trainer_state.json

{
  "best_metric": 0.9034666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0590/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": 41.67326736450195,
      "learning_rate": 0.00042857142857142855,
      "loss": 1.4047,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8298666666666666,
      "eval_loss": 0.571993350982666,
      "eval_runtime": 15.2502,
      "eval_samples_per_second": 245.898,
      "eval_steps_per_second": 3.869,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 18.86758041381836,
      "learning_rate": 0.00035714285714285714,
      "loss": 0.4668,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.4690561890602112,
      "eval_runtime": 15.2675,
      "eval_samples_per_second": 245.62,
      "eval_steps_per_second": 3.864,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 29.338115692138672,
      "learning_rate": 0.0002857142857142857,
      "loss": 0.2755,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.3981682360172272,
      "eval_runtime": 15.1981,
      "eval_samples_per_second": 246.742,
      "eval_steps_per_second": 3.882,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 28.509403228759766,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.1527,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8946666666666667,
      "eval_loss": 0.38306862115859985,
      "eval_runtime": 15.2625,
      "eval_samples_per_second": 245.701,
      "eval_steps_per_second": 3.866,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 35.28919982910156,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.0965,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8970666666666667,
      "eval_loss": 0.383367657661438,
      "eval_runtime": 15.1632,
      "eval_samples_per_second": 247.309,
      "eval_steps_per_second": 3.891,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 53.91545104980469,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.0557,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9032,
      "eval_loss": 0.3702581524848938,
      "eval_runtime": 15.3141,
      "eval_samples_per_second": 244.873,
      "eval_steps_per_second": 3.853,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 30.892024993896484,
      "learning_rate": 0.0,
      "loss": 0.0357,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.3582104742527008,
      "eval_runtime": 15.2774,
      "eval_samples_per_second": 245.461,
      "eval_steps_per_second": 3.862,
      "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
}