trainer_state.json

{
  "best_metric": 0.8856,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0392/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 133.8059844970703,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 3.197,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6784,
      "eval_loss": 1.5079265832901,
      "eval_runtime": 7.732,
      "eval_samples_per_second": 484.995,
      "eval_steps_per_second": 7.631,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 80.31352996826172,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 0.9092,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8229333333333333,
      "eval_loss": 0.6081333160400391,
      "eval_runtime": 7.762,
      "eval_samples_per_second": 483.123,
      "eval_steps_per_second": 7.601,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 166.04136657714844,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 0.5293,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8616,
      "eval_loss": 0.45934614539146423,
      "eval_runtime": 7.6199,
      "eval_samples_per_second": 492.132,
      "eval_steps_per_second": 7.743,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 126.33173370361328,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 0.4067,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.4193449318408966,
      "eval_runtime": 7.6211,
      "eval_samples_per_second": 492.052,
      "eval_steps_per_second": 7.742,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 85.41637420654297,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 0.3424,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8749333333333333,
      "eval_loss": 0.39347684383392334,
      "eval_runtime": 7.6102,
      "eval_samples_per_second": 492.762,
      "eval_steps_per_second": 7.753,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 60.22658920288086,
      "learning_rate": 9.999999999999999e-06,
      "loss": 0.2919,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.3777764141559601,
      "eval_runtime": 7.6532,
      "eval_samples_per_second": 489.993,
      "eval_steps_per_second": 7.709,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 297.2530212402344,
      "learning_rate": 0.0,
      "loss": 0.272,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8856,
      "eval_loss": 0.36249491572380066,
      "eval_runtime": 8.0396,
      "eval_samples_per_second": 466.439,
      "eval_steps_per_second": 7.339,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}