metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: speech_id
dtype: string
- name: speaker_id
dtype: string
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 1585742876.38
num_examples: 3446
- name: test
num_bytes: 276840471
num_examples: 680
download_size: 1668931471
dataset_size: 1862583347.38
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*