---
dataset_info:
  features:
  - name: img_good
    dtype: image
  - name: img_bad
    dtype: image
  - name: caption
    dtype: string
  splits:
  - name: train
    num_bytes: 6111116673.0
    num_examples: 302768
  - name: validation
    num_bytes: 95749355.5
    num_examples: 5198
  - name: test
    num_bytes: 93727711.0
    num_examples: 5220
  download_size: 6226398399
  dataset_size: 6300593739.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---