trainer_state.json

{
  "best_metric": 0.40745484828948975,
  "best_model_checkpoint": "distilbert_lda_50_v1_qnli/checkpoint-820",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2870,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 3.832184076309204,
      "learning_rate": 4.9e-05,
      "loss": 0.5532,
      "step": 410
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7627677100494233,
      "eval_loss": 0.5176936984062195,
      "eval_runtime": 2.289,
      "eval_samples_per_second": 2386.653,
      "eval_steps_per_second": 9.611,
      "step": 410
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.9373576641082764,
      "learning_rate": 4.8e-05,
      "loss": 0.4263,
      "step": 820
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8211605345048508,
      "eval_loss": 0.40745484828948975,
      "eval_runtime": 2.2919,
      "eval_samples_per_second": 2383.66,
      "eval_steps_per_second": 9.599,
      "step": 820
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.5357537269592285,
      "learning_rate": 4.7e-05,
      "loss": 0.3371,
      "step": 1230
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8132894014277869,
      "eval_loss": 0.4168780446052551,
      "eval_runtime": 2.3267,
      "eval_samples_per_second": 2347.919,
      "eval_steps_per_second": 9.455,
      "step": 1230
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.284133434295654,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.2505,
      "step": 1640
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8116419549697969,
      "eval_loss": 0.4765816330909729,
      "eval_runtime": 2.3197,
      "eval_samples_per_second": 2355.085,
      "eval_steps_per_second": 9.484,
      "step": 1640
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.09841251373291,
      "learning_rate": 4.5e-05,
      "loss": 0.18,
      "step": 2050
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7940691927512356,
      "eval_loss": 0.6080189943313599,
      "eval_runtime": 2.3468,
      "eval_samples_per_second": 2327.804,
      "eval_steps_per_second": 9.374,
      "step": 2050
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.979333400726318,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1275,
      "step": 2460
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8120080541826835,
      "eval_loss": 0.676863431930542,
      "eval_runtime": 2.307,
      "eval_samples_per_second": 2368.062,
      "eval_steps_per_second": 9.536,
      "step": 2460
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.858701467514038,
      "learning_rate": 4.3e-05,
      "loss": 0.0953,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7973640856672158,
      "eval_loss": 0.6988027095794678,
      "eval_runtime": 2.35,
      "eval_samples_per_second": 2324.65,
      "eval_steps_per_second": 9.362,
      "step": 2870
    },
    {
      "epoch": 7.0,
      "step": 2870,
      "total_flos": 4.856261458098893e+16,
      "train_loss": 0.2814387039025071,
      "train_runtime": 709.1788,
      "train_samples_per_second": 7384.809,
      "train_steps_per_second": 28.907
    }
  ],
  "logging_steps": 1,
  "max_steps": 20500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.856261458098893e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}