metadata
dataset_info:
features:
- name: text
dtype: string
- name: path
dtype: string
- name: id
dtype: string
- name: text_description
dtype: string
- name: audio_tokens
sequence:
sequence: int64
splits:
- name: train
num_bytes: 78115189
num_examples: 32215
- name: validation
num_bytes: 13105273
num_examples: 5589
download_size: 19785174
dataset_size: 91220462
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*