trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4366428852081299,
  "best_model_checkpoint": "bert_base_km_50_v1_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 7.707717418670654,
      "learning_rate": 4.9e-05,
      "loss": 0.4219,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.4366428852081299,
      "eval_runtime": 0.5638,
      "eval_samples_per_second": 1546.526,
      "eval_steps_per_second": 7.094,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.482503890991211,
      "learning_rate": 4.8e-05,
      "loss": 0.2286,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7947247706422018,
      "eval_loss": 0.48875945806503296,
      "eval_runtime": 0.565,
      "eval_samples_per_second": 1543.356,
      "eval_steps_per_second": 7.08,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 8.51332950592041,
      "learning_rate": 4.7e-05,
      "loss": 0.1615,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.622405469417572,
      "eval_runtime": 0.567,
      "eval_samples_per_second": 1537.964,
      "eval_steps_per_second": 7.055,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.031186580657959,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1168,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.6123579740524292,
      "eval_runtime": 0.5662,
      "eval_samples_per_second": 1540.044,
      "eval_steps_per_second": 7.064,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.508831977844238,
      "learning_rate": 4.5e-05,
      "loss": 0.0868,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7889908256880734,
      "eval_loss": 0.7037513852119446,
      "eval_runtime": 0.5784,
      "eval_samples_per_second": 1507.644,
      "eval_steps_per_second": 6.916,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.161043167114258,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0646,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7889908256880734,
      "eval_loss": 0.7526697516441345,
      "eval_runtime": 0.5701,
      "eval_samples_per_second": 1529.681,
      "eval_steps_per_second": 7.017,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.18004320125387172,
      "train_runtime": 663.1458,
      "train_samples_per_second": 5077.993,
      "train_steps_per_second": 19.905
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}