metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
sequence: float32
- name: samplerate
dtype: int64
- name: secs
dtype: float64
- name: speaker_id
dtype: int64
- name: _speaker_id
dtype: int64
- name: flag
dtype: int64
splits:
- name: train
num_bytes: 57330762
num_examples: 498
download_size: 47646166
dataset_size: 57330762
configs:
- config_name: default
data_files:
- split: train
path: data/train-*