metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcription
dtype: string
- name: duration
dtype: float32
- name: voice
dtype: string
- name: speaker_id
dtype: string
- name: chapter
dtype: string
splits:
- name: train
num_bytes: 283194600
num_examples: 5510
download_size: 283057664
dataset_size: 283194600
configs:
- config_name: default
data_files:
- split: train
path: data/train-*