A thread to list all duplicate tasks

[Palipoor]

Hi
I am creating this thread to list duplicate tasks that we encounter, so we can merge them later on, preferably using the "multiple definitions" @danyaljj suggested.

[Palipoor]

I will also use this to ask @danyaljj
If we use a dataset as a classification task into "0" or "1", does it really make sense to have it as a classification task into "Yes" or "No" or any other pair? I don't think it does.

[danyaljj]

I am actually for keeping those separate. It's a measure of whether our models generalize to slightly different output encodings.