fix(sam3_image): prevent unwrap panic caused by LruCache eviction during mixed queries#254
Open
darkSuperman wants to merge 2 commits intojamjamjon:mainfrom
Open
fix(sam3_image): prevent unwrap panic caused by LruCache eviction during mixed queries#254darkSuperman wants to merge 2 commits intojamjamjon:mainfrom
darkSuperman wants to merge 2 commits intojamjamjon:mainfrom
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
🐛 The Bug
Calling
self.text_cache.get(&p.text).unwrap()occasionally panics withOption::unwrap()on a None value.This happens when:
The number of unique prompts in a single batch exceeds the default LruCache capacity (16).
Or, a previously cached prompt is pushed to the tail (least recently used), and a subsequent query mixes old (tail) prompts with new uncached ones.
🕵️ Root Cause
The original code used
self.text_cache.contains(&p.text)which only checks existence but does not promote the item to the Most Recently Used (MRU) position.When new prompts are put() into the cache, the capacity overflows, evicting the oldest items (even if they are about to be accessed in the current batch).
The subsequent unwrap() encounters a None value.
🛠️ The Fix
Promote MRU: Replaced
.contains()with.get(). This inherently promotes accessed items to the front, granting them immunity from eviction during the same batch processing.Dynamic Resize: Dynamically check
seen.len(). If it exceeds current capacity,.resize()the cache to guarantee the entire batch fits safely.Avoid Double Encoding: Extracted encode_texts into a variable encoded_feats to prevent running the heavy vision-language model twice for the same inputs.