Quantize KV Cache of TabPFN-3 run with fit_mode="fit_with_cache"#983
Merged
Conversation
Collaborator
Author
|
This change is part of the following stack: Change managed by git-spice. |
bejaeger
requested review from
eliott-kalfon
and removed request for
a team and
eliott-kalfon
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request introduces per-tensor symmetric int8 quantization for the KV cache in TabPFN-3 models to reduce memory footprint during inference with minimal accuracy loss. The feedback suggests using a fully symmetric range of [-127, 127] for int8 quantization to prevent asymmetry, and updating the type annotations in the attention layer's forward method signature to include QuantizedKVCacheEntry to ensure static type checking safety.
priorphil
approved these changes
May 28, 2026
priorphil
reviewed
May 28, 2026
Collaborator
priorphil
left a comment
priorphil
left a comment
There was a problem hiding this comment.
Just to double check, there's no native pytorch quantized tensor that would de-quantize on the fly?
github-merge-queue
Bot
removed this pull request from the merge queue due to failed status checks
github-merge-queue
Bot
removed this pull request from the merge queue due to failed status checks
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
No description provided.