Use the latest HabanaAI/vllm-fork release tag to build vllm-gaudi image

.github/workflows/_comps-workflow.yml

@@ -70,7 +70,10 @@ jobs:
               cd ./vllm-openvino && git checkout v0.6.1 && git rev-parse HEAD && cd ../
           fi
           if [[ $(grep -c "vllm-gaudi:" ${docker_compose_yml}) != 0 ]]; then
-              git clone --depth 1 --branch v0.6.4.post2+Gaudi-1.19.0 https://github.com/HabanaAI/vllm-fork.git
+              git clone https://github.com/HabanaAI/vllm-fork.git && cd vllm-fork
+              VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+              echo "Check out vLLM tag ${VLLM_VER}"
+              git checkout ${VLLM_VER} &> /dev/null && cd ../
           fi
       - name: Get build list
         id: get-build-list

.github/workflows/push-image-build.yml

@@ -95,7 +95,10 @@ jobs:
               cd ./vllm-openvino && git checkout v0.6.1 && git rev-parse HEAD && cd ../
           fi
           if [[ $(grep -c "vllm-gaudi:" ${docker_compose_path}) != 0 ]]; then
-              git clone --depth 1 --branch v0.6.4.post2+Gaudi-1.19.0 https://github.com/HabanaAI/vllm-fork.git
+              git clone https://github.com/HabanaAI/vllm-fork.git && cd vllm-fork
+              VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+              echo "Check out vLLM tag ${VLLM_VER}"
+              git checkout ${VLLM_VER} &> /dev/null && cd ../
           fi
 
       - name: Build Image

.github/workflows/scripts/get_test_matrix.sh

@@ -10,7 +10,7 @@ cd $WORKSPACE
 changed_files_full=$changed_files_full
 run_matrix="{\"include\":["
 
-# add test services when comps code change
+
 function find_test_1() {
     local pre_service_path=$1
     local n=$2
@@ -97,7 +97,7 @@ function _fill_in_matrix() {
     sleep 1s
 }
 
-# add test case when test scripts code change
+
 function find_test_2() {
     test_files=$(printf '%s\n' "${changed_files[@]}" | grep -E "\.sh") || true
     for test_file in ${test_files}; do
@@ -108,6 +108,7 @@ function find_test_2() {
     done
 }
 
+
 function find_test_3() {
     yaml_files=${changed_files}
     for yaml_file in ${yaml_files}; do
@@ -127,8 +128,10 @@ function find_test_3() {
     done
 }
 
+
 function main() {
 
+    # add test services when comps code change
     changed_files=$(printf '%s\n' "${changed_files_full[@]}" | grep 'comps/' | grep -vE '\.md|comps/cores|comps/third_parties|deployment|\.yaml') || true
     echo "===========start find_test_1============"
     echo "changed_files=${changed_files}"
@@ -137,6 +140,7 @@ function main() {
     echo "run_matrix=${run_matrix}"
     echo "===========finish find_test_1============"
 
+    # add test case when test scripts code change
     changed_files=$(printf '%s\n' "${changed_files_full[@]}" | grep 'tests/' | grep -vE '\.md|\.txt|tests/cores') || true
     echo "===========start find_test_2============"
     echo "changed_files=${changed_files}"
@@ -145,6 +149,7 @@ function main() {
     echo "run_matrix=${run_matrix}"
     echo "===========finish find_test_2============"
 
+    # add test case when docker-compose code change
     changed_files=$(printf '%s\n' "${changed_files_full[@]}" | grep 'deployment/docker_compose/compose' | grep '.yaml') || true
     echo "===========start find_test_3============"
     echo "changed_files=${changed_files}"

comps/agent/src/README.md

@@ -131,7 +131,8 @@ export vllm_volume=${YOUR_LOCAL_DIR_FOR_MODELS}
 # build vLLM image
 git clone https://github.com/HabanaAI/vllm-fork.git
 cd ./vllm-fork
-git checkout v0.6.4.post2+Gaudi-1.19.0
+VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+git checkout ${VLLM_VER} &> /dev/null
 docker build -f Dockerfile.hpu -t opea/vllm-gaudi:latest --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
 
 # vllm serving on 4 Gaudi2 cards

comps/third_parties/vllm/src/build_docker_vllm.sh

@@ -35,9 +35,10 @@ fi
 
 # Build the docker image for vLLM based on the hardware mode
 if [ "$hw_mode" = "hpu" ]; then
-    git clone https://github.com/HabanaAI/vllm-fork.git
-    cd ./vllm-fork/
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    git clone https://github.com/HabanaAI/vllm-fork.git && cd vllm-fork
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build -f Dockerfile.hpu -t opea/vllm-gaudi:latest --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
     cd ..
     rm -rf vllm-fork

