metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 108209141028.79991
num_examples: 1337649
- name: test
num_bytes: 13886414744.109455
num_examples: 167207
- name: valid
num_bytes: 13372345801.458637
num_examples: 167206
download_size: 109246997469
dataset_size: 135467901574.368
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*