trainer_state.json

{
  "best_metric": 0.6965780854225159,
  "best_model_checkpoint": "bert_uncased_L-2_H-256_A-4_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.8095260858535767,
      "learning_rate": 4.9e-05,
      "loss": 0.7188,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4084507042253521,
      "eval_loss": 0.7157241106033325,
      "eval_runtime": 0.0304,
      "eval_samples_per_second": 2331.848,
      "eval_steps_per_second": 32.843,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.639072835445404,
      "learning_rate": 4.8e-05,
      "loss": 0.6947,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5211267605633803,
      "eval_loss": 0.6965780854225159,
      "eval_runtime": 0.0298,
      "eval_samples_per_second": 2381.717,
      "eval_steps_per_second": 33.545,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.7304646968841553,
      "learning_rate": 4.7e-05,
      "loss": 0.693,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5352112676056338,
      "eval_loss": 0.697733461856842,
      "eval_runtime": 0.0305,
      "eval_samples_per_second": 2326.037,
      "eval_steps_per_second": 32.761,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2466614246368408,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.699,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5492957746478874,
      "eval_loss": 0.7025748491287231,
      "eval_runtime": 0.0303,
      "eval_samples_per_second": 2345.845,
      "eval_steps_per_second": 33.04,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1505616903305054,
      "learning_rate": 4.5e-05,
      "loss": 0.6941,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.39436619718309857,
      "eval_loss": 0.7083516120910645,
      "eval_runtime": 0.0317,
      "eval_samples_per_second": 2243.045,
      "eval_steps_per_second": 31.592,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7813518047332764,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6908,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3380281690140845,
      "eval_loss": 0.7167142033576965,
      "eval_runtime": 0.0304,
      "eval_samples_per_second": 2336.935,
      "eval_steps_per_second": 32.915,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.7492716908454895,
      "learning_rate": 4.3e-05,
      "loss": 0.6915,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.323943661971831,
      "eval_loss": 0.7229864597320557,
      "eval_runtime": 0.0313,
      "eval_samples_per_second": 2268.141,
      "eval_steps_per_second": 31.946,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 11240405621760.0,
      "train_loss": 0.6974171002705892,
      "train_runtime": 3.8825,
      "train_samples_per_second": 8177.712,
      "train_steps_per_second": 38.635
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "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": 11240405621760.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}