EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

Refactor Dockerfile: unify CUDA logic and reduce image size by ~2.6 GB (#3749)

Browse Source 
Signed-off-by: Kebe <mail@kebe7jun.com>
This commit is contained in:
Kebe
2025-03-08 19:01:13 +08:00
committed by GitHub
parent 8d323e95e4
commit 4a893d142d
1 changed files with 7 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
docker/Dockerfile
View File

@@ -30,44 +30,19 @@ ARG CUDA_VERSION
RUN python3 -m pip install --upgrade pip setuptools wheel html5lib six \ RUN python3 -m pip install --upgrade pip setuptools wheel html5lib six \
&& git clone --depth=1 https://github.com/sgl-project/sglang.git \ && git clone --depth=1 https://github.com/sgl-project/sglang.git \
&& if [ "$CUDA_VERSION" = "12.1.1" ]; then \ && if [ "$CUDA_VERSION" = "12.1.1" ]; then \
python3 -m pip install torch --index-url https://download.pytorch.org/whl/cu121; \ export CUINDEX=121; \
elif [ "$CUDA_VERSION" = "12.4.1" ]; then \ elif [ "$CUDA_VERSION" = "12.4.1" ]; then \
python3 -m pip install torch --index-url https://download.pytorch.org/whl/cu124; \ export CUINDEX=124; \
elif [ "$CUDA_VERSION" = "12.5.1" ]; then \ elif [ "$CUDA_VERSION" = "12.5.1" ]; then \
python3 -m pip install torch --index-url https://download.pytorch.org/whl/cu124; \ export CUINDEX=124; \
elif [ "$CUDA_VERSION" = "11.8.0" ]; then \ elif [ "$CUDA_VERSION" = "11.8.0" ]; then \
python3 -m pip install torch --index-url https://download.pytorch.org/whl/cu118; \ export CUINDEX=118; \
python3 -m pip install sgl-kernel -i https://docs.sglang.ai/whl/cu118; \ python3 -m pip install --no-cache-dir sgl-kernel -i https://docs.sglang.ai/whl/cu118; \
else \ else \
echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1; \ echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1; \
fi \ fi \
&& python3 -m pip install --no-cache-dir torch --index-url https://download.pytorch.org/whl/cu${CUINDEX} \
&& cd sglang \ && cd sglang \
&& if [ "$BUILD_TYPE" = "srt" ]; then \ && python3 -m pip --no-cache-dir install -e "python[${BUILD_TYPE}]" --find-links https://flashinfer.ai/whl/cu${CUINDEX}/torch2.5/flashinfer-python
if [ "$CUDA_VERSION" = "12.1.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[srt]" --find-links https://flashinfer.ai/whl/cu121/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "12.4.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[srt]" --find-links https://flashinfer.ai/whl/cu124/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "12.5.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[srt]" --find-links https://flashinfer.ai/whl/cu124/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "11.8.0" ]; then \
python3 -m pip --no-cache-dir install -e "python[srt]" --find-links https://flashinfer.ai/whl/cu118/torch2.5/flashinfer-python; \
python3 -m pip install sgl-kernel -i https://docs.sglang.ai/whl/cu118; \
else \
echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1; \
fi; \
else \
if [ "$CUDA_VERSION" = "12.1.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[all]" --find-links https://flashinfer.ai/whl/cu121/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "12.4.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[all]" --find-links https://flashinfer.ai/whl/cu124/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "12.5.1" ]; then \
python3 -m pip --no-cache-dir install -e "python[all]" --find-links https://flashinfer.ai/whl/cu124/torch2.5/flashinfer-python; \
elif [ "$CUDA_VERSION" = "11.8.0" ]; then \
python3 -m pip --no-cache-dir install -e "python[all]" --find-links https://flashinfer.ai/whl/cu118/torch2.5/flashinfer-python; \
python3 -m pip install sgl-kernel -i https://docs.sglang.ai/whl/cu118; \
else \
echo "Unsupported CUDA version: $CUDA_VERSION" && exit 1; \
fi; \
fi
ENV DEBIAN_FRONTEND=interactive ENV DEBIAN_FRONTEND=interactive