trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6225768327713013,
  "best_model_checkpoint": "bert_base_km_20_v1_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": 1.3155686855316162,
      "learning_rate": 4.9e-05,
      "loss": 0.6182,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6225768327713013,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.6878,
      "eval_samples_per_second": 1516.325,
      "eval_steps_per_second": 7.269,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.696873426437378,
      "learning_rate": 4.8e-05,
      "loss": 0.591,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6893576383590698,
      "eval_loss": 0.6297030448913574,
      "eval_matthews_correlation": 0.07656947466775693,
      "eval_runtime": 0.675,
      "eval_samples_per_second": 1545.239,
      "eval_steps_per_second": 7.408,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9348201751708984,
      "learning_rate": 4.7e-05,
      "loss": 0.5432,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6797698736190796,
      "eval_loss": 0.6300162076950073,
      "eval_matthews_correlation": 0.08088406701239864,
      "eval_runtime": 0.6962,
      "eval_samples_per_second": 1498.188,
      "eval_steps_per_second": 7.182,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.5529346466064453,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4606,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6519654989242554,
      "eval_loss": 0.6926795244216919,
      "eval_matthews_correlation": 0.14975092131840292,
      "eval_runtime": 0.6643,
      "eval_samples_per_second": 1570.163,
      "eval_steps_per_second": 7.527,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.714582681655884,
      "learning_rate": 4.5e-05,
      "loss": 0.3815,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6567593216896057,
      "eval_loss": 0.7553606033325195,
      "eval_matthews_correlation": 0.09193331749307261,
      "eval_runtime": 0.6983,
      "eval_samples_per_second": 1493.613,
      "eval_steps_per_second": 7.16,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.6072070598602295,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3147,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6366251111030579,
      "eval_loss": 0.8696136474609375,
      "eval_matthews_correlation": 0.09631786888423775,
      "eval_runtime": 0.6804,
      "eval_samples_per_second": 1532.842,
      "eval_steps_per_second": 7.348,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 6749587903150080.0,
      "train_loss": 0.484881274840411,
      "train_runtime": 101.9729,
      "train_samples_per_second": 4192.78,
      "train_steps_per_second": 16.671
    }
  ],
  "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
}