metadata
dataset_info:
config_name: ja
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: 1559812591.162
num_examples: 1794
- name: validation
num_bytes: 959331712.57
num_examples: 1111
- name: test
num_bytes: 1022858975.424
num_examples: 1192
download_size: 3285322107
dataset_size: 3542003279.156
configs:
- config_name: ja
data_files:
- split: train
path: ja/train-*
- split: validation
path: ja/validation-*
- split: test
path: ja/test-*