Matt Whiteway
Matt Whiteway
awesome, thanks so much for looking into this. yeah maybe we can switch over to the discord? https://discord.gg/tDUPdRj4BM to answer your most recent questions here though: * if you set...
That's a good observation, it is indeed possible that oversampling would force the model to focus more on these less-frequent keypoints. One way to do this would be to implement...
We do not currently support multi-GPU training but it should not be difficult to implement with our current setup (at least for the supervised models). We'll be happy to look...
Hi @EugenioBertolini, thanks for the kind words! Glad you were able to get the training up and running without issues :) It would be great if you could add a...
Awesome, thank you so much! No rush On Tue, Mar 5, 2024, 09:39 Eugenio Bertolini ***@***.***> wrote: > Ok, as soon as I have some spare time today or tomorrow...
Hi @EugenioBertolini, just checking in - is this still a PR you'd like to make?
No problem, let me know if I can be of any help!
@wyclearnpy it looks like there is an issue with how you have organized/labeled your data. The multiview option requires labels from all views at a given time point. So for...
Great, you should still be able to get comparable performance to DLC even without the multiview losses. You might also try to extract the context frames for each of your...
@wyclearnpy I'm not able to open the image that you linked above. Can you copy/paste the error from the command line here?