metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio_tokens
sequence:
sequence: int64
splits:
- name: train
num_bytes: 115521510
num_examples: 54472
- name: validation
num_bytes: 3471081
num_examples: 1400
download_size: 26204194
dataset_size: 118992591
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*