Update config, Dockerfiles, and grafana settings

Author: YongZhuIntel

config/config.yaml

@@ -45,17 +45,28 @@ prompt_guard:
 # Supported formats: 127.0.0.1, 192.168.1.1, ::1, 2001:db8::1
 # NOT supported: domain names (example.com), protocol prefixes (http://), paths (/api), ports in address (use 'port' field)
 vllm_endpoints:
-  - name: "endpoint1"
-    address: "172.28.0.20"  # Static IPv4 of llm-katan within docker compose network
-    port: 8002
-    weight: 1
+  - name: "math-endpoint"  
+    address: "172.28.0.30"      # 数学模型服务器 IP  
+    port: 8006                # 数学模型端口  
+    weight: 1  
+    
+  - name: "text-endpoint"  
+    address: "172.28.0.30"      # 文字模型服务器 IP  
+    port: 8007                # 文字模型端口  
+    weight: 1  
 
-model_config:
-  "qwen3":
-    reasoning_family: "qwen3"  # This model uses Qwen-3 reasoning syntax
-    preferred_endpoints: ["endpoint1"]  # Optional: omit to let upstream handle endpoint selection
-    pii_policy:
-      allow_by_default: true
+model_config:  
+  "DeepSeek-R1-Distill-Qwen-7B":  
+    reasoning_family: "deepseek"    # DeepSeek 使用 deepseek reasoning 语法  
+    preferred_endpoints: ["math-endpoint"]  
+    pii_policy:  
+      allow_by_default: true  
+    
+  "qwen3":  
+    reasoning_family: "qwen3"       # Qwen3 使用 qwen3 reasoning 语法  
+    preferred_endpoints: ["text-endpoint"]  
+    pii_policy:  
+      allow_by_default: true  
 
 # Classifier configuration
 classifier:
@@ -139,7 +150,7 @@ categories:
   - name: math
     system_prompt: "You are a mathematics expert. Provide step-by-step solutions, show your work clearly, and explain mathematical concepts in an understandable way."
     model_scores:
-      - model: qwen3
+      - model: DeepSeek-R1-Distill-Qwen-7B
         score: 1.0
         use_reasoning: true  # Enable reasoning for complex math
   - name: physics

dashboard/Dockerfile

@@ -1,5 +1,7 @@
 # Build frontend
 FROM node:18-alpine AS frontend-builder
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 WORKDIR /app/frontend
 COPY frontend/package*.json ./
 RUN npm ci
@@ -8,6 +10,8 @@ RUN npm run build
 
 # Build backend
 FROM golang:1.21-alpine AS backend-builder
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 WORKDIR /app/backend
 COPY backend/go.* ./
 RUN go mod download
@@ -16,6 +20,8 @@ RUN CGO_ENABLED=0 GOOS=linux go build -o dashboard-server .
 
 # Final image
 FROM alpine:3.18
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 RUN apk add --no-cache ca-certificates
 WORKDIR /app
 COPY --from=backend-builder /app/backend/dashboard-server .

dashboard/backend/Dockerfile

@@ -1,5 +1,7 @@
 # Stage 1: Build frontend with Node.js
 FROM node:20-alpine AS frontend-builder
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 WORKDIR /app/frontend
 COPY dashboard/frontend/package.json dashboard/frontend/package-lock.json dashboard/frontend/tsconfig.json dashboard/frontend/tsconfig.node.json dashboard/frontend/vite.config.ts ./
 COPY dashboard/frontend/src ./src
@@ -11,7 +13,8 @@ RUN npm run build
 # Stage 2: Build backend with Go
 FROM golang:1.24 AS backend-builder
 WORKDIR /app
-
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 # Use Chinese Go proxy to avoid network timeout issues
 ENV GOPROXY=https://goproxy.cn,direct
 ENV GOSUMDB=sum.golang.google.cn
@@ -31,6 +34,8 @@ RUN CGO_ENABLED=0 GOOS=${TARGETOS} GOARCH=${TARGETARCH} go build -ldflags="-w -s
 
 # Stage 3: Final runtime image
 FROM alpine:3.19
+ENV http_proxy=http://proxy.iil.intel.com:911
+ENV https_proxy=http://proxy.iil.intel.com:911
 RUN apk --no-cache add ca-certificates wget
 WORKDIR /app
 COPY --from=backend-builder /app/dashboard-backend /app/dashboard-backend

deploy/docker-compose/addons/grafana.ini

@@ -36,7 +36,8 @@ enabled = false
 
 [server]
 # Configure root URL
-root_url = %(protocol)s://%(domain)s:%(http_port)s/
+# root_url = %(protocol)s://%(domain)s:%(http_port)s/
+root_url = http://10.112.229.41:8700/embedded/grafana/
 
 # Disable serving from sub path
 serve_from_sub_path = false