---
dataset_info:
  features:
  - name: caption
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  - name: ib_embed
    sequence: float64
    length: 1024
  splits:
  - name: train
    num_bytes: 1589587882
    num_examples: 186088
  download_size: 1251917269
  dataset_size: 1589587882
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---