{
  "best_metric": 0.8307240704500979,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-9/checkpoint-5362",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 6894,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.658814907073975,
      "learning_rate": 5.312164802258909e-05,
      "loss": 0.5144,
      "step": 766
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7847358121330724,
      "eval_f1": 0.8053097345132744,
      "eval_loss": 0.4449388384819031,
      "eval_precision": 0.7350565428109854,
      "eval_recall": 0.8904109589041096,
      "eval_runtime": 131.9263,
      "eval_samples_per_second": 7.747,
      "eval_steps_per_second": 1.94,
      "step": 766
    },
    {
      "epoch": 2.0,
      "grad_norm": 21.609981536865234,
      "learning_rate": 4.648144201976546e-05,
      "loss": 0.4571,
      "step": 1532
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.815068493150685,
      "eval_f1": 0.8298829882988299,
      "eval_loss": 0.4288274347782135,
      "eval_precision": 0.7683333333333333,
      "eval_recall": 0.9021526418786693,
      "eval_runtime": 132.7828,
      "eval_samples_per_second": 7.697,
      "eval_steps_per_second": 1.928,
      "step": 1532
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.596930027008057,
      "learning_rate": 3.984123601694182e-05,
      "loss": 0.444,
      "step": 2298
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.812133072407045,
      "eval_f1": 0.8218923933209648,
      "eval_loss": 0.4297088086605072,
      "eval_precision": 0.781305114638448,
      "eval_recall": 0.8669275929549902,
      "eval_runtime": 133.7052,
      "eval_samples_per_second": 7.644,
      "eval_steps_per_second": 1.915,
      "step": 2298
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.766616344451904,
      "learning_rate": 3.320103001411819e-05,
      "loss": 0.4335,
      "step": 3064
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8160469667318982,
      "eval_f1": 0.8278388278388279,
      "eval_loss": 0.43032753467559814,
      "eval_precision": 0.7779690189328744,
      "eval_recall": 0.8845401174168297,
      "eval_runtime": 132.4524,
      "eval_samples_per_second": 7.716,
      "eval_steps_per_second": 1.933,
      "step": 3064
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.048282623291016,
      "learning_rate": 2.6560824011294545e-05,
      "loss": 0.4137,
      "step": 3830
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.824853228962818,
      "eval_f1": 0.8350230414746544,
      "eval_loss": 0.4249579608440399,
      "eval_precision": 0.789198606271777,
      "eval_recall": 0.8864970645792564,
      "eval_runtime": 130.8699,
      "eval_samples_per_second": 7.809,
      "eval_steps_per_second": 1.956,
      "step": 3830
    },
    {
      "epoch": 6.0,
      "grad_norm": 19.333023071289062,
      "learning_rate": 1.992061800847091e-05,
      "loss": 0.4157,
      "step": 4596
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8180039138943248,
      "eval_f1": 0.8351063829787234,
      "eval_loss": 0.43369239568710327,
      "eval_precision": 0.7633711507293355,
      "eval_recall": 0.9217221135029354,
      "eval_runtime": 131.2119,
      "eval_samples_per_second": 7.789,
      "eval_steps_per_second": 1.951,
      "step": 4596
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.5701804161071777,
      "learning_rate": 1.3280412005647272e-05,
      "loss": 0.4096,
      "step": 5362
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8307240704500979,
      "eval_f1": 0.8464951197870452,
      "eval_loss": 0.4228073060512543,
      "eval_precision": 0.7743506493506493,
      "eval_recall": 0.9334637964774951,
      "eval_runtime": 131.6336,
      "eval_samples_per_second": 7.764,
      "eval_steps_per_second": 1.945,
      "step": 5362
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.1991465091705322,
      "learning_rate": 6.640206002823636e-06,
      "loss": 0.4081,
      "step": 6128
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8307240704500979,
      "eval_f1": 0.8451208594449418,
      "eval_loss": 0.4199325144290924,
      "eval_precision": 0.7788778877887789,
      "eval_recall": 0.923679060665362,
      "eval_runtime": 132.6978,
      "eval_samples_per_second": 7.702,
      "eval_steps_per_second": 1.929,
      "step": 6128
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.31607940793037415,
      "learning_rate": 0.0,
      "loss": 0.4011,
      "step": 6894
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8307240704500979,
      "eval_f1": 0.8422971741112123,
      "eval_loss": 0.4193461537361145,
      "eval_precision": 0.78839590443686,
      "eval_recall": 0.9041095890410958,
      "eval_runtime": 133.6999,
      "eval_samples_per_second": 7.644,
      "eval_steps_per_second": 1.915,
      "step": 6894
    }
  ],
  "logging_steps": 500,
  "max_steps": 6894,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "total_flos": 2121256775520.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.7939954008608587,
    "learning_rate": 5.976185402541273e-05,
    "num_train_epochs": 9,
    "temperature": 18
  }
}