metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: speaker_id
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 3287914474
num_examples: 10000
download_size: 3153212824
dataset_size: 3287914474
configs:
- config_name: default
data_files:
- split: train
path: data/train-*