biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[NVIDIA] Bugfix NVFP4 DGX Spark and RTX50 (#38423)

Browse Source 
Signed-off-by: johnnynunez <johnnynuca14@gmail.com>
Signed-off-by: Johnny <johnnynuca14@gmail.com>
This commit is contained in:
Johnny
2026-03-30 18:36:18 +02:00
committed by GitHub
parent 8e6293e838
commit b4a2f3ac36
15 changed files with 86 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
docker/Dockerfile
View File

@@ -590,7 +590,10 @@ RUN --mount=type=cache,target=/root/.cache/uv \
# Install FlashInfer JIT cache (requires CUDA-version-specific index URL)
# https://docs.flashinfer.ai/installation.html
# From versions.json: .flashinfer.version
ARG FLASHINFER_VERSION=0.6.6
# 0.6.7: CUTLASS 4.4.2 bump, fixes TMA grouped GEMM on SM12x (flashinfer#2798)
# TODO: bump to 0.6.8 when released for NVFP4/MXFP4 group GEMMs on
# SM120/SM121 (RTX 50 / DGX Spark) via flashinfer#2738
ARG FLASHINFER_VERSION=0.6.7
RUN --mount=type=cache,target=/root/.cache/uv \
uv pip install --system flashinfer-jit-cache==${FLASHINFER_VERSION} \
--extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \