trainer_state.json

{
  "best_metric": NaN,
  "best_model_checkpoint": "mobilebert_add_GLUE_Experiment_logit_kd_pretrain_qnli/checkpoint-819",
  "epoch": 6.0,
  "global_step": 4914,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 4.9e-05,
      "loss": 0.0,
      "step": 819
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4418,
      "eval_samples_per_second": 578.599,
      "eval_steps_per_second": 4.554,
      "step": 819
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.8e-05,
      "loss": 0.0,
      "step": 1638
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4458,
      "eval_samples_per_second": 578.352,
      "eval_steps_per_second": 4.552,
      "step": 1638
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.7e-05,
      "loss": 0.0,
      "step": 2457
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4513,
      "eval_samples_per_second": 578.017,
      "eval_steps_per_second": 4.55,
      "step": 2457
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.0,
      "step": 3276
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4759,
      "eval_samples_per_second": 576.517,
      "eval_steps_per_second": 4.538,
      "step": 3276
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.5e-05,
      "loss": 0.0,
      "step": 4095
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4604,
      "eval_samples_per_second": 577.457,
      "eval_steps_per_second": 4.545,
      "step": 4095
    },
    {
      "epoch": 6.0,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0,
      "step": 4914
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4946000366099213,
      "eval_loss": NaN,
      "eval_runtime": 9.4869,
      "eval_samples_per_second": 575.847,
      "eval_steps_per_second": 4.533,
      "step": 4914
    },
    {
      "epoch": 6.0,
      "step": 4914,
      "total_flos": 1.884356181019853e+16,
      "train_loss": 0.0,
      "train_runtime": 2688.3146,
      "train_samples_per_second": 1948.116,
      "train_steps_per_second": 15.233
    }
  ],
  "max_steps": 40950,
  "num_train_epochs": 50,
  "total_flos": 1.884356181019853e+16,
  "trial_name": null,
  "trial_params": null
}