trainer_state.json

{
  "best_metric": 0.388401061296463,
  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_pretrain_rte/checkpoint-60",
  "epoch": 8.0,
  "global_step": 160,
  "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": 0.4107,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.3951400816440582,
      "eval_runtime": 0.5432,
      "eval_samples_per_second": 509.897,
      "eval_steps_per_second": 5.522,
      "step": 20
    },
    {
      "epoch": 2.0,
      "learning_rate": 4.8e-05,
      "loss": 0.3757,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.3914218842983246,
      "eval_runtime": 0.5449,
      "eval_samples_per_second": 508.396,
      "eval_steps_per_second": 5.506,
      "step": 40
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.7e-05,
      "loss": 0.347,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.388401061296463,
      "eval_runtime": 0.5432,
      "eval_samples_per_second": 509.919,
      "eval_steps_per_second": 5.523,
      "step": 60
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3072,
      "step": 80
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.402240514755249,
      "eval_runtime": 0.5439,
      "eval_samples_per_second": 509.256,
      "eval_steps_per_second": 5.515,
      "step": 80
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.5e-05,
      "loss": 0.2762,
      "step": 100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.41157519817352295,
      "eval_runtime": 0.542,
      "eval_samples_per_second": 511.088,
      "eval_steps_per_second": 5.535,
      "step": 100
    },
    {
      "epoch": 6.0,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2457,
      "step": 120
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.40726301074028015,
      "eval_runtime": 0.5446,
      "eval_samples_per_second": 508.611,
      "eval_steps_per_second": 5.508,
      "step": 120
    },
    {
      "epoch": 7.0,
      "learning_rate": 4.3e-05,
      "loss": 0.2215,
      "step": 140
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.4115126132965088,
      "eval_runtime": 0.5433,
      "eval_samples_per_second": 509.804,
      "eval_steps_per_second": 5.521,
      "step": 140
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.2e-05,
      "loss": 0.2059,
      "step": 160
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.4230565130710602,
      "eval_runtime": 0.5423,
      "eval_samples_per_second": 510.767,
      "eval_steps_per_second": 5.532,
      "step": 160
    },
    {
      "epoch": 8.0,
      "step": 160,
      "total_flos": 632613383438336.0,
      "train_loss": 0.29872527420520784,
      "train_runtime": 143.5614,
      "train_samples_per_second": 867.225,
      "train_steps_per_second": 6.966
    }
  ],
  "max_steps": 1000,
  "num_train_epochs": 50,
  "total_flos": 632613383438336.0,
  "trial_name": null,
  "trial_params": null
}