trainer_state.json

{
  "best_metric": 0.9056,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0920/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": 38.20867156982422,
      "learning_rate": 0.000375,
      "loss": 1.25,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8576,
      "eval_loss": 0.4800911843776703,
      "eval_runtime": 7.0981,
      "eval_samples_per_second": 528.311,
      "eval_steps_per_second": 8.312,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 51.30622100830078,
      "learning_rate": 0.00012500000000000006,
      "loss": 0.3252,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8978666666666667,
      "eval_loss": 0.3401147425174713,
      "eval_runtime": 7.3724,
      "eval_samples_per_second": 508.656,
      "eval_steps_per_second": 8.003,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 30.860273361206055,
      "learning_rate": 0.0,
      "loss": 0.1228,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.3356272280216217,
      "eval_runtime": 7.2284,
      "eval_samples_per_second": 518.787,
      "eval_steps_per_second": 8.162,
      "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
}