{
  "best_metric": 0.8189823874755382,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-16/checkpoint-3830",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 3830,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.586636066436768,
      "learning_rate": 0.0002815362185318002,
      "loss": 0.5012,
      "step": 766
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7847358121330724,
      "eval_f1": 0.8090277777777778,
      "eval_loss": 0.4424646198749542,
      "eval_precision": 0.7269890795631825,
      "eval_recall": 0.9119373776908023,
      "eval_runtime": 131.6873,
      "eval_samples_per_second": 7.761,
      "eval_steps_per_second": 1.944,
      "step": 766
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.197028160095215,
      "learning_rate": 0.00021115216389885013,
      "loss": 0.4492,
      "step": 1532
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.799412915851272,
      "eval_f1": 0.8064211520302172,
      "eval_loss": 0.440277099609375,
      "eval_precision": 0.7791970802919708,
      "eval_recall": 0.8356164383561644,
      "eval_runtime": 131.1398,
      "eval_samples_per_second": 7.793,
      "eval_steps_per_second": 1.952,
      "step": 1532
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.803006172180176,
      "learning_rate": 0.0001407681092659001,
      "loss": 0.4291,
      "step": 2298
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.812133072407045,
      "eval_f1": 0.8218923933209648,
      "eval_loss": 0.4386424720287323,
      "eval_precision": 0.781305114638448,
      "eval_recall": 0.8669275929549902,
      "eval_runtime": 130.5843,
      "eval_samples_per_second": 7.826,
      "eval_steps_per_second": 1.96,
      "step": 2298
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.590163230895996,
      "learning_rate": 7.038405463295006e-05,
      "loss": 0.4241,
      "step": 3064
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.815068493150685,
      "eval_f1": 0.8177434908389585,
      "eval_loss": 0.4371667206287384,
      "eval_precision": 0.8060836501901141,
      "eval_recall": 0.8297455968688845,
      "eval_runtime": 130.7093,
      "eval_samples_per_second": 7.819,
      "eval_steps_per_second": 1.959,
      "step": 3064
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.153400421142578,
      "learning_rate": 0.0,
      "loss": 0.4008,
      "step": 3830
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8189823874755382,
      "eval_f1": 0.8310502283105022,
      "eval_loss": 0.4148101210594177,
      "eval_precision": 0.7791095890410958,
      "eval_recall": 0.8904109589041096,
      "eval_runtime": 131.3758,
      "eval_samples_per_second": 7.779,
      "eval_steps_per_second": 1.949,
      "step": 3830
    }
  ],
  "logging_steps": 500,
  "max_steps": 3830,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1178475986400.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7716344444380391,
    "learning_rate": 0.00035192027316475024,
    "num_train_epochs": 5,
    "temperature": 10
  }
}