feat(wren-ai-service): multi-llm deployment and setting loading mechanism

.github/workflows/ai-service-test.yaml

@@ -46,7 +46,7 @@ jobs:
       - name: Install Just
         uses: extractions/setup-just@v2
         with:
-          just-version: "1.31.0"
+          just-version: "1.36.0"
       - name: Prepare testing environment and Run tests
         run: |
           just test

.gitignore

@@ -26,7 +26,8 @@ wren-ai-service/demo/poetry.lock
 wren-ai-service/demo/custom_dataset
 wren-ai-service/demo/.env
 wren-ai-service/tools/dev/etc/**
-wren-ai-service/.deepeval-cache.json
+.deepeval-cache.json
+docker/config.yaml
 
 # python
 .python-version

deployment/kustomizations/base/cm.yaml

@@ -3,12 +3,10 @@ kind: ConfigMap
 metadata:
   name: wren-config
 data:
-
   # Wren Engine Service Port
   WREN_ENGINE_PORT: "8080"
   # Wren AI Service Port
   WREN_AI_SERVICE_PORT: "5555"
-  WREN_AI_SERVICE_ENABLE_TIMER: "1"
 
   WREN_UI_ENDPOINT: http://wren-ui-svc:3000
 
@@ -18,25 +16,14 @@ data:
   WREN_AI_SERVICE_VERSION: "0.10.4"
   WREN_UI_VERSION: "0.16.0"
 
-  # LLM and Embedder Configurations
-  LLM_PROVIDER: "openai_llm"
-  LLM_OPENAI_API_BASE: "https://api.openai.com/v1"
-  GENERATION_MODEL: "gpt-4o-mini"
-
-  EMBEDDER_PROVIDER: "openai_embedder"
-  EMBEDDER_OPENAI_API_BASE: "https://api.openai.com/v1"
-
   # Document store related
   QDRANT_HOST: "wren-qdrant"
-  DOCUMENT_STORE_PROVIDER: "qdrant"
-
-  # Langfuse: for LLM tracing
-  LANGFUSE_HOST: "https://cloud.langfuse.com"
-  LANGFUSE_ENABLE: ""
 
   # Telemetry
   POSTHOG_HOST: "https://app.posthog.com"
   TELEMETRY_ENABLED: "false"
+  # this is for telemetry to know the model, i think ai-service might be able to provide a endpoint to get the information
+  GENERATION_MODEL: "gpt-4o-mini"
 
   # service endpoints of AI service & engine service
   WREN_ENGINE_ENDPOINT: "http://wren-engine-svc:8080"
@@ -55,3 +42,118 @@ data:
   LOGGING_LEVEL: INFO
 
   IBIS_SERVER_ENDPOINT: http://wren-ibis-server-svc:8000
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: wren-ai-service-config
+data:
+  config.yaml: |
+    type: llm
+    provider: openai_llm
+    models:
+      - model: gpt-4o-mini
+        kwargs:
+          {
+            "temperature": 0,
+            "n": 1,
+            "max_tokens": 4096,
+            "response_format": { "type": "json_object" },
+          }
+      - model: gpt-4o
+        kwargs:
+          {
+            "temperature": 0,
+            "n": 1,
+            "max_tokens": 4096,
+            "response_format": { "type": "json_object" },
+          }
+    api_base: https://api.openai.com/v1
+    timeout: 120
+
+    ---
+    type: embedder
+    provider: openai_embedder
+    models:
+      - model: text-embedding-3-large
+        dimension: 3072
+    api_base: https://api.openai.com/v1
+    timeout: 120
+
+    ---
+    type: engine
+    provider: wren_ui
+    endpoint: http://wren-ui-svc:3000
+
+    ---
+    type: document_store
+    provider: qdrant
+    location: http://wren-qdrant:6333
+    embedding_model_dim: 3072
+    timeout: 120
+
+    ---
+    type: pipeline
+    pipes:
+      - name: indexing
+        embedder: openai_embedder.text-embedding-3-large
+        document_store: qdrant
+      - name: retrieval
+        llm: openai_llm.gpt-4o-mini
+        embedder: openai_embedder.text-embedding-3-large
+        document_store: qdrant
+      - name: historical_question
+        embedder: openai_embedder.text-embedding-3-large
+        document_store: qdrant
+      - name: sql_generation
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: sql_correction
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: followup_sql_generation
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: sql_summary
+        llm: openai_llm.gpt-4o-mini
+      - name: sql_answer
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: sql_breakdown
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: sql_expansion
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: sql_explanation
+        llm: openai_llm.gpt-4o-mini
+      - name: sql_regeneration
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: semantics_description
+        llm: openai_llm.gpt-4o-mini
+      - name: relationship_recommendation
+        llm: openai_llm.gpt-4o-mini
+        engine: wren_ui
+      - name: question_recommendation
+        llm: openai_llm.gpt-4o-mini
+      - name: intent_classification
+        llm: openai_llm.gpt-4o-mini
+        embedder: openai_embedder.text-embedding-3-large
+        document_store: qdrant
+      - name: data_assistance
+        llm: openai_llm.gpt-4o-mini
+
+    ---
+    settings:
+      column_indexing_batch_size: 50
+      table_retrieval_size: 10
+      table_column_retrieval_size: 100
+      allow_using_db_schemas_without_pruning: false
+      query_cache_maxsize: 1000
+      query_cache_ttl: 3600
+      langfuse_host: https://cloud.langfuse.com
+      langfuse_enable: true
+      enable_timer: false
+      logging_level: DEBUG
+      development: false

