Taking too much time

#7
by wajahatalikhan - opened

Hi,
I am trying to download VGGSound dataset by using the following command on my machine but it is taking too much time.

from datasets import load_dataset
ds = load_dataset("Loie/VGGSound")

any suggestions?

hi, if possible, you could try downloading it manually, and it may be faster.

Thank you, I did.
Actually I am not able to properly access the video files. When I run the code it gives me an error saying that the file in csv doesnot match the video file. Maybe it is because of a little different naming of the video files? What do you think and how can I solve it?

The first column of the CSV file is 'youtube_id', and the second column is 'start time'. You need to process these columns yourself. For example, '--0PQM4-hqg,30,waterfall burbling,train' should be processed to '--0PQM4-hqg_000030'.

Sign up or log in to comment