metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: text
dtype: string
- name: text_with_preprocessing
dtype: string
- name: audio_tokens
sequence: int64
splits:
- name: train
num_bytes: 10815126293
num_examples: 53218
- name: validation
num_bytes: 813854183
num_examples: 4006
download_size: 11221145747
dataset_size: 11628980476
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*