---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 2241515242.832
    num_examples: 7168
  - name: test
    num_bytes: 284202360.0
    num_examples: 891
  - name: val
    num_bytes: 261038797.0
    num_examples: 859
  download_size: 2782146548
  dataset_size: 2786756399.832
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: val
    path: data/val-*
---