Adding files to deploy SearchQnA application on ROCm vLLM

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

@@ -8,9 +8,9 @@ services:
     image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.5
     container_name: search-tei-embedding-server
     ports:
-      - "3001:80"
+      - "${SEARCH_TEI_EMBEDDING_PORT:-3001}:80"
     volumes:
-      - "${MODEL_PATH:-./data}:/data"
+      - "${MODEL_CACHE:-./data}:/data"
     shm_size: 1g
     environment:
       no_proxy: ${no_proxy}
@@ -20,13 +20,14 @@ services:
       HUGGING_FACE_HUB_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
       HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
     command: --model-id ${SEARCH_EMBEDDING_MODEL_ID} --auto-truncate
+
   search-embedding:
     image: ${REGISTRY:-opea}/embedding:${TAG:-latest}
     container_name: search-embedding-server
     depends_on:
       - search-tei-embedding-service
     ports:
-      - "3002:6000"
+      - "${SEARCH_EMBEDDING_SERVICE_PORT:-3002}:6000"
     ipc: host
     environment:
       no_proxy: ${no_proxy}
@@ -36,11 +37,12 @@ services:
       TEI_EMBEDDING_ENDPOINT: ${SEARCH_TEI_EMBEDDING_ENDPOINT}
       HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
     restart: unless-stopped
+
   search-web-retriever:
     image: ${REGISTRY:-opea}/web-retriever:${TAG:-latest}
     container_name: search-web-retriever-server
     ports:
-      - "3003:7077"
+      - "${SEARCH_WEB_RETRIEVER_SERVICE_PORT:-3003}:7077"
     ipc: host
     environment:
       no_proxy: ${no_proxy}
@@ -50,26 +52,28 @@ services:
       GOOGLE_API_KEY: ${SEARCH_GOOGLE_API_KEY}
       GOOGLE_CSE_ID: ${SEARCH_GOOGLE_CSE_ID}
     restart: unless-stopped
+
   search-tei-reranking-service:
     image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.5
     container_name: search-tei-reranking-server
     ports:
-      - "3004:80"
+      - "${SEARCH_TEI_RERANKING_PORT:-3004}:80"
     volumes:
-      - "${MODEL_PATH:-./data}:/data"
+      - "${MODEL_CACHE:-./data}:/data"
     shm_size: 1g
     environment:
       no_proxy: ${no_proxy}
       http_proxy: ${http_proxy}
       https_proxy: ${https_proxy}
     command: --model-id ${SEARCH_RERANK_MODEL_ID} --auto-truncate
+
   search-reranking:
     image: ${REGISTRY:-opea}/reranking:${TAG:-latest}
     container_name: search-reranking-server
     depends_on:
       - search-tei-reranking-service
     ports:
-      - "3005:8000"
+      - "${SEARCH_RERANK_SERVICE_PORT:-3005}:8000"
     ipc: host
     environment:
       no_proxy: ${no_proxy}
@@ -80,13 +84,14 @@ services:
       HUGGING_FACE_HUB_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
       HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
     restart: unless-stopped
+
   search-tgi-service:
     image: ghcr.io/huggingface/text-generation-inference:2.3.1-rocm
     container_name: search-tgi-service
     ports:
-      - "3006:80"
+      - "${SEARCH_TGI_SERVICE_PORT:-3006}:80"
     volumes:
-      - "${MODEL_PATH:-./data}:/data"
+      - "${MODEL_CACHE:-./data}:/data"
