metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 9214214437.921028
num_examples: 34092
- name: test
num_bytes: 1143301837.041485
num_examples: 4262
- name: valid
num_bytes: 1134736508.4274852
num_examples: 4262
download_size: 11008882557
dataset_size: 11492252783.389997
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*