[CI/Build][Docker] Bump nvidia-cutlass-dsl to 4.6.0 and drop packaging workarounds (#47442)

Signed-off-by: Artem Perevedentsev <aperevedents@nvidia.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
This commit is contained in:
Artem Perevedentsev
2026-07-16 13:51:41 +00:00
committed by GitHub
co-authored by Matthew Bonanni
parent b8168e33e0
commit 7cd1d57b74
7 changed files with 8 additions and 120 deletions
+2 -2
View File
@@ -16,7 +16,7 @@ PyNvVideoCodec==2.0.4
--extra-index-url https://flashinfer.ai/whl/
flashinfer-python==0.6.14
flashinfer-cubin==0.6.14
apache-tvm-ffi==0.1.9
apache-tvm-ffi==0.1.10
tilelang==0.1.9
nvidia-cudnn-frontend>=1.19.1
# Required for LLM_NVTX_SCOPES_FOR_PROFILING=1
@@ -25,7 +25,7 @@ nvtx==0.2.15
fastsafetensors >= 0.3.2
# QuACK and Cutlass DSL for FA4 (cute-DSL implementation)
nvidia-cutlass-dsl[cu13]==4.5.2
nvidia-cutlass-dsl[cu13]==4.6.0
quack-kernels>=0.4.0 # Required for tml-fa4
# Tokenspeed_MLA for faster mla with spec decode
+1 -1
View File
@@ -43,7 +43,7 @@ anyio==4.14.1
# sse-starlette
# starlette
# watchfiles
apache-tvm-ffi==0.1.9
apache-tvm-ffi==0.1.10
# via
# -c requirements/cuda.txt
# xgrammar