run-0/checkpoint-384/trainer_state.json

{
  "best_metric": 0.8003913894324853,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-384",
  "epoch": 4.0,
  "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": 1.6740443706512451,
      "learning_rate": 0.0001218354408608861,
      "loss": 0.4816,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7514677103718199,
      "eval_f1": 0.7519531250000001,
      "eval_loss": 0.41538161039352417,
      "eval_precision": 0.7504873294346979,
      "eval_recall": 0.7534246575342466,
      "eval_runtime": 26.5175,
      "eval_samples_per_second": 38.541,
      "eval_steps_per_second": 1.207,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7866544723510742,
      "learning_rate": 9.137658064566457e-05,
      "loss": 0.4182,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7798434442270059,
      "eval_f1": 0.7817652764306499,
      "eval_loss": 0.3980446457862854,
      "eval_precision": 0.775,
      "eval_recall": 0.7886497064579256,
      "eval_runtime": 25.5088,
      "eval_samples_per_second": 40.065,
      "eval_steps_per_second": 1.254,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1238555908203125,
      "learning_rate": 6.091772043044305e-05,
      "loss": 0.4044,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.8205574912891985,
      "eval_loss": 0.39847832918167114,
      "eval_precision": 0.7394034536891679,
      "eval_recall": 0.9217221135029354,
      "eval_runtime": 25.6502,
      "eval_samples_per_second": 39.844,
      "eval_steps_per_second": 1.248,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.6613142490386963,
      "learning_rate": 3.0458860215221525e-05,
      "loss": 0.3971,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8003913894324853,
      "eval_f1": 0.799212598425197,
      "eval_loss": 0.38756656646728516,
      "eval_precision": 0.803960396039604,
      "eval_recall": 0.7945205479452054,
      "eval_runtime": 26.1487,
      "eval_samples_per_second": 39.084,
      "eval_steps_per_second": 1.224,
      "step": 384
    }
  ],
  "logging_steps": 500,
  "max_steps": 480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 942780789120.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.679174768290245,
    "learning_rate": 0.00015229430107610762,
    "num_train_epochs": 5,
    "temperature": 27
  }
}