chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

Try to fix B200 DeepEP error (#14399)

Browse Source
This commit is contained in:
fzyzcjy
2025-12-04 16:41:24 +08:00
committed by GitHub
parent 70d2587324
commit e3ab23c1a6
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docker/Dockerfile
View File

@@ -138,7 +138,7 @@ RUN --mount=type=cache,target=/root/.cache/pip python3 -m pip install --upgrade
# We use Tom's DeepEP fork for GB200 for now; the 1fd57b0276311d035d16176bb0076426166e52f3 commit is https://github.com/fzyzcjy/DeepEP/tree/gb200_blog_part_2
RUN set -eux; \
if [ "${CUDA_VERSION%%.*}" = "12" ]; then \
pip install nvidia-nvshmem-cu12==${NVSHMEM_VERSION} ; \
pip install nvidia-nvshmem-cu12==3.3.20 ; \
elif [ "${CUDA_VERSION%%.*}" = "13" ]; then \
pip install nvidia-nvshmem-cu13==${NVSHMEM_VERSION} ; \
else \
@@ -203,7 +203,7 @@ RUN --mount=type=cache,target=/root/.cache/pip python3 -m pip install \
RUN --mount=type=cache,target=/root/.cache/pip if [ "${CUDA_VERSION%%.*}" = "12" ]; then \
python3 -m pip install nvidia-nccl-cu12==2.28.3 --force-reinstall --no-deps ; \
python3 -m pip install nvidia-cudnn-cu12==9.16.0.29 --force-reinstall --no-deps; \
python3 -m pip install nvidia-nvshmem-cu12==${NVSHMEM_VERSION} --force-reinstall --no-deps; \
python3 -m pip install nvidia-nvshmem-cu12==3.3.20 --force-reinstall --no-deps; \
python3 -m pip install nvidia-cutlass-dsl==4.3.0 --force-reinstall --no-deps; \
elif [ "${CUDA_VERSION%%.*}" = "13" ]; then \
python3 -m pip install nvidia-nccl-cu13==2.28.3 --force-reinstall --no-deps ; \