iq2_xxs: tune quantization#5320
Merged
Merged
Conversation
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way.
|
Could this applied to other IQ quants？ |
ggerganov
approved these changes
Feb 5, 2024
Contributor
Author
Yes, but with much less gain. I.e., one gets an increase in PPL if one reduces the scale search range as aggressively as here, or one can keep about the same PPL but with much lower speedup. |
Contributor
|
Any noticeable speed-up you can offer us with a close to equal perplexity is interresting for the CPU poor, @ikawrakow! |
jordankanter
pushed a commit
to jordankanter/llama.cpp
that referenced
this pull request
Mar 13, 2024
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
Seunghhon
pushed a commit
to Seunghhon/llama.cpp
that referenced
this pull request
Apr 26, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
phuongncn
pushed a commit
to phuongncn/llama.cpp-gx10-dgx-sparks-deepseekv4
that referenced
this pull request
Apr 28, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
ljubomirj
pushed a commit
to ljubomirj/llama.cpp
that referenced
this pull request
May 6, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
my-other-github-account
pushed a commit
to my-other-github-account/llama.cpp
that referenced
this pull request
May 15, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
my-other-github-account
pushed a commit
to my-other-github-account/llama.cpp
that referenced
this pull request
May 15, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
AlexiAlp
pushed a commit
to minghaop/llama.cpp
that referenced
this pull request
Jun 2, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
AlexiAlp
pushed a commit
to minghaop/llama.cpp
that referenced
this pull request
Jun 2, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
fukuro-kun
pushed a commit
to fukuro-kun/fukuro-llama-cpp-turboquant
that referenced
this pull request
Jul 5, 2026
We get slightly better PPL, and we cut quantization time in nearly half. The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way. Co-authored-by: Iwan Kawrakow <[email protected]>
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
4 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.
We get slightly better PPL, and we cut quantization time in nearly half.
The trick is to 1st quantize without forcing points onto the E8-lattice. We can then use a narrower search range around the block scale that we got that way, which gives the significant reduction in quantization time.
The code becomes simpler too, so it is a win-win.
Here is a comparison between PPL with this PR and PR #4773 for a context of 4096