     environment:
       no_proxy: ${no_proxy}
       http_proxy: ${http_proxy}
@@ -96,7 +101,7 @@ services:
     shm_size: 1g
     devices:
       - /dev/kfd:/dev/kfd
-      - /dev/dri/:/dev/dri/
+      - /dev/dri:/dev/dri
     cap_add:
       - SYS_PTRACE
     group_add:
@@ -105,25 +110,26 @@ services:
       - seccomp:unconfined
     ipc: host
     command: --model-id ${SEARCH_LLM_MODEL_ID} --max-input-length 1024 --max-total-tokens 2048
+
   search-llm:
     image: ${REGISTRY:-opea}/llm-textgen:${TAG:-latest}
     container_name: search-llm-server
     depends_on:
       - search-tgi-service
     ports:
-      - "3007:9000"
+      - "${SEARCH_LLM_SERVICE_PORT:-3007}:9000"
     ipc: host
     environment:
       no_proxy: ${no_proxy}
       http_proxy: ${http_proxy}
       https_proxy: ${https_proxy}
-      TGI_LLM_ENDPOINT: ${SEARCH_TGI_LLM_ENDPOINT}
       HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
       LLM_ENDPOINT: ${SEARCH_TGI_LLM_ENDPOINT}
       LLM_MODEL_ID: ${SEARCH_LLM_MODEL_ID}
       LLM_MODEL: ${SEARCH_LLM_MODEL_ID}
       HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
-      OPENAI_API_KEY: ${SEARCH_OPENAI_API_KEY}
+      LLM_COMPONENT_NAME: "OpeaTextGenService"
+
     restart: unless-stopped
   search-backend-server:
     image: ${REGISTRY:-opea}/searchqna:${TAG:-latest}
@@ -139,18 +145,18 @@ services:
     ports:
       - "${SEARCH_BACKEND_SERVICE_PORT:-3008}:8888"
     environment:
-      - no_proxy=${no_proxy}
-      - https_proxy=${https_proxy}
-      - http_proxy=${http_proxy}
-      - MEGA_SERVICE_HOST_IP=${SEARCH_MEGA_SERVICE_HOST_IP}
-      - EMBEDDING_SERVICE_HOST_IP=${SEARCH_EMBEDDING_SERVICE_HOST_IP}
-      - WEB_RETRIEVER_SERVICE_HOST_IP=${SEARCH_WEB_RETRIEVER_SERVICE_HOST_IP}
-      - RERANK_SERVICE_HOST_IP=${SEARCH_RERANK_SERVICE_HOST_IP}
-      - LLM_SERVICE_HOST_IP=${SEARCH_LLM_SERVICE_HOST_IP}
-      - EMBEDDING_SERVICE_PORT=${SEARCH_EMBEDDING_SERVICE_PORT}
-      - WEB_RETRIEVER_SERVICE_PORT=${SEARCH_WEB_RETRIEVER_SERVICE_PORT}
-      - RERANK_SERVICE_PORT=${SEARCH_RERANK_SERVICE_PORT}
-      - LLM_SERVICE_PORT=${SEARCH_LLM_SERVICE_PORT}
+      no_proxy: ${no_proxy}
+      https_proxy: ${https_proxy}
+      http_proxy: ${http_proxy}
+      MEGA_SERVICE_HOST_IP: ${SEARCH_MEGA_SERVICE_HOST_IP}
+      EMBEDDING_SERVICE_HOST_IP: ${SEARCH_EMBEDDING_SERVICE_HOST_IP}
+      WEB_RETRIEVER_SERVICE_HOST_IP: ${SEARCH_WEB_RETRIEVER_SERVICE_HOST_IP}
+      RERANK_SERVICE_HOST_IP: ${SEARCH_RERANK_SERVICE_HOST_IP}
+      LLM_SERVICE_HOST_IP: ${SEARCH_LLM_SERVICE_HOST_IP}
+      EMBEDDING_SERVICE_PORT: ${SEARCH_EMBEDDING_SERVICE_PORT}
+      WEB_RETRIEVER_SERVICE_PORT: ${SEARCH_WEB_RETRIEVER_SERVICE_PORT}
+      RERANK_SERVICE_PORT: ${SEARCH_RERANK_SERVICE_PORT}
+      LLM_SERVICE_PORT: ${SEARCH_LLM_SERVICE_PORT}
     ipc: host
     restart: always
   search-ui-server:
@@ -161,10 +167,10 @@ services:
     ports:
       - "${SEARCH_FRONTEND_SERVICE_PORT:-5173}:5173"
     environment:
-      - no_proxy=${no_proxy}
-      - https_proxy=${https_proxy}
