run-40/checkpoint-384/trainer_state.json

{
  "best_metric": 0.8091976516634051,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-40/checkpoint-384",
  "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": 7.570804595947266,
      "learning_rate": 9.127982348433665e-05,
      "loss": 0.5774,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8033268101761253,
      "eval_f1": 0.8222811671087534,
      "eval_loss": 0.45489490032196045,
      "eval_mcc": 0.6209445541298213,
      "eval_precision": 0.75,
      "eval_recall": 0.9099804305283757,
      "eval_runtime": 67.0612,
      "eval_samples_per_second": 15.24,
      "eval_steps_per_second": 0.477,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.614089012145996,
      "learning_rate": 6.0853215656224434e-05,
      "loss": 0.4642,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8091976516634051,
      "eval_f1": 0.8202764976958525,
      "eval_loss": 0.4278040826320648,
      "eval_mcc": 0.6231493257080627,
      "eval_precision": 0.7752613240418118,
      "eval_recall": 0.8708414872798435,
      "eval_runtime": 66.8566,
      "eval_samples_per_second": 15.286,
      "eval_steps_per_second": 0.479,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 768,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 471390394560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9602502972445941,
    "learning_rate": 0.00012170643131244887,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "temperature": 29
  }
}