---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: image_path
    dtype: string
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: image
    dtype: image
  splits:
  - name: train
    num_bytes: 8209758077.2033
    num_examples: 49969
  download_size: 8137261186
  dataset_size: 8209758077.2033
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---