metadata
dataset_info:
config_name: ko
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: 77586232
num_examples: 92
- name: validation
num_bytes: 38561398
num_examples: 49
- name: test
num_bytes: 47260349
num_examples: 57
download_size: 148028604
dataset_size: 163407979
configs:
- config_name: ko
data_files:
- split: train
path: ko/train-*
- split: validation
path: ko/validation-*
- split: test
path: ko/test-*