metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float64
- name: speaker_id
dtype: string
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 4634693726.917114
num_examples: 11806
- name: test
num_bytes: 257526707.7054429
num_examples: 656
- name: valid
num_bytes: 257526675.7054429
num_examples: 656
download_size: 5054759735
dataset_size: 5149747110.327999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*