trainer_state.json

{
  "best_global_step": 68,
  "best_metric": 0.6238862872123718,
  "best_model_checkpoint": "distilbert_km_50_v2_cola/checkpoint-68",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 238,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1716675758361816,
      "learning_rate": 4.9e-05,
      "loss": 0.6213,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6252191066741943,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4786,
      "eval_samples_per_second": 2179.47,
      "eval_steps_per_second": 10.448,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2596843242645264,
      "learning_rate": 4.8e-05,
      "loss": 0.6057,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6238862872123718,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4632,
      "eval_samples_per_second": 2251.748,
      "eval_steps_per_second": 10.795,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.412390112876892,
      "learning_rate": 4.7e-05,
      "loss": 0.5869,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6845637559890747,
      "eval_loss": 0.6300022006034851,
      "eval_matthews_correlation": 0.017393698817017886,
      "eval_runtime": 0.454,
      "eval_samples_per_second": 2297.283,
      "eval_steps_per_second": 11.013,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0707762241363525,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5567,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6692233681678772,
      "eval_loss": 0.6484770774841309,
      "eval_matthews_correlation": 0.07351652355856254,
      "eval_runtime": 0.4888,
      "eval_samples_per_second": 2133.674,
      "eval_steps_per_second": 10.229,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.1753809452056885,
      "learning_rate": 4.5e-05,
      "loss": 0.5056,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6577181220054626,
      "eval_loss": 0.6887763738632202,
      "eval_matthews_correlation": 0.055491890520777944,
      "eval_runtime": 0.4435,
      "eval_samples_per_second": 2351.553,
      "eval_steps_per_second": 11.273,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.5411596298217773,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4515,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.641418993473053,
      "eval_loss": 0.7534975409507751,
      "eval_matthews_correlation": 0.011521571495439718,
      "eval_runtime": 0.4661,
      "eval_samples_per_second": 2237.777,
      "eval_steps_per_second": 10.728,
      "step": 204
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.098875045776367,
      "learning_rate": 4.3e-05,
      "loss": 0.3961,
      "step": 238
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6366251111030579,
      "eval_loss": 0.8222112059593201,
      "eval_matthews_correlation": 0.021546575667040293,
      "eval_runtime": 0.4561,
      "eval_samples_per_second": 2286.942,
      "eval_steps_per_second": 10.963,
      "step": 238
    },
    {
      "epoch": 7.0,
      "step": 238,
      "total_flos": 3964550540676096.0,
      "train_loss": 0.5319601547818223,
      "train_runtime": 69.2371,
      "train_samples_per_second": 6175.161,
      "train_steps_per_second": 24.553
    }
  ],
  "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": 3964550540676096.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}