Repository · GitHub

thu-ml/SpargeAttn

[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

Owner	thu-ml
License (as declared)	Apache-2.0
Stars	1,261
Forks	110
Main language	Cuda
Created	25 February 2025
Last updated at source	25 February 2026

cudaai-infraattentioninference-accelerationllmmlsysquantizationsageattentionsparse-attentionvideo-generationvision-transformervit

Embed this entry

<iframe src="https://silytics.si/embed/item/65baaea6-a308-4ae4-b4e8-4f598c65df24" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

More from thu-ml

tianshouRepository
SageAttentionRepository
TurboDiffusionRepository
zhusuanRepository

Silytics shows public metadata from GitHub and links to the original. Information may be incomplete or out of date. Last refreshed 15 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.