Implementing DivideMix #10561
Unanswered
dreaquil
asked this question in
code help: CV
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi! Recently started using pytorch lightning and really enjoying it so far! However, I'm still getting used to the way of doing things. I'm currently working on implementing a DivideMix model which is based on MixMatch (Semi supervised learning). This has a rather complicated training loop as it contains two models and uses the output of one model to sample the training data for the other. I'm struggling to figure out how best to fit this into the pytorch lightning framework or if this may be beyond the framework's structure. Does anyone have any idea what the best way to structure this would be?
Training starts with:
From my understanding, the following steps need to take place on each epoch:
Beta Was this translation helpful? Give feedback.
All reactions