metadata
dataset_info:
features:
- name: _id
dtype: string
- name: speech
dtype: audio
- name: transcription
dtype: string
- name: questions
sequence: string
- name: answers
sequence:
sequence: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 71044170590.61
num_examples: 13285
download_size: 65033191852
dataset_size: 71044170590.61
configs:
- config_name: default
data_files:
- split: train
path: data/train-*