---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype:
      class_label:
        names:
          '0': fake
          '1': real
  splits:
  - name: train
    num_bytes: 10881873439.327
    num_examples: 98163
  - name: validation
    num_bytes: 574289333.296
    num_examples: 5168
  - name: test
    num_bytes: 592123012.48
    num_examples: 5440
  download_size: 13085799986
  dataset_size: 12048285785.102999
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
license: mit
task_categories:
- image-classification
language:
- en
size_categories:
- 10K<n<100K
---
#This dataset is collected from ImageReward for the fake class and COCO for the real class