run-7/checkpoint-960/trainer_state.json

{
  "best_metric": 0.7915851272015656,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-7/checkpoint-768",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 960,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.15159010887146,
      "learning_rate": 2.5124348703312046e-05,
      "loss": 0.6503,
      "step": 192
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6908023483365949,
      "eval_f1": 0.6623931623931624,
      "eval_loss": 0.6152858734130859,
      "eval_mcc": 0.38712657099283027,
      "eval_precision": 0.7294117647058823,
      "eval_recall": 0.6066536203522505,
      "eval_runtime": 66.1991,
      "eval_samples_per_second": 15.438,
      "eval_steps_per_second": 0.483,
      "step": 192
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.791621208190918,
      "learning_rate": 1.884326152748403e-05,
      "loss": 0.5656,
      "step": 384
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.761252446183953,
      "eval_f1": 0.7761467889908258,
      "eval_loss": 0.5115423202514648,
      "eval_mcc": 0.5271935842180233,
      "eval_precision": 0.7305699481865285,
      "eval_recall": 0.8277886497064579,
      "eval_runtime": 66.4101,
      "eval_samples_per_second": 15.389,
      "eval_steps_per_second": 0.482,
      "step": 384
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.52553653717041,
      "learning_rate": 1.2562174351656023e-05,
      "loss": 0.512,
      "step": 576
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.786692759295499,
      "eval_f1": 0.8080985915492958,
      "eval_loss": 0.48333269357681274,
      "eval_mcc": 0.5882099387037968,
      "eval_precision": 0.7344,
      "eval_recall": 0.898238747553816,
      "eval_runtime": 66.6309,
      "eval_samples_per_second": 15.338,
      "eval_steps_per_second": 0.48,
      "step": 576
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.329979419708252,
      "learning_rate": 6.281087175828012e-06,
      "loss": 0.494,
      "step": 768
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7915851272015656,
      "eval_f1": 0.8136482939632547,
      "eval_loss": 0.474974662065506,
      "eval_mcc": 0.6002406744389225,
      "eval_precision": 0.7357594936708861,
      "eval_recall": 0.9099804305283757,
      "eval_runtime": 66.491,
      "eval_samples_per_second": 15.37,
      "eval_steps_per_second": 0.481,
      "step": 768
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.863868713378906,
      "learning_rate": 0.0,
      "loss": 0.4871,
      "step": 960
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7710371819960861,
      "eval_f1": 0.802030456852792,
      "eval_loss": 0.4800633490085602,
      "eval_mcc": 0.5707750105508812,
      "eval_precision": 0.706408345752608,
      "eval_recall": 0.9275929549902152,
      "eval_runtime": 66.5281,
      "eval_samples_per_second": 15.362,
      "eval_steps_per_second": 0.481,
      "step": 960
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9315469293191833,
    "learning_rate": 3.1405435879140055e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 16,
    "temperature": 39
  }
}