docker: add patch to increase GPU deepep timeout (#18298)
This commit is contained in:
@@ -230,18 +230,21 @@ RUN set -eux; \
|
|||||||
cd DeepEP && \
|
cd DeepEP && \
|
||||||
git checkout ${GRACE_BLACKWELL_DEEPEP_BRANCH} && \
|
git checkout ${GRACE_BLACKWELL_DEEPEP_BRANCH} && \
|
||||||
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
||||||
|
sed -i 's/#define NUM_TIMEOUT_CYCLES 200000000000ull/#define NUM_TIMEOUT_CYCLES 2000000000000ull/' csrc/kernels/configs.cuh && \
|
||||||
cd .. ; \
|
cd .. ; \
|
||||||
elif [ "$HOPPER_SBO" = "1" ]; then \
|
elif [ "$HOPPER_SBO" = "1" ]; then \
|
||||||
git clone https://github.com/deepseek-ai/DeepEP.git -b antgroup-opt && \
|
git clone https://github.com/deepseek-ai/DeepEP.git -b antgroup-opt && \
|
||||||
cd DeepEP && \
|
cd DeepEP && \
|
||||||
git checkout ${HOPPER_SBO_DEEPEP_COMMIT} && \
|
git checkout ${HOPPER_SBO_DEEPEP_COMMIT} && \
|
||||||
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
||||||
|
sed -i 's/#define NUM_TIMEOUT_CYCLES 200000000000ull/#define NUM_TIMEOUT_CYCLES 2000000000000ull/' csrc/kernels/configs.cuh && \
|
||||||
cd .. ; \
|
cd .. ; \
|
||||||
else \
|
else \
|
||||||
curl --retry 3 --retry-delay 2 -fsSL -o ${DEEPEP_COMMIT}.zip \
|
curl --retry 3 --retry-delay 2 -fsSL -o ${DEEPEP_COMMIT}.zip \
|
||||||
https://${GITHUB_ARTIFACTORY}/deepseek-ai/DeepEP/archive/${DEEPEP_COMMIT}.zip && \
|
https://${GITHUB_ARTIFACTORY}/deepseek-ai/DeepEP/archive/${DEEPEP_COMMIT}.zip && \
|
||||||
unzip -q ${DEEPEP_COMMIT}.zip && rm ${DEEPEP_COMMIT}.zip && mv DeepEP-${DEEPEP_COMMIT} DeepEP && cd DeepEP && \
|
unzip -q ${DEEPEP_COMMIT}.zip && rm ${DEEPEP_COMMIT}.zip && mv DeepEP-${DEEPEP_COMMIT} DeepEP && cd DeepEP && \
|
||||||
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
sed -i 's/#define NUM_CPU_TIMEOUT_SECS 100/#define NUM_CPU_TIMEOUT_SECS 1000/' csrc/kernels/configs.cuh && \
|
||||||
|
sed -i 's/#define NUM_TIMEOUT_CYCLES 200000000000ull/#define NUM_TIMEOUT_CYCLES 2000000000000ull/' csrc/kernels/configs.cuh && \
|
||||||
cd .. ; \
|
cd .. ; \
|
||||||
fi
|
fi
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user