metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: filename
dtype: string
- name: duration
dtype: float64
- name: text
dtype: string
- name: lang
dtype: string
- name: id
dtype: string
- name: merged
dtype: string
- name: key
dtype: string
- name: latents
list:
list: float16
- name: n_chars
dtype: int64
- name: rate
dtype: float64
- name: speed_class
dtype: int64
- name: speed_tag
dtype: string
splits:
- name: train
num_bytes: 51569912265
num_examples: 3680120
download_size: 42018690415
dataset_size: 51569912265