---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: file_name
    dtype: string
  - name: language
    dtype: string
  - name: caption
    dtype: string
  - name: question
    dtype: string
  - name: weight
    dtype: float64
  splits:
  - name: train
    num_bytes: 32196135
    num_examples: 166057
  download_size: 15999769
  dataset_size: 32196135
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---