metadata
dataset_info:
features:
- name: sentence_id
dtype: int32
- name: sentence
dtype: string
- name: input_speaker
dtype: string
- name: output_speaker
dtype: string
- name: source_audio
dtype:
audio:
sampling_rate: 16000
- name: target_audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 12948486312
num_examples: 72000
- name: validation
num_bytes: 1618542789
num_examples: 9000
- name: test
num_bytes: 1618542141
num_examples: 9000
download_size: 5603210527
dataset_size: 16185571242
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*