Improve documentation clarity for OLLAMA_HOST and Caddy rate limiting

Co-authored-by: hlohaus <983577+hlohaus@users.noreply.github.com>

Caddyfile.example

@@ -42,14 +42,11 @@ your-domain.com {
     }
 
     # Optional: Rate limiting
-    # Uncomment to enable
-    # rate_limit {
-    #     zone dynamic {
-    #         key {remote_host}
-    #         events 100
-    #         window 1m
-    #     }
-    # }
+    # Note: Rate limiting in Caddy requires a plugin or external middleware
+    # For built-in rate limiting, consider using:
+    # 1. Caddy rate-limit plugin: https://github.com/mholt/caddy-ratelimit
+    # 2. External rate limiting service
+    # 3. Nginx as described in docs/reverse-proxy.md
 }
 
 # For local development without a domain (comment out the above and use this instead)

docker-compose-proxy.yml

@@ -20,6 +20,8 @@ services:
     networks:
       - gpt4free-network
     environment:
+      # Optional: Only needed if using Ollama for local model inference
+      # Remove this line if not using Ollama
       - OLLAMA_HOST=host.docker.internal
 
   # Caddy reverse proxy with automatic HTTPS