
      "step": 428
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7258347978910369,
      "eval_f1": 0.5185185185185186,
      "eval_loss": 0.469474196434021,
      "eval_mcc": 0.3443248673024818,
      "eval_precision": 0.6253101736972705,
      "eval_recall": 0.4428822495606327,
      "eval_runtime": 3.1499,
      "eval_samples_per_second": 541.924,
      "eval_steps_per_second": 17.143,
      "step": 428
    }
  ],
  "logging_steps": 500,
  "max_steps": 428,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 1049551328880.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8445790915524267,
    "learning_rate": 0.00019342829981097668,
    "num_train_epochs": 2,
    "temperature": 17
  }
}