dutta18's picture
Update README.md
6f1037f verified
metadata
dataset_info:
  features:
    - name: multiple_choice_answer
      dtype: string
    - name: question
      dtype: string
    - name: image
      dtype: image
  splits:
    - name: train
      num_bytes: 16427957.954119343
      num_examples: 100
  download_size: 16332826
  dataset_size: 16427957.954119343
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

A small VQAv2 Dataset For Testing Script or Notebook on Google Collab. Use split='train'.