metadata
dataset_info:
config_name: ru
features:
- name: path
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: condition_on_prev
sequence: int64
- name: whisper_transcript
dtype: string
splits:
- name: train
num_bytes: 4364362221.793
num_examples: 5041
- name: validation
num_bytes: 1771814770.088
num_examples: 2054
- name: test
num_bytes: 1832794707.363
num_examples: 2131
download_size: 7357091175
dataset_size: 7968971699.244
configs:
- config_name: ru
data_files:
- split: train
path: ru/train-*
- split: validation
path: ru/validation-*
- split: test
path: ru/test-*