for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
from fuel.datasets.hdf5 import H5PYDataset
from fuel.utils import find_in_data_path
class YouTubeAudio(H5PYDataset):
def __init__(self, youtube_id, **kwargs):
super(YouTubeAudio, self).__init__(
file_or_path=find_in_data_path('{}.hdf5'.format(youtube_id)),
which_sets=('train',), **kwargs
)
mila-udem /
fuel
