metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text_latin
dtype: string
- name: text_cyrillic
dtype: string
- name: duration
dtype: float64
- name: words_count
dtype: int64
splits:
- name: train
num_bytes: 6007316552.28
num_examples: 11444
- name: validation
num_bytes: 338242143
num_examples: 648
- name: test
num_bytes: 470705713
num_examples: 899
download_size: 6684255545
dataset_size: 6816264408.28
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*