metadata
dataset_info:
features:
- name: video_index
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 3646755
num_examples: 3208
download_size: 1205976
dataset_size: 3646755
configs:
- config_name: default
data_files:
- split: train
path: data/train-*