feat: salt KV block hashes with LoRA adapter name#243
Merged
Conversation
Signed-off-by: Sage Ahrac <sagiahrak@gmail.com>
sagearc
requested review from
dannyharnik,
elevran,
kfirtoledo and
vMaroon
as code owners
Contributor
There was a problem hiding this comment.
Pull request overview
This PR adds support for LoRA (Low-Rank Adaptation) adapters by using the adapter name as the model identifier in KV block keys, preventing hash collisions between base models and their LoRA variants.
Key Changes:
- Introduces
effectiveModelNamelogic that uses the LoRA adapter name (if present) instead of the base model name for KV block identification - Fixes an error handling bug where parentRequestKey lookup failures were silently ignored
- Ensures consistent use of the effective model name across all KV block operations (storage, parent key lookup, and token processing)
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Member
|
Awesome work @sagearc /lgtm |
github-actions
Bot
added
the
lgtm
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Previously, KV blocks were hashed using only the base model name, which could
cause collisions between the base model and its lora adapters.
We now use the LoRA name (if present) to generate unique block hashes.
This ensures cache isolation and completes the support needed for LoRA
adapters.
closes #167