metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 598099904.5376791
num_examples: 1934
- name: test
num_bytes: 74221544.33416046
num_examples: 242
- name: valid
num_bytes: 74699082.33416046
num_examples: 242
download_size: 665620175
dataset_size: 747020531.206
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*