add support for json based content types in chat completion request

arch/envoy.template.yaml

@@ -625,7 +625,7 @@ static_resources:
               - endpoint:
                   address:
                     socket_address:
-                      address: 0.0.0.0
+                      address: host.docker.internal
                       port_value: 9091
                   hostname: localhost
 

arch/tools/cli/docker_cli.py

@@ -52,6 +52,8 @@ def docker_start_archgw_detached(
     port_mappings = [
         f"{prompt_gateway_port}:{prompt_gateway_port}",
         f"{llm_gateway_port}:{llm_gateway_port}",
+        # expose llm gateway as well without tracing support
+        f"{llm_gateway_port + 1}:{llm_gateway_port + 1}",
         "9901:19901",
     ]
     port_mappings_args = [item for port in port_mappings for item in ("-p", port)]