feat: update Dockerfile and docker-compose for improved networking and add new make targets for Ollama builds

ai_stack/ollama/Dockerfile

@@ -9,6 +9,8 @@ ENV OLLAMA_HOST=0.0.0.0
 RUN apt-get update && apt-get install -y \
     curl \
     ca-certificates \
+    wget \
+    gnupg \
     && rm -rf /var/lib/apt/lists/*
 
 # Install Ollama

ai_stack/ollama/docker-compose.yml

@@ -1,6 +1,8 @@
 services:
   ollama:
-    build: .
+    build: 
+      context: .
+      network: host
     image: localhost/ollama-server:latest
     ports:
       - "11434:11434" # Map host port 11434 to container port 11434 (Ollama default)
@@ -11,6 +13,7 @@ services:
     environment:
       - OLLAMA_HOST=0.0.0.0
     restart: unless-stopped
+    network_mode: slirp4netns
     # Uncomment the following lines if you have NVIDIA GPU support
     # deploy:
     #   resources:

makefile

@@ -96,6 +96,8 @@ help:
 	@echo "  make ai-launcher          Interactive AI backend launcher (recommended for new users)"
 	@echo ""
 	@echo "Ollama-specific targets:"
+	@echo "  make ollama-build         Build Ollama container with no cache (using slirp4netns)"
+	@echo "  make ollama-build-host    Build Ollama container with host networking (alternative)"
 	@echo "  make ollama-up            Start only the Ollama container"
 	@echo "  make ollama-down          Stop only the Ollama container"
 	@echo "  make ollama-logs          View Ollama container logs"
@@ -135,6 +137,14 @@ info:
 	podman compose --env-file $(ENV_FILE) -f docker-compose.yml -f docker-compose.override.yml config
 
 # Ollama-specific targets
+ollama-build:
+	cd ai_stack/ollama && DOCKER_BUILDKIT=0 CONTAINERS_NETNS=slirp4netns podman compose build --no-cache
+	@echo "Ollama container built with no cache. Use 'make ollama-up' to start it."
+
+ollama-build-host:
+	cd ai_stack/ollama && DOCKER_BUILDKIT=0 podman build --network=host --no-cache -t localhost/ollama-server:latest .
+	@echo "Ollama container built with host networking. Use 'make ollama-up' to start it."
+
 ollama-up:
 	cd ai_stack/ollama && podman compose up -d
 	@echo "Ollama container started. Use 'make pull-deepseek-model' to download and load the DeepSeek model."