metadata
dataset_info:
features:
- name: text
dtype: string
- name: text_tokens
list: int64
- name: audio_tokens
list:
list: int64
- name: length
dtype: float64
- name: n_text_tokens
dtype: int64
- name: n_audio_tokens
dtype: int64
splits:
- name: train
num_bytes: 270555870
num_examples: 40816
download_size: 267754043
dataset_size: 270555870
configs:
- config_name: default
data_files:
- split: train
path: data/train-*