trainer_state.json

{
  "best_metric": 1.2641148567199707,
  "best_model_checkpoint": "mobilebert_add_GLUE_Experiment_logit_kd_sst2_256/checkpoint-3162",
  "epoch": 11.0,
  "global_step": 5797,
  "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": 1.5438,
      "step": 527
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5814220183486238,
      "eval_loss": 1.4011807441711426,
      "eval_runtime": 1.9244,
      "eval_samples_per_second": 453.117,
      "eval_steps_per_second": 3.637,
      "step": 527
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.8e-05,
      "loss": 1.364,
      "step": 1054
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5412844036697247,
      "eval_loss": 1.5474371910095215,
      "eval_runtime": 1.9252,
      "eval_samples_per_second": 452.943,
      "eval_steps_per_second": 3.636,
      "step": 1054
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.7e-05,
      "loss": 1.2907,
      "step": 1581
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5642201834862385,
      "eval_loss": 1.5137693881988525,
      "eval_runtime": 1.9325,
      "eval_samples_per_second": 451.237,
      "eval_steps_per_second": 3.622,
      "step": 1581
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.257,
      "step": 2108
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5665137614678899,
      "eval_loss": 1.4408754110336304,
      "eval_runtime": 1.9137,
      "eval_samples_per_second": 455.672,
      "eval_steps_per_second": 3.658,
      "step": 2108
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.5e-05,
      "loss": 1.2417,
      "step": 2635
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5928899082568807,
      "eval_loss": 1.4473283290863037,
      "eval_runtime": 1.9221,
      "eval_samples_per_second": 453.677,
      "eval_steps_per_second": 3.642,
      "step": 2635
    },
    {
      "epoch": 6.0,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.2056,
      "step": 3162
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7075688073394495,
      "eval_loss": 1.2641148567199707,
      "eval_runtime": 2.067,
      "eval_samples_per_second": 421.868,
      "eval_steps_per_second": 3.387,
      "step": 3162
    },
    {
      "epoch": 7.0,
      "learning_rate": 4.3e-05,
      "loss": 0.6274,
      "step": 3689
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": NaN,
      "eval_runtime": 1.9207,
      "eval_samples_per_second": 453.996,
      "eval_steps_per_second": 3.644,
      "step": 3689
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.2e-05,
      "loss": 0.0,
      "step": 4216
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": NaN,
      "eval_runtime": 1.9183,
      "eval_samples_per_second": 454.57,
      "eval_steps_per_second": 3.649,
      "step": 4216
    },
    {
      "epoch": 9.0,
      "learning_rate": 4.1e-05,
      "loss": 0.0,
      "step": 4743
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": NaN,
      "eval_runtime": 1.917,
      "eval_samples_per_second": 454.883,
      "eval_steps_per_second": 3.652,
      "step": 4743
    },
    {
      "epoch": 10.0,
      "learning_rate": 4e-05,
      "loss": 0.0,
      "step": 5270
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": NaN,
      "eval_runtime": 1.9509,
      "eval_samples_per_second": 446.978,
      "eval_steps_per_second": 3.588,
      "step": 5270
    },
    {
      "epoch": 11.0,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0,
      "step": 5797
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": NaN,
      "eval_runtime": 1.9254,
      "eval_samples_per_second": 452.883,
      "eval_steps_per_second": 3.636,
      "step": 5797
    },
    {
      "epoch": 11.0,
      "step": 5797,
      "total_flos": 1.9095284072054784e+16,
      "train_loss": 0.7754758815426487,
      "train_runtime": 4341.2127,
      "train_samples_per_second": 775.693,
      "train_steps_per_second": 6.07
    }
  ],
  "max_steps": 26350,
  "num_train_epochs": 50,
  "total_flos": 1.9095284072054784e+16,
  "trial_name": null,
  "trial_params": null
}