metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 2488721141.145171
num_examples: 12890
- name: test
num_bytes: 326389692.0569508
num_examples: 1612
- name: valid
num_bytes: 308662185.54587823
num_examples: 1611
download_size: 3086264891
dataset_size: 3123773018.748
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*