Changes to DL models to work on TUAB

meeg-ml-benchmarks / brain-age-benchmark-paper

M/EEG brain age benchmark paper

https://meeg-ml-benchmarks.github.io/brain-age-benchmark-paper/

BSD 3-Clause "New" or "Revised" License

24 stars 10 forks source link

Changes to DL models to work on TUAB #22

Closed hubertjb closed 3 years ago

hubertjb commented 3 years ago

This PR fixes a few minor things so that we can train and evaluate deep learning models on TUAB. I rebased on #20 to include @gemeinl's latest changes.

Some important points:

I couldn't get the script to run with n_jobs>1 and preload=False at the same time. It seems one of the processes suddenly has trouble reshaping windows when loading them.
I changed the default batch size to 256, as this made training a bit faster.

gemeinl commented 3 years ago

Great, thanks!