run-10/checkpoint-96/trainer_state.json

{
  "best_metric": 0.7847358121330724,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-10/checkpoint-96",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 96,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.5607351064682007,
      "learning_rate": 0.0008549093212842401,
      "loss": 0.5996,
      "step": 48
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7632093933463796,
      "eval_f1": 0.7949152542372881,
      "eval_loss": 0.502604067325592,
      "eval_precision": 0.7010463378176383,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 27.7883,
      "eval_samples_per_second": 36.778,
      "eval_steps_per_second": 0.576,
      "step": 48
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.883798122406006,
      "learning_rate": 0.0007124244344035335,
      "loss": 0.4919,
      "step": 96
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7847358121330724,
      "eval_f1": 0.8066783831282953,
      "eval_loss": 0.4441834092140198,
      "eval_precision": 0.7320574162679426,
      "eval_recall": 0.898238747553816,
      "eval_runtime": 27.0257,
      "eval_samples_per_second": 37.816,
      "eval_steps_per_second": 0.592,
      "step": 96
    }
  ],
  "logging_steps": 500,
  "max_steps": 336,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9873851004059778,
    "learning_rate": 0.0009973942081649468,
    "num_train_epochs": 7,
    "temperature": 16
  }
}