trainer_state.json

{
  "best_metric": 0.7232,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0106/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 116.78289031982422,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 3.6236,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5616,
      "eval_loss": 2.901484727859497,
      "eval_runtime": 4.9243,
      "eval_samples_per_second": 761.536,
      "eval_steps_per_second": 11.981,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 117.45916748046875,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 1.8607,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6997333333333333,
      "eval_loss": 1.3036363124847412,
      "eval_runtime": 5.5955,
      "eval_samples_per_second": 670.18,
      "eval_steps_per_second": 10.544,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 562.3662109375,
      "learning_rate": 0.0,
      "loss": 1.1747,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7232,
      "eval_loss": 1.1437582969665527,
      "eval_runtime": 4.9744,
      "eval_samples_per_second": 753.859,
      "eval_steps_per_second": 11.861,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 999,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}