run-2/checkpoint-180/trainer_state.json

{
  "best_metric": 0.8215994541006149,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-stsb/run-2/checkpoint-180",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 180,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 2.4969136714935303,
      "eval_pearson": 0.32223422122729983,
      "eval_runtime": 0.9052,
      "eval_samples_per_second": 1657.071,
      "eval_spearmanr": 0.32647520010278785,
      "eval_steps_per_second": 103.843,
      "step": 36
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.1902464628219604,
      "eval_pearson": 0.774068032504475,
      "eval_runtime": 0.9073,
      "eval_samples_per_second": 1653.256,
      "eval_spearmanr": 0.7685396079860547,
      "eval_steps_per_second": 103.604,
      "step": 72
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.845661461353302,
      "eval_pearson": 0.8156384301074189,
      "eval_runtime": 0.9602,
      "eval_samples_per_second": 1562.214,
      "eval_spearmanr": 0.8142363416044488,
      "eval_steps_per_second": 97.899,
      "step": 108
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.7359773516654968,
      "eval_pearson": 0.820309116112808,
      "eval_runtime": 0.9488,
      "eval_samples_per_second": 1580.967,
      "eval_spearmanr": 0.8207010328870485,
      "eval_steps_per_second": 99.074,
      "step": 144
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.7456313371658325,
      "eval_pearson": 0.8215994541006149,
      "eval_runtime": 0.9399,
      "eval_samples_per_second": 1595.857,
      "eval_spearmanr": 0.8210025491088271,
      "eval_steps_per_second": 100.007,
      "step": 180
    }
  ],
  "logging_steps": 500,
  "max_steps": 180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 2.5099285971844188e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 16,
    "seed": 8
  }
}