metadata
dataset_info:
features:
- name: text
dtype: string
- name: text_tokens
list: string
- name: audio_tokens
list: int64
- name: duration
dtype: float64
- name: n_text_tokens
dtype: int64
- name: n_audio_tokens
dtype: int64
- name: speaker_id
dtype: string
splits:
- name: train
num_bytes: 1594752406
num_examples: 316543
download_size: 1511904429
dataset_size: 1594752406
configs:
- config_name: default
data_files:
- split: train
path: data/train-*