metadata
dataset_info:
features:
- name: answer
dtype: string
- name: category_id
dtype: int64
- name: id
dtype: int64
- name: question
dtype: string
- name: video_id
dtype: int64
- name: audio
dtype: audio
- name: image
dtype: image
- name: category
dtype: string
splits:
- name: train
num_bytes: 10102519235.746
num_examples: 140554
- name: valid
num_bytes: 831734018.921
num_examples: 11143
download_size: 1790605662
dataset_size: 10934253254.667
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*