tests/agent/build_vllm_gaudi.sh

@@ -9,9 +9,9 @@ function build_vllm_docker_images() {
         git clone https://github.com/HabanaAI/vllm-fork.git
     fi
     cd ./vllm-fork
-    # git fetch --all
-    # git checkout v0.6.4.post2+Gaudi-1.19.0
-    # sed -i 's/triton/triton==3.1.0/g' requirements-hpu.txt
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:comps --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi:comps failed"

tests/agent/sql_agent_test/test_sql_agent.sh

@@ -108,6 +108,9 @@ function build_vllm_docker_images() {
         git clone https://github.com/HabanaAI/vllm-fork.git
     fi
     cd ./vllm-fork
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:comps --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi:comps failed"

tests/agent/test_agent_langchain_on_intel_hpu.sh

@@ -59,7 +59,9 @@ function build_vllm_docker_images() {
         git clone https://github.com/HabanaAI/vllm-fork.git
     fi
     cd ./vllm-fork
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t opea/vllm-gaudi:comps --shm-size=128g . --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi:comps failed"

tests/guardrails/test_guardrails_hallucination_detection_on_intel_hpu.sh

@@ -12,7 +12,9 @@ function build_docker_images() {
     cd $WORKPATH
     git clone https://github.com/HabanaAI/vllm-fork.git
     cd vllm-fork/
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache --build-arg https_proxy=$https_proxy --build-arg http_proxy=$http_proxy -f Dockerfile.hpu -t opea/vllm-gaudi:comps --shm-size=128g .
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi built fail"
@@ -44,8 +46,16 @@ function start_service() {
     cd $WORKPATH
     cd comps/guardrails/deployment/docker_compose/
     docker compose up ${service_name} -d
+    if [ $? -ne 0 ]; then
+        echo "Microservice failed to start!"
+        for service in $service_name; do
+            echo "Logs for $service..."
+            docker logs $service
+        done
+        exit 1
+    fi
     echo "Microservice started"
-    sleep 15
+    sleep 1m
 }
 
 function validate_microservice() {

tests/llms/test_llms_doc-summarization_vllm_on_intel_hpu.sh

@@ -19,7 +19,9 @@ function build_docker_images() {
     cd $WORKPATH
     git clone https://github.com/HabanaAI/vllm-fork.git
     cd vllm-fork/
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t ${REGISTRY:-opea}/vllm-gaudi:${TAG:-latest} --shm-size=128g .
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi built fail"
@@ -54,10 +56,11 @@ function start_service() {
     cd $WORKPATH/comps/llms/deployment/docker_compose
     docker compose -f compose_doc-summarization.yaml up ${service_name} -d > ${LOG_PATH}/start_services_with_compose.log
 
-    sleep 30s
+    sleep 1m
 }
 
 function validate_services() {
+    date
     local URL="$1"
     local EXPECTED_RESULT="$2"
     local SERVICE_NAME="$3"
@@ -87,6 +90,7 @@ function validate_services() {
         docker logs ${DOCKER_NAME} >> ${LOG_PATH}/${SERVICE_NAME}.log
         exit 1
     fi
+    date
     sleep 1s
 }
 

tests/llms/test_llms_faq-generation_tgi_on_intel_hpu.sh

@@ -38,8 +38,16 @@ function start_service() {
 
     cd $WORKPATH/comps/llms/deployment/docker_compose
     docker compose -f compose_faq-generation.yaml up ${service_name} -d > ${LOG_PATH}/start_services_with_compose.log
-
-    sleep 30s
+    if [ $? -ne 0 ]; then
+        echo "Microservice failed to start!"
+        for service in $service_name; do
+            echo "Logs for $service..."
+            docker logs $service
+        done
+        exit 1
+    fi
+    echo "Microservice started"
+    sleep 1m
 }
 
 function validate_services() {

tests/llms/test_llms_faq-generation_vllm_on_intel_hpu.sh

@@ -19,7 +19,9 @@ function build_docker_images() {
     cd $WORKPATH
     git clone https://github.com/HabanaAI/vllm-fork.git
     cd vllm-fork/
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t ${REGISTRY:-opea}/vllm-gaudi:${TAG:-latest} --shm-size=128g .
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi built fail"

tests/llms/test_llms_text-generation_service_vllm_on_intel_hpu.sh

@@ -19,7 +19,9 @@ function build_docker_images() {
     cd $WORKPATH
     git clone https://github.com/HabanaAI/vllm-fork.git
     cd vllm-fork/
-    git checkout v0.6.4.post2+Gaudi-1.19.0
+    VLLM_VER=$(git describe --tags "$(git rev-list --tags --max-count=1)")
+    echo "Check out vLLM tag ${VLLM_VER}"
+    git checkout ${VLLM_VER} &> /dev/null
     docker build --no-cache -f Dockerfile.hpu -t ${REGISTRY:-opea}/vllm-gaudi:${TAG:-latest} --shm-size=128g .
     if [ $? -ne 0 ]; then
         echo "opea/vllm-gaudi built fail"