-      - http_proxy=${http_proxy}
-      - BACKEND_BASE_URL=${SEARCH_BACKEND_SERVICE_ENDPOINT}
+      no_proxy: ${no_proxy}
+      https_proxy: ${https_proxy}
+      http_proxy: ${http_proxy}
+      BACKEND_BASE_URL: ${SEARCH_BACKEND_SERVICE_ENDPOINT}
     ipc: host
     restart: always
 

SearchQnA/docker_compose/amd/gpu/rocm/compose_vllm.yaml

@@ -0,0 +1,185 @@
+# Copyright (C) 2024 Intel Corporation
+# SPDX-License-Identifier: Apache-2.0
+
+# SPDX-License-Identifier: Apache-2.0
+
+services:
+  search-tei-embedding-service:
+    image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.5
+    container_name: search-tei-embedding-server
+    ports:
+      - "${SEARCH_TEI_EMBEDDING_PORT:-3001}:80"
+    volumes:
+      - "${MODEL_CACHE:-./data}:/data"
+    shm_size: 1g
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HUGGING_FACE_HUB_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+    command: --model-id ${SEARCH_EMBEDDING_MODEL_ID} --auto-truncate
+
+  search-embedding:
+    image: ${REGISTRY:-opea}/embedding:${TAG:-latest}
+    container_name: search-embedding-server
+    depends_on:
+      - search-tei-embedding-service
+    ports:
+      - "${SEARCH_EMBEDDING_SERVICE_PORT:-3002}:6000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      TEI_EMBEDDING_HOST_IP: ${SEARCH_HOST_IP}
+      TEI_EMBEDDING_PORT: ${SEARCH_TEI_EMBEDDING_PORT}
+      TEI_EMBEDDING_ENDPOINT: ${SEARCH_TEI_EMBEDDING_ENDPOINT}
+      HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+    restart: unless-stopped
+
+  search-web-retriever:
+    image: ${REGISTRY:-opea}/web-retriever:${TAG:-latest}
+    container_name: search-web-retriever-server
+    ports:
+      - "${SEARCH_WEB_RETRIEVER_SERVICE_PORT:-3003}:7077"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      TEI_EMBEDDING_ENDPOINT: ${SEARCH_TEI_EMBEDDING_ENDPOINT}
+      GOOGLE_API_KEY: ${SEARCH_GOOGLE_API_KEY}
+      GOOGLE_CSE_ID: ${SEARCH_GOOGLE_CSE_ID}
+    restart: unless-stopped
+
+  search-tei-reranking-service:
+    image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.5
+    container_name: search-tei-reranking-server
+    ports:
+      - "${SEARCH_TEI_RERANKING_PORT:-3004}:80"
+    volumes:
+      - "${MODEL_CACHE:-./data}:/data"
+    shm_size: 1g
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+    command: --model-id ${SEARCH_RERANK_MODEL_ID} --auto-truncate
+
+  search-reranking:
+    image: ${REGISTRY:-opea}/reranking:${TAG:-latest}
+    container_name: search-reranking-server
+    depends_on:
+      - search-tei-reranking-service
+    ports:
+      - "${SEARCH_RERANK_SERVICE_PORT:-3005}:8000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      TEI_RERANKING_ENDPOINT: ${SEARCH_TEI_RERANKING_ENDPOINT}
+      HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HUGGING_FACE_HUB_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+    restart: unless-stopped
+
+  search-vllm-service:
+    image: ${REGISTRY:-opea}/vllm-rocm:${TAG:-latest}
+    container_name: search-vllm-service
+    ports:
