Search results

The detected CUDA version (12.0) mismatches the version that was used to compile PyTorch (11.8). Please make sure to use the same CUDA versions.

Status: Open.

#212 In thu-ml/SageAttention;

· gateway opened on Jul 15, 2025

sageattention-1 论文中的CogvideoX的测试

Status: Open.

#211 In thu-ml/SageAttention;

· Romaosir opened on Jul 15, 2025

SageAttentionV3

Status: Open.

#210 In thu-ml/SageAttention;

· Bezdarnost opened on Jul 9, 2025

Compiling from main branch results in 2.1.1 version?

Status: Open.

#209 In thu-ml/SageAttention;

· SamTyurenkov opened on Jul 9, 2025

How to use NVFP4 gemm

Status: Open.

#208 In thu-ml/SageAttention;

· nanzh-19 opened on Jul 9, 2025

sageattention2.2 can't be used on ubuntu20.04?

Status: Open.

#206 In thu-ml/SageAttention;

· fangg2000 opened on Jul 6, 2025

Do you have plans to support qk_head_dim = 192 in DeepSeek MLA ?

Status: Open.

#204 In thu-ml/SageAttention;

· artetaout opened on Jul 3, 2025

The performance of sageattention2.2 is worse than sageattention2.1.

Status: Open.

#203 In thu-ml/SageAttention;

· triple-Mu opened on Jul 2, 2025

How to leverage Sage Attn 2++ ?

Status: Open.

#200 In thu-ml/SageAttention;

· deepbeepmeep opened on Jul 2, 2025

Unspecified CUDA launch failure on L4 GPU (sm89) with pv_accum_dtype="fp32+fp16"

Status: Open.

#199 In thu-ml/SageAttention;

· Li1177 opened on Jul 2, 2025

[ COLLAB ] Integration into Axolotl framework

Status: Open.

#198 In thu-ml/SageAttention;

· NanoCode012 opened on Jul 1, 2025

add support for 2060 gpu (Solved using sageatt 1 and triton 3.1 post17)

Status: Open.

#197 In thu-ml/SageAttention;

· noise2014 opened on Jul 1, 2025