Skip to content

Issues: thu-ml/SageAttention

Beta
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

53 Open 109 Closed
53 Open 109 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

SegaAttention2, 4090, 3 hours, 8secs.
#183 opened Jun 4, 2025 by K-O-N-B
3
Where is the SageAttention2++ code?
#182 opened May 28, 2025 by aikitoria
1
fix-gcc15
#180 opened May 25, 2025 by al-swaiti
RuntimeError: Cannot find CUDA_HOME. CUDA must be available to build the package.
#179 opened May 25, 2025 by wujpia
1
Question about usage of make_smem_desc
#178 opened May 24, 2025 by botbw
Error in worker loop (rank 3): SM89 kernel is not available. Make sure you GPUs with compute capability 8.9.
#177 opened May 22, 2025 by LiuYinfeng01
1
Release SageAttention3 on Hugging Face
#176 opened May 22, 2025 by NielsRogge
MultiQuery Attention
#169 opened May 13, 2025 by asahni-sc
wonder why using different(128/64) block-size in sageattn-v1 for quant q, k
#167 opened May 8, 2025 by ZJLi2013
head-dim lower than 128
#166 opened May 6, 2025 by XuekuanWang
1
How to simply verify the successful installation of sageattention?
#165 opened May 6, 2025 by aswordok
2
BUG: RTX 50XX nan returned by _fused.mean_scale_fuse_quant_cuda and _fused.scale_fuse_quant_cuda
#164 opened Apr 30, 2025 by deepbeepmeep
Sage Attention make time longer
#163 opened Apr 30, 2025 by lonelygoogle
4
Sage Attention 2 not work with Pytorch Compile
#162 opened Apr 30, 2025 by pivtienduc
9
SageAttention2.0 is slower VS Flash attention 2
#161 opened Apr 23, 2025 by XuekuanWang
3
Transformers intergrate
#159 opened Apr 20, 2025 by lucasjinreal
SageAttention errors on 1.0.6 and 2.1.1 on 5090
#158 opened Apr 18, 2025 by BBBAAA2
5
No GPUs found
#157 opened Apr 17, 2025 by rkfg
2
The accuracy loss in the CUDA version is much more than Triton version for llama-3.2
#154 opened Apr 7, 2025 by WanliZhong
2
Install failed. plz help.
#153 opened Apr 3, 2025 by lgs777
1
K Sampler [WinError 2] The system cannot find the file specified.
#152 opened Apr 2, 2025 by REG-0422
1
Sage Attention vs Flash Attention Speed Comparison with Wan 2.1 - Sage Attention is 37% Faster - 720p - 14b model - tested on Windows Python VENV - no WSL
#150 opened Mar 25, 2025 by FurkanGozukara
4
qk_int_sv_f8_cuda_sm90 kernel failed to run on H20
#149 opened Mar 25, 2025 by SimonSongg
4
SageAttention fails to work on RTX 50xx series GPU (e.g. 5090) despite clean venv install
#148 opened Mar 24, 2025 by Teskun
54
I've built Windows wheels
#146 opened Mar 24, 2025 by woct0rdho
1
ProTip! Mix and match filters to narrow down what you’re looking for.