---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: height
    dtype: int64
  - name: width
    dtype: int64
  - name: caption
    dtype: string
  - name: caption_llava
    dtype: string
  splits:
  - name: train
    num_bytes: 123287434981.32
    num_examples: 951312
  download_size: 129572126531
  dataset_size: 123287434981.32
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---