run-6/checkpoint-300/trainer_state.json

{
  "best_metric": 0.7367906066536204,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-300",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.0887755155563354,
      "learning_rate": 0.0002701514784814016,
      "loss": 0.3405,
      "step": 50
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5440313111545988,
      "eval_f1": 0.21548821548821548,
      "eval_loss": 0.3118150532245636,
      "eval_precision": 0.7710843373493976,
      "eval_recall": 0.12524461839530332,
      "eval_runtime": 28.2208,
      "eval_samples_per_second": 36.214,
      "eval_steps_per_second": 1.134,
      "step": 50
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1933519840240479,
      "learning_rate": 0.00023155841012691565,
      "loss": 0.3081,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6154598825831703,
      "eval_f1": 0.43776824034334766,
      "eval_loss": 0.2966625690460205,
      "eval_precision": 0.8138297872340425,
      "eval_recall": 0.299412915851272,
      "eval_runtime": 28.3705,
      "eval_samples_per_second": 36.023,
      "eval_steps_per_second": 1.128,
      "step": 100
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.296247124671936,
      "learning_rate": 0.00019296534177242969,
      "loss": 0.2998,
      "step": 150
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6379647749510763,
      "eval_f1": 0.4917582417582418,
      "eval_loss": 0.29152899980545044,
      "eval_precision": 0.8248847926267281,
      "eval_recall": 0.350293542074364,
      "eval_runtime": 28.0611,
      "eval_samples_per_second": 36.421,
      "eval_steps_per_second": 1.14,
      "step": 150
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.9188393950462341,
      "learning_rate": 0.00015437227341794375,
      "loss": 0.2935,
      "step": 200
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5724070450097848,
      "eval_f1": 0.2962962962962963,
      "eval_loss": 0.29398053884506226,
      "eval_precision": 0.8363636363636363,
      "eval_recall": 0.18003913894324852,
      "eval_runtime": 27.6467,
      "eval_samples_per_second": 36.966,
      "eval_steps_per_second": 1.157,
      "step": 200
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5853505730628967,
      "learning_rate": 0.00011577920506345782,
      "loss": 0.291,
      "step": 250
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7270058708414873,
      "eval_f1": 0.6698224852071005,
      "eval_loss": 0.29069557785987854,
      "eval_precision": 0.8473053892215568,
      "eval_recall": 0.5538160469667319,
      "eval_runtime": 28.5484,
      "eval_samples_per_second": 35.799,
      "eval_steps_per_second": 1.121,
      "step": 250
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.0324815511703491,
      "learning_rate": 7.718613670897188e-05,
      "loss": 0.2884,
      "step": 300
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7367906066536204,
      "eval_f1": 0.6911595866819747,
      "eval_loss": 0.2900073230266571,
      "eval_precision": 0.8361111111111111,
      "eval_recall": 0.589041095890411,
      "eval_runtime": 28.0011,
      "eval_samples_per_second": 36.499,
      "eval_steps_per_second": 1.143,
      "step": 300
    }
  ],
  "logging_steps": 500,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 1414171183680.0,
  "train_batch_size": 62,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.43186153278710804,
    "learning_rate": 0.0003087445468358875,
    "num_train_epochs": 8,
    "per_device_train_batch_size": 62,
    "temperature": 10
  }
}