---
dataset_info:
  features:
  - name: text_input
    dtype: string
  - name: text_output
    dtype: string
  - name: image_input
    dtype: image
  - name: image_output
    dtype: image
  - name: dataset
    dtype: string
  - name: id
    dtype: string
  - name: n_aux
    dtype: int32
  splits:
  - name: train
    num_bytes: 33539542629.0
    num_examples: 40000
  download_size: 26872716432
  dataset_size: 33539542629.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---