{
  "best_metric": 0.8326810176125244,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-32/checkpoint-672",
  "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": 1.7263739109039307,
      "learning_rate": 0.00011867778606566929,
      "loss": 0.5622,
      "step": 96
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7710371819960861,
      "eval_f1": 0.8006814310051107,
      "eval_loss": 0.4762067198753357,
      "eval_precision": 0.7088989441930619,
      "eval_recall": 0.9197651663405088,
      "eval_runtime": 25.5548,
      "eval_samples_per_second": 39.993,
      "eval_steps_per_second": 1.252,
      "step": 96
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.8613439798355103,
      "learning_rate": 0.00010384306280746064,
      "loss": 0.4724,
      "step": 192
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8072407045009785,
      "eval_f1": 0.8273444347063978,
      "eval_loss": 0.43931350111961365,
      "eval_precision": 0.7492063492063492,
      "eval_recall": 0.923679060665362,
      "eval_runtime": 25.0105,
      "eval_samples_per_second": 40.863,
      "eval_steps_per_second": 1.279,
      "step": 192
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.853325366973877,
      "learning_rate": 8.900833954925197e-05,
      "loss": 0.4412,
      "step": 288
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8258317025440313,
      "eval_f1": 0.8363970588235294,
      "eval_loss": 0.42110058665275574,
      "eval_precision": 0.7885615251299827,
      "eval_recall": 0.8904109589041096,
      "eval_runtime": 25.7114,
      "eval_samples_per_second": 39.749,
      "eval_steps_per_second": 1.245,
      "step": 288
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.9698286056518555,
      "learning_rate": 7.417361629104331e-05,
      "loss": 0.4294,
      "step": 384
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8160469667318982,
      "eval_f1": 0.8192307692307692,
      "eval_loss": 0.41999757289886475,
      "eval_precision": 0.8052930056710775,
      "eval_recall": 0.8336594911937377,
      "eval_runtime": 25.0795,
      "eval_samples_per_second": 40.75,
      "eval_steps_per_second": 1.276,
      "step": 384
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.9139764308929443,
      "learning_rate": 5.9338893032834646e-05,
      "loss": 0.4153,
      "step": 480
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8297455968688845,
      "eval_f1": 0.843806104129264,
      "eval_loss": 0.4074234664440155,
      "eval_precision": 0.7794361525704809,
      "eval_recall": 0.9197651663405088,
      "eval_runtime": 25.2807,
      "eval_samples_per_second": 40.426,
      "eval_steps_per_second": 1.266,
      "step": 480
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.3249385356903076,
      "learning_rate": 4.4504169774625984e-05,
      "loss": 0.4067,
      "step": 576
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.824853228962818,
      "eval_f1": 0.8397493285586393,
      "eval_loss": 0.4091223180294037,
      "eval_precision": 0.7739273927392739,
      "eval_recall": 0.9178082191780822,
      "eval_runtime": 25.2618,
      "eval_samples_per_second": 40.456,
      "eval_steps_per_second": 1.267,
      "step": 576
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.5627498626708984,
      "learning_rate": 2.9669446516417323e-05,
      "loss": 0.3993,
      "step": 672
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8326810176125244,
      "eval_f1": 0.848,
      "eval_loss": 0.4094063341617584,
      "eval_precision": 0.7768729641693811,
      "eval_recall": 0.9334637964774951,
      "eval_runtime": 26.0578,
      "eval_samples_per_second": 39.221,
      "eval_steps_per_second": 1.228,
      "step": 672
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.446059465408325,
      "learning_rate": 1.4834723258208661e-05,
      "loss": 0.4005,
      "step": 768
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8287671232876712,
      "eval_f1": 0.8424842484248425,
      "eval_loss": 0.4037325978279114,
      "eval_precision": 0.78,
      "eval_recall": 0.9158512720156555,
      "eval_runtime": 25.5767,
      "eval_samples_per_second": 39.958,
      "eval_steps_per_second": 1.251,
      "step": 768
    },
    {
      "epoch": 9.0,
      "grad_norm": 2.3571696281433105,
      "learning_rate": 0.0,
      "loss": 0.3936,
      "step": 864
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8326810176125244,
      "eval_f1": 0.8423963133640553,
      "eval_loss": 0.40162914991378784,
      "eval_precision": 0.7961672473867596,
      "eval_recall": 0.8943248532289628,
      "eval_runtime": 25.5021,
      "eval_samples_per_second": 40.075,
      "eval_steps_per_second": 1.255,
      "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": 0.8475925399213161,
    "learning_rate": 0.00013351250932387796,
    "num_train_epochs": 9,
    "temperature": 10
  }
}