metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 9302913443.561954
num_examples: 111200
- name: test
num_bytes: 1119354595.6598015
num_examples: 13901
- name: valid
num_bytes: 1125525152.5452442
num_examples: 13900
download_size: 9232284149
dataset_size: 11547793191.767
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*