metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 621015746.8685446
num_examples: 894
- name: test
num_bytes: 187135363.08763692
num_examples: 256
- name: valid
num_bytes: 98381996.04381846
num_examples: 128
download_size: 775819937
dataset_size: 906533106
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*