---
dataset_info:
  features:
  - name: conversations
    dtype: string
  - name: type
    dtype: string
  - name: id
    dtype: int64
  - name: image
    dtype: string
  splits:
  - name: train
    num_bytes: 71942597
    num_examples: 40002
  download_size: 28578667
  dataset_size: 71942597
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---