trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4533984661102295,
  "best_model_checkpoint": "tiny_bert_rand_5_v1_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 6.7664947509765625,
      "learning_rate": 4.9e-05,
      "loss": 0.4063,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.4533984661102295,
      "eval_runtime": 0.2729,
      "eval_samples_per_second": 3195.702,
      "eval_steps_per_second": 14.659,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.511747360229492,
      "learning_rate": 4.8e-05,
      "loss": 0.2372,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.4990834891796112,
      "eval_runtime": 0.2691,
      "eval_samples_per_second": 3241.015,
      "eval_steps_per_second": 14.867,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.99404239654541,
      "learning_rate": 4.7e-05,
      "loss": 0.1798,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.5129833817481995,
      "eval_runtime": 0.2822,
      "eval_samples_per_second": 3089.656,
      "eval_steps_per_second": 14.173,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.237902641296387,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1434,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.5429797172546387,
      "eval_runtime": 0.2741,
      "eval_samples_per_second": 3181.646,
      "eval_steps_per_second": 14.595,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.562696933746338,
      "learning_rate": 4.5e-05,
      "loss": 0.1149,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.6099434494972229,
      "eval_runtime": 0.2641,
      "eval_samples_per_second": 3302.166,
      "eval_steps_per_second": 15.148,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.458475112915039,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0949,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.801605504587156,
      "eval_loss": 0.6271517872810364,
      "eval_runtime": 0.267,
      "eval_samples_per_second": 3265.915,
      "eval_steps_per_second": 14.981,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.19609037673834598,
      "train_runtime": 221.6128,
      "train_samples_per_second": 15195.197,
      "train_steps_per_second": 59.563
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}