{
  "best_metric": 0.8434442270058709,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-864",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 864,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.464465618133545,
      "learning_rate": 0.0001794635836369996,
      "loss": 0.565,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7788649706457925,
      "eval_f1": 0.8078231292517007,
      "eval_loss": 0.4691426157951355,
      "eval_precision": 0.7142857142857143,
      "eval_recall": 0.9295499021526419,
      "eval_runtime": 25.0337,
      "eval_samples_per_second": 40.825,
      "eval_steps_per_second": 1.278,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.053184509277344,
      "learning_rate": 0.00015703063568237466,
      "loss": 0.4676,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7984344422700587,
      "eval_f1": 0.8170515097690941,
      "eval_loss": 0.4367915391921997,
      "eval_precision": 0.7479674796747967,
      "eval_recall": 0.9001956947162426,
      "eval_runtime": 25.2941,
      "eval_samples_per_second": 40.405,
      "eval_steps_per_second": 1.265,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.9524621963500977,
      "learning_rate": 0.00013459768772774972,
      "loss": 0.4358,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8209393346379648,
      "eval_f1": 0.8334849863512284,
      "eval_loss": 0.4263817071914673,
      "eval_precision": 0.7789115646258503,
      "eval_recall": 0.8962818003913894,
      "eval_runtime": 24.9367,
      "eval_samples_per_second": 40.984,
      "eval_steps_per_second": 1.283,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.636496067047119,
      "learning_rate": 0.00011216473977312478,
      "loss": 0.4174,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8160469667318982,
      "eval_f1": 0.8181818181818182,
      "eval_loss": 0.41675594449043274,
      "eval_precision": 0.8087954110898662,
      "eval_recall": 0.8277886497064579,
      "eval_runtime": 24.9785,
      "eval_samples_per_second": 40.915,
      "eval_steps_per_second": 1.281,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.0662448406219482,
      "learning_rate": 8.97317918184998e-05,
      "loss": 0.4077,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8356164383561644,
      "eval_f1": 0.8502673796791443,
      "eval_loss": 0.40537795424461365,
      "eval_precision": 0.7806873977086743,
      "eval_recall": 0.9334637964774951,
      "eval_runtime": 25.8814,
      "eval_samples_per_second": 39.488,
      "eval_steps_per_second": 1.236,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.637610912322998,
      "learning_rate": 6.729884386387486e-05,
      "loss": 0.4004,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8258317025440313,
      "eval_f1": 0.842756183745583,
      "eval_loss": 0.40995147824287415,
      "eval_precision": 0.7681159420289855,
      "eval_recall": 0.9334637964774951,
      "eval_runtime": 25.8126,
      "eval_samples_per_second": 39.593,
      "eval_steps_per_second": 1.24,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.4253664016723633,
      "learning_rate": 4.48658959092499e-05,
      "loss": 0.391,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8326810176125244,
      "eval_f1": 0.8488063660477454,
      "eval_loss": 0.40309804677963257,
      "eval_precision": 0.7741935483870968,
      "eval_recall": 0.9393346379647749,
      "eval_runtime": 26.0268,
      "eval_samples_per_second": 39.267,
      "eval_steps_per_second": 1.23,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.7185893058776855,
      "learning_rate": 2.243294795462495e-05,
      "loss": 0.3869,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8395303326810176,
      "eval_f1": 0.851985559566787,
      "eval_loss": 0.39594146609306335,
      "eval_precision": 0.7906197654941374,
      "eval_recall": 0.923679060665362,
      "eval_runtime": 25.1887,
      "eval_samples_per_second": 40.574,
      "eval_steps_per_second": 1.27,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.016538619995117,
      "learning_rate": 0.0,
      "loss": 0.3828,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8434442270058709,
      "eval_f1": 0.8542805100182149,
      "eval_loss": 0.39343148469924927,
      "eval_precision": 0.7989778534923339,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 25.5945,
      "eval_samples_per_second": 39.931,
      "eval_steps_per_second": 1.25,
      "step": 864
    }
  ],
  "logging_steps": 500,
  "max_steps": 864,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 2121256775520.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.8956145079239534,
    "learning_rate": 0.00020189653159162458,
    "num_train_epochs": 9,
    "temperature": 8
  }
}