trainer_state.json

{
  "best_metric": 0.5403890609741211,
  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_data_aug_rte/checkpoint-2272",
  "epoch": 7.0,
  "global_step": 7952,
  "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": 0.2735,
      "step": 1136
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4657039711191336,
      "eval_loss": 0.5752855539321899,
      "eval_runtime": 0.7236,
      "eval_samples_per_second": 382.814,
      "eval_steps_per_second": 4.146,
      "step": 1136
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.8e-05,
      "loss": 0.2182,
      "step": 2272
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4620938628158845,
      "eval_loss": 0.5403890609741211,
      "eval_runtime": 0.7138,
      "eval_samples_per_second": 388.063,
      "eval_steps_per_second": 4.203,
      "step": 2272
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.7e-05,
      "loss": 0.2119,
      "step": 3408
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.47653429602888087,
      "eval_loss": 0.5687434077262878,
      "eval_runtime": 0.716,
      "eval_samples_per_second": 386.892,
      "eval_steps_per_second": 4.19,
      "step": 3408
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.2089,
      "step": 4544
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48375451263537905,
      "eval_loss": 0.5697432160377502,
      "eval_runtime": 0.7118,
      "eval_samples_per_second": 389.139,
      "eval_steps_per_second": 4.214,
      "step": 4544
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.5e-05,
      "loss": 0.2072,
      "step": 5680
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.48014440433212996,
      "eval_loss": 0.5590192079544067,
      "eval_runtime": 0.7136,
      "eval_samples_per_second": 388.179,
      "eval_steps_per_second": 4.204,
      "step": 5680
    },
    {
      "epoch": 6.0,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2057,
      "step": 6816
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.48375451263537905,
      "eval_loss": 0.5585924983024597,
      "eval_runtime": 0.7129,
      "eval_samples_per_second": 388.565,
      "eval_steps_per_second": 4.208,
      "step": 6816
    },
    {
      "epoch": 7.0,
      "learning_rate": 4.3e-05,
      "loss": 0.2047,
      "step": 7952
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.48014440433212996,
      "eval_loss": 0.5555509924888611,
      "eval_runtime": 0.7091,
      "eval_samples_per_second": 390.621,
      "eval_steps_per_second": 4.231,
      "step": 7952
    },
    {
      "epoch": 7.0,
      "step": 7952,
      "total_flos": 3.1911928058085376e+16,
      "train_loss": 0.21857677858840052,
      "train_runtime": 6161.1259,
      "train_samples_per_second": 1179.963,
      "train_steps_per_second": 9.219
    }
  ],
  "max_steps": 56800,
  "num_train_epochs": 50,
  "total_flos": 3.1911928058085376e+16,
  "trial_name": null,
  "trial_params": null
}