metadata
dataset_info:
features:
- name: id
dtype: string
- name: speaker_id
dtype: string
- name: ref_text
dtype: string
- name: target_text
dtype: string
- name: target_tokens
dtype: string
- name: codec
dtype: string
- name: codec_vocab_size
dtype: int64
- name: codec_sample_rate
dtype: int64
- name: language
dtype: string
- name: duration_ref
dtype: float64
- name: duration_target
dtype: float64
- name: ref_audio
dtype:
audio:
sampling_rate: 16000
- name: target_audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 2190524962
num_examples: 9393
- name: test
num_bytes: 261977777
num_examples: 1043
download_size: 2672192337
dataset_size: 2452502739
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*