metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 499683339.5520001
num_examples: 4192
- name: test
num_bytes: 60941107.3
num_examples: 524
- name: valid
num_bytes: 62724755.3
num_examples: 524
download_size: 392924700
dataset_size: 623349202.1520001
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*