---
dataset_info:
  features:
  - name: images
    dtype: image
  - name: texts
    list:
    - name: assistant
      dtype: string
    - name: source
      dtype: string
    - name: user
      dtype: string
  splits:
  - name: train
    num_bytes: 6196136599.875
    num_examples: 15937
  download_size: 5990306276
  dataset_size: 6196136599.875
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---