metadata
dataset_info:
features:
- name: speaker_id
dtype: string
- name: audio
dtype:
audio:
decode: false
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 1385216611
num_examples: 14976
- name: validation
num_bytes: 189113973
num_examples: 2066
download_size: 1451848965
dataset_size: 1574330584
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*