From d6312a73b9e4c9144afb3cc53dcd134038aa1920 Mon Sep 17 00:00:00 2001 From: XinyaoWa Date: Thu, 16 Jan 2025 17:48:06 +0800 Subject: [PATCH] Fix vllm hpu to a stable release (#1156) There exist risks with vllm-fork main branch, change to latest stable release v0.6.4.post2+Gaudi-1.19.0 Signed-off-by: Xinyao Wang --- .github/workflows/_comps-workflow.yml | 1 + comps/third_parties/vllm/src/build_docker_vllm.sh | 1 + .../test_llms_doc-summarization_langchain_vllm_on_intel_hpu.sh | 1 + .../llms/test_llms_faq-generation_langchain_vllm_on_intel_hpu.sh | 1 + tests/llms/test_llms_text-generation_opea_vllm_on_intel_hpu.sh | 1 + 5 files changed, 5 insertions(+) diff --git a/.github/workflows/_comps-workflow.yml b/.github/workflows/_comps-workflow.yml index 0e693e75e..86b41239c 100644 --- a/.github/workflows/_comps-workflow.yml +++ b/.github/workflows/_comps-workflow.yml @@ -65,6 +65,7 @@ jobs: fi if [[ $(grep -c "vllm-gaudi:" ${docker_compose_yml}) != 0 ]]; then git clone https://github.com/HabanaAI/vllm-fork.git vllm-fork + cd vllm-fork && git checkout v0.6.4.post2+Gaudi-1.19.0 && cd ../ fi - name: Get build list id: get-build-list diff --git a/comps/third_parties/vllm/src/build_docker_vllm.sh b/comps/third_parties/vllm/src/build_docker_vllm.sh index 20d4f8df5..bec3a0c8f 100644 --- a/comps/third_parties/vllm/src/build_docker_vllm.sh +++ b/comps/third_parties/vllm/src/build_docker_vllm.sh @@ -37,6 +37,7 @@ fi if [ "$hw_mode" = "hpu" ]; then git clone https://github.com/HabanaAI/vllm-fork.git cd ./vllm-fork/ + git checkout v0.6.4.post2+Gaudi-1.19.0 docker build -f Dockerfile.hpu -t opea/vllm-gaudi:latest --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy cd .. rm -rf vllm-fork diff --git a/tests/llms/test_llms_doc-summarization_langchain_vllm_on_intel_hpu.sh b/tests/llms/test_llms_doc-summarization_langchain_vllm_on_intel_hpu.sh index 0e97d8e13..8dc3952cd 100644 --- a/tests/llms/test_llms_doc-summarization_langchain_vllm_on_intel_hpu.sh +++ b/tests/llms/test_llms_doc-summarization_langchain_vllm_on_intel_hpu.sh @@ -12,6 +12,7 @@ function build_docker_images() { cd $WORKPATH git clone https://github.com/HabanaAI/vllm-fork.git cd vllm-fork/ + git checkout v0.6.4.post2+Gaudi-1.19.0 docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:latest --shm-size=128g . if [ $? -ne 0 ]; then echo "opea/vllm-gaudi built fail" diff --git a/tests/llms/test_llms_faq-generation_langchain_vllm_on_intel_hpu.sh b/tests/llms/test_llms_faq-generation_langchain_vllm_on_intel_hpu.sh index 57d4f4207..0af6e7133 100644 --- a/tests/llms/test_llms_faq-generation_langchain_vllm_on_intel_hpu.sh +++ b/tests/llms/test_llms_faq-generation_langchain_vllm_on_intel_hpu.sh @@ -12,6 +12,7 @@ function build_docker_images() { cd $WORKPATH git clone https://github.com/HabanaAI/vllm-fork.git cd vllm-fork/ + git checkout v0.6.4.post2+Gaudi-1.19.0 docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:latest --shm-size=128g . if [ $? -ne 0 ]; then echo "opea/vllm-gaudi built fail" diff --git a/tests/llms/test_llms_text-generation_opea_vllm_on_intel_hpu.sh b/tests/llms/test_llms_text-generation_opea_vllm_on_intel_hpu.sh index 05c644ef0..cd37b88d2 100644 --- a/tests/llms/test_llms_text-generation_opea_vllm_on_intel_hpu.sh +++ b/tests/llms/test_llms_text-generation_opea_vllm_on_intel_hpu.sh @@ -12,6 +12,7 @@ function build_docker_images() { cd $WORKPATH git clone https://github.com/HabanaAI/vllm-fork.git cd vllm-fork/ + git checkout v0.6.4.post2+Gaudi-1.19.0 docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:comps --shm-size=128g . if [ $? -ne 0 ]; then echo "opea/vllm-gaudi built fail"