trainer_state.json

{
  "best_global_step": 34,
  "best_metric": 0.6151070594787598,
  "best_model_checkpoint": "distilbert_rand_20_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": 1.3837093114852905,
      "learning_rate": 4.9e-05,
      "loss": 0.6147,
      "step": 34
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6912751793861389,
      "eval_loss": 0.6151070594787598,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.4606,
      "eval_samples_per_second": 2264.206,
      "eval_steps_per_second": 10.854,
      "step": 34
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2729761600494385,
      "learning_rate": 4.8e-05,
      "loss": 0.5912,
      "step": 68
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6883988380432129,
      "eval_loss": 0.6195095181465149,
      "eval_matthews_correlation": -0.016334423518002312,
      "eval_runtime": 0.4886,
      "eval_samples_per_second": 2134.865,
      "eval_steps_per_second": 10.234,
      "step": 68
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.290869951248169,
      "learning_rate": 4.7e-05,
      "loss": 0.5468,
      "step": 102
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6864812970161438,
      "eval_loss": 0.6209526062011719,
      "eval_matthews_correlation": 0.054255997232782946,
      "eval_runtime": 0.4854,
      "eval_samples_per_second": 2148.567,
      "eval_steps_per_second": 10.3,
      "step": 102
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.986222982406616,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4978,
      "step": 136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6423777341842651,
      "eval_loss": 0.693848729133606,
      "eval_matthews_correlation": 0.08015877090661065,
      "eval_runtime": 0.441,
      "eval_samples_per_second": 2365.037,
      "eval_steps_per_second": 11.338,
      "step": 136
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.9276981353759766,
      "learning_rate": 4.5e-05,
      "loss": 0.4357,
      "step": 170
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6548417806625366,
      "eval_loss": 0.7163270115852356,
      "eval_matthews_correlation": 0.0812620979720463,
      "eval_runtime": 0.4589,
      "eval_samples_per_second": 2273.048,
      "eval_steps_per_second": 10.897,
      "step": 170
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.8644490242004395,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3843,
      "step": 204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6529242396354675,
      "eval_loss": 0.8270009756088257,
      "eval_matthews_correlation": 0.08445604290824693,
      "eval_runtime": 0.4736,
      "eval_samples_per_second": 2202.425,
      "eval_steps_per_second": 10.558,
      "step": 204
    },
    {
      "epoch": 6.0,
      "step": 204,
      "total_flos": 3398186177722368.0,
      "train_loss": 0.5117585098042208,
      "train_runtime": 58.7225,
      "train_samples_per_second": 7280.859,
      "train_steps_per_second": 28.95
    }
  ],
  "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": 3398186177722368.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}