metadata
dataset_info:
features:
- name: number
dtype: string
- name: image
dtype: image
- name: audio
dtype:
audio:
sampling_rate: 16000
mono: false
- name: text
dtype: string
splits:
- name: train
num_bytes: 24300453
num_examples: 533
download_size: 16173814
dataset_size: 24300453
configs:
- config_name: default
data_files:
- split: train
path: data/train-*