From 23c85343fba6dd1bc346dd98e6ff3d9472e58b60 Mon Sep 17 00:00:00 2001 From: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Date: Thu, 14 May 2026 05:00:20 -0400 Subject: [PATCH] [Bug] Fix DeepSeek V4 `AttributeError: module 'cutlass.cute.nvgpu' has no attribute 'LoadCacheMode'` (#42342) Signed-off-by: yewentao256 Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by: Roger Wang --- requirements/cuda.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/requirements/cuda.txt b/requirements/cuda.txt index b5f6074d29..43b26e2b42 100644 --- a/requirements/cuda.txt +++ b/requirements/cuda.txt @@ -21,7 +21,7 @@ nvidia-cudnn-frontend>=1.13.0,<1.19.0 fastsafetensors >= 0.2.2 # QuACK and Cutlass DSL for FA4 (cute-DSL implementation) -nvidia-cutlass-dsl[cu13]>=4.4.2 +nvidia-cutlass-dsl[cu13]==4.5.0 quack-kernels>=0.3.3 # Tokenspeed_MLA for faster mla with spec decode