run-6/checkpoint-800/trainer_state.json

{
  "best_metric": 0.6666666666666666,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-6/checkpoint-160",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 800,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.8079733848571777,
      "learning_rate": 5.379067903396322e-05,
      "loss": 0.0738,
      "step": 160
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0,
      "eval_loss": 0.05636660009622574,
      "eval_mcc": 0.0,
      "eval_precision": 0.0,
      "eval_recall": 0.0,
      "eval_runtime": 1.8691,
      "eval_samples_per_second": 682.132,
      "eval_steps_per_second": 21.4,
      "step": 160
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5784248113632202,
      "learning_rate": 4.3032543227170566e-05,
      "loss": 0.0565,
      "step": 320
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.040632054176072234,
      "eval_loss": 0.05314622446894646,
      "eval_mcc": 0.04230810954948887,
      "eval_precision": 0.5,
      "eval_recall": 0.021176470588235293,
      "eval_runtime": 1.867,
      "eval_samples_per_second": 682.92,
      "eval_steps_per_second": 21.425,
      "step": 320
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.1806308478116989,
      "learning_rate": 3.227440742037793e-05,
      "loss": 0.0547,
      "step": 480
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1": 0.0700218818380744,
      "eval_loss": 0.051966577768325806,
      "eval_mcc": 0.05672760277973996,
      "eval_precision": 0.5,
      "eval_recall": 0.03764705882352941,
      "eval_runtime": 1.877,
      "eval_samples_per_second": 679.267,
      "eval_steps_per_second": 21.31,
      "step": 480
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3304363191127777,
      "learning_rate": 2.1516271613585283e-05,
      "loss": 0.0539,
      "step": 640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6643137254901961,
      "eval_f1": 0.06140350877192983,
      "eval_loss": 0.05162987858057022,
      "eval_mcc": 0.03960834081045404,
      "eval_precision": 0.45161290322580644,
      "eval_recall": 0.03294117647058824,
      "eval_runtime": 1.8749,
      "eval_samples_per_second": 680.029,
      "eval_steps_per_second": 21.334,
      "step": 640
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.38088977336883545,
      "learning_rate": 1.0758135806792642e-05,
      "loss": 0.0536,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6658823529411765,
      "eval_f1": 0.0616740088105727,
      "eval_loss": 0.05147842690348625,
      "eval_mcc": 0.04835822308072435,
      "eval_precision": 0.4827586206896552,
      "eval_recall": 0.03294117647058824,
      "eval_runtime": 1.8714,
      "eval_samples_per_second": 681.301,
      "eval_steps_per_second": 21.374,
      "step": 800
    }
  ],
  "logging_steps": 500,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "total_flos": 1458777189600.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.062310076275025716,
    "learning_rate": 6.454881484075586e-05,
    "num_train_epochs": 6,
    "temperature": 27
  }
}