Hubert Banville

Haha, time to start neurodecode, a competing package for time-frequency-domain EEG? :P Actually I don't think this would break anything, even code. We might have to adapt a few things,...

I agree, an example around checkpointing would be useful. For reference, there is checkpointing in the relative positioning example although it's not the main point of the example: https://braindecode.org/auto_examples/plot_relative_positioning.html#training

Looks like both with and without specifying `picks` I get this on the first 50 subjects: 

I was not expecting this, but it looks like increasing the dropout rate from 50 to 90% fixes this (same 50 subjects):  It could be that subjects 1 and...

I created a wiki page to list these naming conventions: https://github.com/braindecode/braindecode/wiki/Variable-naming-conventions I would vote for `i_xxx` instead of `idx_xxx`, and to use `n_xxx` as suggested by @agramfort.

Working on #65 I've thought a bit more about naming conventions for supercrops vs. windows. So we currently have the following concepts: - Trial - Supercrop - Crop Here are...

I was actually thinking we should use terminology that is general enough to apply to the main use cases of the package. As long as we explicitly define these terms...

I was just able to reproduce the issue on a server, I'll look into it @dcwil

Seems like the issue is with joblib.Parallel's `max_nbytes` argument, which is set to 1 MB by default. The code fails because the data that has to be memmapped is around...

I spent a bit more time looking into this, and I think I was actually looking at the problem the wrong way. The error occurs because of two things: (1)...