FIX: Bug with handling of active adapters #1659
Merged
Conversation
This file contains 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
There was a bug for some models like IA3, LoHa, etc., where calling set_adapter would not correctly update the active_adapter. This is now fixed. Note that this is not about the active_adapter attribute on PeftModel or layers, which are handled separately. This PR also ensures that LoraModel, IA3Model, etc. consistently use self.active_adapters, not self.active_adapter. The latter should be treated more like a private attribute (but this isn't changed for backwards compatibility).
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
pacman100
approved these changes
Apr 17, 2024
There was a problem hiding this comment.
Thank you @BenjaminBossan for the fixes wrt properly setting adapters and updating the active adapters!
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Outdated suggestions cannot be applied.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.
There was a bug for some models like IA3, LoHa, etc., where calling
set_adapterwould not correctly update theactive_adapter. This is now fixed.Note that this is not about the
active_adapterattribute onPeftModelor layers, which are handled separately.This PR also ensures that
LoraModel,IA3Model, etc. consistently useself.active_adapters, notself.active_adapter. The latter should be treated more like a private attribute (but this isn't changed for backwards compatibility).