fix(inference/traeffik): Refactor traeffik to be outside of models and add sd

jan-inference/docker-compose.yml

@@ -0,0 +1,25 @@
+version: '3'
+
+services:
+  # Service for Traefik, a modern HTTP reverse proxy and load balancer.
+  traefik:
+    image: traefik:v2.10
+    ports:
+      # Map port 80 in the container to port 80 on the host.
+      - "80:80"
+      # Map port 8080 in the container (Traefik's dashboard) to port 8080 on the host.
+      - "8080:8080"
+    # Mount the Docker socket to allow Traefik to listen to Docker's API.
+    volumes:
+      - /var/run/docker.sock:/var/run/docker.sock:ro
+      - ./traefik/traefik.yml:/traefik.yml:ro
+      - ./traefik/config.yml:/config.yml:ro
+    # Connect this service to the traefik_public network.
+    networks:
+      - traefik_public
+
+# Define networks used in this docker-compose file.
+networks:
+  # Public-facing network that Traefik uses. Marked as external to indicate it may be defined outside this file.
+  traefik_public:
+    external: true

jan-inference/llm/traefik/config.yml

@@ -1,23 +0,0 @@
-http:
- #region routers 
-  routers:
-    llm:
-      entryPoints:
-        - "web"
-      rule: "PathPrefix(`/inference/llm`)"
-      middlewares:
-        - llmm-strip-prefix
-      service: llm
-#endregion
-#region services
-  services:
-    llm:
-      loadBalancer:
-        servers:
-          - url: "http://llm:8000"
-#endregion
-  middlewares:
-    llmm-strip-prefix:
-      stripPrefix:
-        prefixes: 
-          - "/inference/llm"

jan-inference/traefik/config.yml

@@ -0,0 +1,38 @@
+http:
+ #region routers 
+  routers:
+    llm:
+      entryPoints:
+        - "web"
+      rule: "PathPrefix(`/inference/llm/`)"
+      middlewares:
+        - llmm-strip-prefix
+      service: llm
+    sd:
+      entryPoints:
+        - "web"
+      rule: "PathPrefix(`/inference/sd/`)"
+      middlewares:
+        - sd-strip-prefix
+      service: sd
+#endregion
+#region services
+  services:
+    llm:
+      loadBalancer:
+        servers:
+          - url: "http://llm:8000"
+    sd:
+      loadBalancer:
+        servers:
+          - url: "http://sd:8000"
+#endregion
+  middlewares:
+    llm-strip-prefix:
+      stripPrefix:
+        prefixes: 
+          - "/inference/llm/"
+    sd-strip-prefix:
+      stripPrefix:
+        prefixes: 
+          - "/inference/sd/"