Use model cache for llms docker compose test. (opea-project#1463)

Signed-off-by: ZePan110 <ze.pan@intel.com>

Author: ZePan110

tests/llms/test_llms_doc-summarization_tgi.sh

@@ -9,6 +9,7 @@ export REGISTRY=${IMAGE_REPO}
 export TAG="comps"
 echo "REGISTRY=IMAGE_REPO=${IMAGE_REPO}"
 echo "TAG=${TAG}"
+export DATA_PATH=${model_cache}
 
 WORKPATH=$(dirname "$PWD")
 host_ip=$(hostname -I | awk '{print $1}')

tests/llms/test_llms_doc-summarization_tgi_on_intel_hpu.sh

@@ -36,7 +36,7 @@ function start_service() {
     export MAX_INPUT_TOKENS=2048
     export MAX_TOTAL_TOKENS=4096
     export LOGFLAG=True
-    export DATA_PATH="/data2/hf_model"
+    export DATA_PATH=${model_cache:-./data}
 
     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

tests/llms/test_llms_doc-summarization_vllm.sh

@@ -9,6 +9,7 @@ export REGISTRY=${IMAGE_REPO}
 export TAG="comps"
 echo "REGISTRY=IMAGE_REPO=${IMAGE_REPO}"
 echo "TAG=${TAG}"
+export DATA_PATH=${model_cache}
 
 WORKPATH=$(dirname "$PWD")
 host_ip=$(hostname -I | awk '{print $1}')

tests/llms/test_llms_doc-summarization_vllm_on_intel_hpu.sh

@@ -51,7 +51,7 @@ function start_service() {
     export MAX_TOTAL_TOKENS=4096
     export VLLM_SKIP_WARMUP=true
     export LOGFLAG=True
-    export DATA_PATH="/data2/hf_model"
+    export DATA_PATH=${model_cache:-./data}
 
     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

tests/llms/test_llms_faq-generation_tgi.sh

@@ -9,6 +9,7 @@ export REGISTRY=${IMAGE_REPO}
 export TAG="comps"
 echo "REGISTRY=IMAGE_REPO=${IMAGE_REPO}"
 echo "TAG=${TAG}"
+export DATA_PATH=${model_cache}
 
 WORKPATH=$(dirname "$PWD")
 host_ip=$(hostname -I | awk '{print $1}')

tests/llms/test_llms_faq-generation_tgi_on_intel_hpu.sh

@@ -34,7 +34,7 @@ function start_service() {
     export LLM_ENDPOINT="http://${host_ip}:${LLM_ENDPOINT_PORT}"
     export LLM_MODEL_ID="Intel/neural-chat-7b-v3-3"
     export LOGFLAG=True
-    export DATA_PATH="/data2/hf_model"
+    export DATA_PATH=${model_cache:-./data}
 
     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

tests/llms/test_llms_faq-generation_vllm.sh

@@ -9,6 +9,7 @@ export REGISTRY=${IMAGE_REPO}
 export TAG="comps"
 echo "REGISTRY=IMAGE_REPO=${IMAGE_REPO}"
 echo "TAG=${TAG}"
+export DATA_PATH=${model_cache}
 
 WORKPATH=$(dirname "$PWD")
 host_ip=$(hostname -I | awk '{print $1}')

tests/llms/test_llms_faq-generation_vllm_on_intel_hpu.sh

@@ -50,7 +50,7 @@ function start_service() {
     export LLM_MODEL_ID="Intel/neural-chat-7b-v3-3"
     export VLLM_SKIP_WARMUP=true
     export LOGFLAG=True
-    export DATA_PATH="/data2/hf_model"
+    export DATA_PATH=${model_cache:-./data}
 
     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

tests/llms/test_llms_text-generation_native_on_intel_hpu.sh

@@ -31,7 +31,7 @@ function start_service() {
     export host_ip=${host_ip}
     export LLM_MODEL_ID="Intel/neural-chat-7b-v3-3"
     export LOGFLAG=True
-    export DATA_PATH="/data2/hf_model"
+    export DATA_PATH=${model_cache:-./data}
 
     cd $WORKPATH/comps/llms/deployment/docker_compose
     docker compose -f compose_text-generation.yaml up ${service_name} -d > ${LOG_PATH}/start_services_with_compose.log

tests/llms/test_llms_text-generation_service_tgi.sh

@@ -9,6 +9,7 @@ export REGISTRY=${IMAGE_REPO}
 export TAG="comps"
 echo "REGISTRY=IMAGE_REPO=${IMAGE_REPO}"
 echo "TAG=${TAG}"
+export DATA_PATH=${model_cache}
 
 WORKPATH=$(dirname "$PWD")
 host_ip=$(hostname -I | awk '{print $1}')