metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 480248141.02917224
num_examples: 4119
- name: test
num_bytes: 60651506.28141387
num_examples: 515
- name: valid
num_bytes: 58619378.28141387
num_examples: 515
download_size: 528681604
dataset_size: 599519025.592
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*