---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: id
    dtype: string
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 2987636058.0
    num_examples: 8500
  - name: test
    num_bytes: 189316644.0
    num_examples: 500
  download_size: 3244122431
  dataset_size: 3176952702.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---