deployment/kustomizations/base/deploy-wren-ai-service.yaml

@@ -14,99 +14,56 @@ spec:
     spec:
       containers:
         - name: wren-ai-service
-          image: ghcr.io/canner/wren-ai-service:0.3.7
+          image: ghcr.io/canner/wren-ai-service:latest
+          volumeMounts:
+            - name: config-volume
+              mountPath: /app/data
           env:
-          - name: WREN_AI_SERVICE_PORT
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: WREN_AI_SERVICE_PORT
-          - name: LLM_PROVIDER
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: LLM_PROVIDER
-          - name: LLM_OPENAI_API_KEY
-            valueFrom:
-              secretKeyRef:
-                name: wrenai-secrets
-                key: LLM_OPENAI_API_KEY
-          - name: LLM_OPENAI_API_BASE
-            valueFrom: 
-              configMapKeyRef:
-                name: wren-config
-                key: LLM_OPENAI_API_BASE
-          - name: EMBEDDER_PROVIDER
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: EMBEDDER_PROVIDER
-          - name: EMBEDDER_OPENAI_API_KEY
-            valueFrom:
-              secretKeyRef:
-                name: wrenai-secrets
-                key: EMBEDDER_OPENAI_API_KEY
-          - name: EMBEDDER_OPENAI_API_BASE
-            valueFrom: 
-              configMapKeyRef:
-                name: wren-config
-                key: EMBEDDER_OPENAI_API_BASE
-          - name: GENERATION_MODEL
-            valueFrom: 
-              configMapKeyRef:
-                name: wren-config
-                key: GENERATION_MODEL
-          - name: QDRANT_HOST
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: QDRANT_HOST
-          - name: DOCUMENT_STORE_PROVIDER
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: DOCUMENT_STORE_PROVIDER
-          - name: WREN_ENGINE_ENDPOINT
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: WREN_ENGINE_ENDPOINT
-          - name: LOGGING_LEVEL
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: LOGGING_LEVEL
-          - name: WREN_UI_ENDPOINT
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: WREN_UI_ENDPOINT
-          - name: ENABLE_TIMER
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: WREN_AI_SERVICE_ENABLE_TIMER
-          - name: PYTHONUNBUFFERED
-            value: "1"
-          - name: LANGFUSE_ENABLE
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: LANGFUSE_ENABLE
-          - name: LANGFUSE_HOST
-            valueFrom:
-              configMapKeyRef:
-                name: wren-config
-                key: LANGFUSE_HOST
-          - name: LANGFUSE_PUBLIC_KEY
-            valueFrom:
-              secretKeyRef:
-                name: wrenai-secrets
-                key: LANGFUSE_PUBLIC_KEY
-          - name: LANGFUSE_SECRET_KEY
-            valueFrom:
-              secretKeyRef:
-                name: wrenai-secrets
-                key: LANGFUSE_SECRET_KEY
+            - name: WREN_AI_SERVICE_PORT
+              valueFrom:
+                configMapKeyRef:
+                  name: wren-config
+                  key: WREN_AI_SERVICE_PORT
+            - name: OPENAI_API_KEY
+              valueFrom:
+                secretKeyRef:
+                  name: wrenai-secrets
+                  key: LLM_OPENAI_API_KEY
+            - name: QDRANT_HOST
+              valueFrom:
+                configMapKeyRef:
+                  name: wren-config
+                  key: QDRANT_HOST
+            - name: LOGGING_LEVEL
+              valueFrom:
+                configMapKeyRef:
+                  name: wren-config
+                  key: LOGGING_LEVEL
+            - name: WREN_UI_ENDPOINT
+              valueFrom:
+                configMapKeyRef:
+                  name: wren-config
+                  key: WREN_UI_ENDPOINT
+            - name: PYTHONUNBUFFERED
+              value: "1"
+            - name: LANGFUSE_PUBLIC_KEY
+              valueFrom:
+                secretKeyRef:
+                  name: wrenai-secrets
+                  key: LANGFUSE_PUBLIC_KEY
+            - name: LANGFUSE_SECRET_KEY
+              valueFrom:
+                secretKeyRef:
+                  name: wrenai-secrets
+                  key: LANGFUSE_SECRET_KEY
+            - name: CONFIG_PATH
+              value: /app/data/config.yaml
           ports:
             - containerPort: 5555
+      volumes:
+        - name: config-volume
+          configMap:
+            name: wren-ai-service-config
+            items:
+              - key: config.yaml
+                path: config.yaml

deployment/kustomizations/examples/secret-wren_example.yaml

@@ -7,7 +7,6 @@ type: Opaque
 data:
   # LLM_OPENAI_API_KEY and EMBEDDER_OPENAI_API_KEY are REQUIRED: without a valid key the wren-ai-service-deployment pod will not start
   LLM_OPENAI_API_KEY: UkVRVUlSRUQ6IHNrLXByb2otYWxsLWFjY2Vzcy1wbGFjZWhvbGRlci00LXdyZW4tYWktc2VydmljZS1kZXBsb3ltZW50
-  EMBEDDER_OPENAI_API_KEY: UkVRVUlSRUQ6IHNrLXByb2otYWxsLWFjY2Vzcy1wbGFjZWhvbGRlci00LXdyZW4tYWktc2VydmljZS1kZXBsb3ltZW50
 
   # Azure openai env
   AZURE_CHAT_BASE: bi9h

deployment/kustomizations/kustomization.yaml

@@ -43,7 +43,7 @@ images:
   - name: ghcr.io/canner/wren-ui
     newTag: 0.9.2 # WREN_UI_VERSION
   - name: ghcr.io/canner/wren-ai-service
-    newTag: 0.8.2 # WREN_AI_SERVICE_VERSION
+    newTag: commit-391ea49 # WREN_AI_SERVICE_VERSION
   - name: ghcr.io/canner/wren-engine-ibis
     newTag: 0.9.0 # IBIS_SERVER_VERSION