run-4/checkpoint-320/trainer_state.json

{
  "best_metric": 0.6674509803921569,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-4/checkpoint-160",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 320,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.30153390765190125,
      "learning_rate": 0.0001806546260556293,
      "loss": 0.1164,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6674509803921569,
      "eval_f1": 0.03636363636363636,
      "eval_loss": 0.10561967641115189,
      "eval_mcc": 0.04629100498862757,
      "eval_precision": 0.5333333333333333,
      "eval_recall": 0.018823529411764704,
      "eval_runtime": 1.857,
      "eval_samples_per_second": 686.581,
      "eval_steps_per_second": 21.54,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2800346910953522,
      "learning_rate": 0.00014452370084450344,
      "loss": 0.1067,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6658823529411765,
      "eval_f1": 0.049107142857142856,
      "eval_loss": 0.10315733402967453,
      "eval_mcc": 0.04166956048662825,
      "eval_precision": 0.4782608695652174,
      "eval_recall": 0.02588235294117647,
      "eval_runtime": 1.8766,
      "eval_samples_per_second": 679.417,
      "eval_steps_per_second": 21.315,
      "step": 320
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 583510875840.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.1455909339678303,
    "learning_rate": 0.00021678555126675516,
    "num_train_epochs": 6,
    "temperature": 19
  }
}