metadata
dataset_info:
features:
- name: path
dtype: string
- name: label
sequence: int64
- name: opus
dtype:
audio:
decode: false
splits:
- name: train
num_bytes: 623497693.25
num_examples: 20550
- name: validation
num_bytes: 571092213.25
num_examples: 18886
download_size: 1191953787
dataset_size: 1194589906.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*