+      - "${SEARCH_VLLM_SERVICE_PORT:-8081}:8011"
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      HF_HUB_DISABLE_PROGRESS_BARS: 1
+      HF_HUB_ENABLE_HF_TRANSFER: 0
+      WILM_USE_TRITON_FLASH_ATTENTION: 0
+      PYTORCH_JIT: 0
+    volumes:
+      - "${MODEL_CACHE:-./data}:/data"
+    shm_size: 20G
+    devices:
+      - /dev/kfd:/dev/kfd
+      - /dev/dri/:/dev/dri/
+    cap_add:
+      - SYS_PTRACE
+    group_add:
+      - video
+    security_opt:
+      - seccomp:unconfined
+      - apparmor=unconfined
+    command: "--model ${SEARCH_LLM_MODEL_ID} --swap-space 16 --disable-log-requests --dtype float16 --tensor-parallel-size 4 --host 0.0.0.0 --port 8011 --num-scheduler-steps 1 --distributed-executor-backend \"mp\""
+    ipc: host
+
+  search-llm:
+    image: ${REGISTRY:-opea}/llm-textgen:${TAG:-latest}
+    container_name: search-llm-server
+    depends_on:
+      - search-vllm-service
+    ports:
+      - "${SEARCH_LLM_SERVICE_PORT:-3007}:9000"
+    ipc: host
+    environment:
+      no_proxy: ${no_proxy}
+      http_proxy: ${http_proxy}
+      https_proxy: ${https_proxy}
+      LLM_ENDPOINT: ${SEARCH_LLM_ENDPOINT}
+      HUGGINGFACEHUB_API_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      LLM_MODEL_ID: ${SEARCH_LLM_MODEL_ID}
+      HF_TOKEN: ${SEARCH_HUGGINGFACEHUB_API_TOKEN}
+      LLM_COMPONENT_NAME: "OpeaTextGenService"
+    restart: unless-stopped
+
+  search-backend-server:
+    image: ${REGISTRY:-opea}/searchqna:${TAG:-latest}
+    container_name: search-backend-server
+    depends_on:
+      - search-tei-embedding-service
+      - search-embedding
+      - search-web-retriever
+      - search-tei-reranking-service
+      - search-reranking
+      - search-vllm-service
+      - search-llm
+    ports:
+      - "${SEARCH_BACKEND_SERVICE_PORT:-3008}:8888"
+    environment:
+      no_proxy: ${no_proxy}
+      https_proxy: ${https_proxy}
+      http_proxy: ${http_proxy}
+      MEGA_SERVICE_HOST_IP: ${SEARCH_MEGA_SERVICE_HOST_IP}
+      EMBEDDING_SERVICE_HOST_IP: ${SEARCH_EMBEDDING_SERVICE_HOST_IP}
+      EMBEDDING_SERVICE_PORT: ${SEARCH_EMBEDDING_SERVICE_PORT}
+      WEB_RETRIEVER_SERVICE_HOST_IP: ${SEARCH_WEB_RETRIEVER_SERVICE_HOST_IP}
+      WEB_RETRIEVER_SERVICE_PORT: ${SEARCH_WEB_RETRIEVER_SERVICE_PORT}
+      RERANK_SERVICE_HOST_IP: ${SEARCH_RERANK_SERVICE_HOST_IP}
+      RERANK_SERVICE_PORT: ${SEARCH_RERANK_SERVICE_PORT}
+      LLM_SERVICE_HOST_IP: ${SEARCH_LLM_SERVICE_HOST_IP}
+      LLM_SERVICE_PORT: ${SEARCH_LLM_SERVICE_PORT}
+    ipc: host
+    restart: always
+
+  search-ui-server:
+    image: ${REGISTRY:-opea}/searchqna-ui:${TAG:-latest}
+    container_name: search-ui-server
+    depends_on:
+      - search-backend-server
+    ports:
+      - "${SEARCH_FRONTEND_SERVICE_PORT:-5173}:5173"
+    environment:
+      no_proxy: ${no_proxy}
+      https_proxy: ${https_proxy}
+      http_proxy: ${http_proxy}
+      BACKEND_BASE_URL: ${SEARCH_BACKEND_SERVICE_ENDPOINT}
+    ipc: host
+    restart: always
+
+networks:
+  default:
+    driver: bridge