metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 574468188.1947066
num_examples: 7159
- name: test
num_bytes: 73434870.38864678
num_examples: 895
- name: valid
num_bytes: 77839059.38864678
num_examples: 895
download_size: 680790066
dataset_size: 725742117.9720001
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*