Update model cache for AgentQnA (#1627)

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

AgentQnA/docker_compose/amd/gpu/rocm/compose.yaml

@@ -8,7 +8,7 @@ services:
     ports:
       - "${AGENTQNA_TGI_SERVICE_PORT-8085}:80"
     volumes:
-      - /var/opea/agent-service/:/data
+      - ${HF_CACHE_DIR:-/var/opea/agent-service/}:/data
     environment:
       no_proxy: ${no_proxy}
       http_proxy: ${http_proxy}

AgentQnA/tests/step2_start_retrieval_tool.sh

@@ -9,7 +9,7 @@ echo "WORKDIR=${WORKDIR}"
 export ip_address=$(hostname -I | awk '{print $1}')
 export host_ip=${ip_address}
 
-export HF_CACHE_DIR=$WORKDIR/hf_cache
+export HF_CACHE_DIR=${model_cache:-"$WORKDIR/hf_cache"}
 if [ ! -d "$HF_CACHE_DIR" ]; then
     echo "Creating HF_CACHE directory"
     mkdir -p "$HF_CACHE_DIR"

AgentQnA/tests/step4_launch_and_validate_agent_gaudi.sh

@@ -13,7 +13,7 @@ export HUGGINGFACEHUB_API_TOKEN=${HUGGINGFACEHUB_API_TOKEN}
 HF_TOKEN=${HUGGINGFACEHUB_API_TOKEN}
 model="meta-llama/Llama-3.3-70B-Instruct" #"meta-llama/Meta-Llama-3.1-70B-Instruct"
 
-export HF_CACHE_DIR=/data2/huggingface
+export HF_CACHE_DIR=${model_cache:-"/data2/huggingface"}
 if [ ! -d "$HF_CACHE_DIR" ]; then
     HF_CACHE_DIR=$WORKDIR/hf_cache
     mkdir -p "$HF_CACHE_DIR"

AgentQnA/tests/step4a_launch_and_validate_agent_tgi_on_rocm.sh

@@ -11,7 +11,7 @@ export ip_address=$(hostname -I | awk '{print $1}')
 export TOOLSET_PATH=$WORKDIR/GenAIExamples/AgentQnA/tools/
 export HUGGINGFACEHUB_API_TOKEN=${HUGGINGFACEHUB_API_TOKEN}
 
-export HF_CACHE_DIR=$WORKDIR/hf_cache
+export HF_CACHE_DIR=${model_cache:-"$WORKDIR/hf_cache"}
 if [ ! -d "$HF_CACHE_DIR" ]; then
     mkdir -p "$HF_CACHE_DIR"
 fi