metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
sequence: float32
- name: samplerate
dtype: int64
- name: secs
dtype: 'null'
- name: speaker_id
dtype: 'null'
- name: _speaker_id
dtype: 'null'
- name: flag
dtype: 'null'
splits:
- name: train
num_bytes: 593619279
num_examples: 2839
download_size: 291174953
dataset_size: 593619279
configs:
- config_name: default
data_files:
- split: train
path: data/train-*