Datasets:
metadata
license: mit
dataset_info:
features:
- name: speaker
dtype: string
- name: source
dtype: string
- name: start
dtype: float64
- name: end
dtype: float64
- name: style
dtype: string
- name: noise
dtype: string
- name: transcription
dtype: string
- name: audio
dtype: audio
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 13550244048.138
num_examples: 64659
download_size: 14945729120
dataset_size: 13550244048.138
configs:
- config_name: default
data_files:
- split: train
path: data/train-*