Skip to content

[1/N] Update MMQ kernel catch up upstream llama.cpp implementation #5

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 8 commits into from
Aug 9, 2025
Merged

[1/N] Update MMQ kernel catch up upstream llama.cpp implementation #5

merged 8 commits into from
Aug 9, 2025

Conversation

Isotr0py
Copy link
Owner

@Isotr0py Isotr0py marked this pull request as ready for review August 7, 2025 05:46
Isotr0py added 2 commits August 9, 2025 21:53
@Isotr0py
fix shmem
756a62b 
Signed-off-by: Isotr0py <[email protected]>
@Isotr0py
update benchmark
02918e7 
Signed-off-by: Isotr0py <[email protected]>
@Isotr0py
Copy link
Owner Author

Isotr0py commented Aug 9, 2025

Benchmark results

Hardware: NVIDIA GeForce RTX 3090 24GB
Hidden size=1024, num_tokens=4096

Quantization Main Time (ms) PR Time (ms)
Q2_K 21.37505244 9.757954925
Q3_K 54.36249449 8.700753478
Q4_K 32.7376706 6.185650527
Q5_K 42.49698077 6.589026637
Q6_K 45.83637421 6.769261481
Q4_0 23.81569172 5.485869246
Q5_0 37.47556383 5.97816959
Q8_0 40.68143068 5.784195224

@Isotr0py
use csv
c7f61c5 
Signed-off-by: Isotr0py <[email protected]>
@Isotr0py Isotr0py merged commit 239c8fd into main Aug 9, 2025
@Isotr0py Isotr0py deleted the mmq-update2 branch August 9, 2025 14:18
@Isotr0py Isotr0py mentioned this pull request Sep 28, 2025
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@Isotr0py