---
dataset_info:
  features:
  - name: images
    sequence: image
  - name: conversation
    list:
      list:
      - name: content
        dtype: string
      - name: role
        dtype: string
  splits:
  - name: train
    num_bytes: 554067397.632
    num_examples: 1024
  download_size: 551168187
  dataset_size: 554067397.632
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---