run-14/checkpoint-196/trainer_state.json

{
  "best_metric": 0.8228962818003914,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-14/checkpoint-196",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 196,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.3073023557662964,
      "learning_rate": 0.00021800841920007835,
      "loss": 0.5666,
      "step": 49
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7720156555772995,
      "eval_f1": 0.8010247651579846,
      "eval_loss": 0.48246321082115173,
      "eval_precision": 0.7106060606060606,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 28.2754,
      "eval_samples_per_second": 36.145,
      "eval_steps_per_second": 1.132,
      "step": 49
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.786695957183838,
      "learning_rate": 0.00016350631440005875,
      "loss": 0.4698,
      "step": 98
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8023483365949119,
      "eval_f1": 0.8212389380530973,
      "eval_loss": 0.436308890581131,
      "eval_precision": 0.7495961227786753,
      "eval_recall": 0.9080234833659491,
      "eval_runtime": 28.7331,
      "eval_samples_per_second": 35.569,
      "eval_steps_per_second": 1.114,
      "step": 98
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.9201297760009766,
      "learning_rate": 0.00010900420960003918,
      "loss": 0.4418,
      "step": 147
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8111545988258317,
      "eval_f1": 0.8197945845004668,
      "eval_loss": 0.41809919476509094,
      "eval_precision": 0.7839285714285714,
      "eval_recall": 0.8590998043052838,
      "eval_runtime": 27.7463,
      "eval_samples_per_second": 36.834,
      "eval_steps_per_second": 1.153,
      "step": 147
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.8454147577285767,
      "learning_rate": 5.450210480001959e-05,
      "loss": 0.4241,
      "step": 196
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8228962818003914,
      "eval_f1": 0.8340971585701191,
      "eval_loss": 0.41074350476264954,
      "eval_precision": 0.7844827586206896,
      "eval_recall": 0.8904109589041096,
      "eval_runtime": 27.5953,
      "eval_samples_per_second": 37.035,
      "eval_steps_per_second": 1.16,
      "step": 196
    }
  ],
  "logging_steps": 500,
  "max_steps": 245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 63,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8444228810359851,
    "learning_rate": 0.00027251052400009793,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 63,
    "temperature": 21
  }
}