Add Server-side Checkpointing to Server Base #56
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
… Added tests to verify it works as intended
emersodb
requested review from
fatemetkl,
jewelltaylor,
sanaAyrml and
yuchongzhang
…ointing. Also changing the FLServerWithCheckpointing to include the potential for specification of high types of parameter exchanges (like those with packing)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Ticket
This PR adds in server-side checkpointing into the base FlServer class. In order to use the checkpointing. One needs to include a checkpointing class and implement a server-side model hydration procedure. Hydration is simply a way to insert the Flower Parameters into a torch model for checkpointing. By default: checkpointing is performed on the aggregated loss using the overridden
evaluate_roundfunction.Tests were added to make sure the flow of the checkpointing hydration procedure makes sense. I also implemented server-side checkpointing into the basic example as a demonstration of the concept.
Note: I'm not entirely happy with the dancing I had to do to hydrate a class level model, but
evaluate_rounddoesn't take a model and therefore we have to assume it exists at the class level. If anyone has a better idea I'm open to it.