trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6229148507118225,
  "best_model_checkpoint": "bert_base_km_10_v2_cola/checkpoint-34",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1589508056640625,
      "learning_rate": 4.9e-05,
      "loss": 0.6185,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6229148507118225,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6639,
      "eval_samples_per_second": 1570.963,
      "eval_steps_per_second": 7.531,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3882426023483276,
      "learning_rate": 4.8e-05,
      "loss": 0.593,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6797698736190796,
      "eval_loss": 0.6264082193374634,
      "eval_matthews_correlation": -0.019508732494421744,
      "eval_runtime": 0.689,
      "eval_samples_per_second": 1513.768,
      "eval_steps_per_second": 7.257,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.7106853723526,
      "learning_rate": 4.7e-05,
      "loss": 0.5528,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6788111329078674,
      "eval_loss": 0.629378616809845,
      "eval_matthews_correlation": 0.04020040476471409,
      "eval_runtime": 0.6665,
      "eval_samples_per_second": 1564.856,
      "eval_steps_per_second": 7.502,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.5940418243408203,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4973,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6529242396354675,
      "eval_loss": 0.6712858080863953,
      "eval_matthews_correlation": 0.06667066349742735,
      "eval_runtime": 0.6783,
      "eval_samples_per_second": 1537.714,
      "eval_steps_per_second": 7.372,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.508085250854492,
      "learning_rate": 4.5e-05,
      "loss": 0.4291,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6558005809783936,
      "eval_loss": 0.7182372808456421,
      "eval_matthews_correlation": 0.1191844484330856,
      "eval_runtime": 0.6784,
      "eval_samples_per_second": 1537.431,
      "eval_steps_per_second": 7.37,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.1077468395233154,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3516,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6088207364082336,
      "eval_loss": 0.8184990286827087,
      "eval_matthews_correlation": 0.0850925091452348,
      "eval_runtime": 0.7374,
      "eval_samples_per_second": 1414.377,
      "eval_steps_per_second": 6.78,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.5070589663935643,
      "train_runtime": 99.0295,
      "train_samples_per_second": 4317.4,
      "train_steps_per_second": 17.167
    }
  ],
  "logging_steps": 1,
  "max_steps": 1700,
  "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": 6749587903150080.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}