Feature model initialization (#7)

* Enhance devcontainer setup and improve RAG functionality

- Updated devcontainer configuration to include multiple docker-compose files.
- Added new docker-compose.devcontainer.yml for devcontainer overriddes
- Refined environment variables for text and image collections in .env file.
- Improved prompt generation in rag.py for better user interaction.
- Streamlined chat.sh for continuous question and image identification modes.
- Adjusted docker-compose.yml to use model_cache for caching.

* Refactor prompts for clarity and implement lazy loading for models in MemoryAlphaRAG

Author: aniongithub

.devcontainer/devcontainer.json

@@ -1,10 +1,14 @@
 {
 	"name": "Memoryalpha RAG API",
-	"dockerComposeFile": "../docker-compose.yml",
+	"dockerComposeFile": [
+		"../docker-compose.yml",
+		"docker-compose.devcontainer.yml"
+	],
 	"service": "lcars",
 	"overrideCommand": true,
 	"remoteUser": "vscode",
 	"postAttachCommand": "/workspace/memoryalpha-rag-api/wait-for-ollama.sh",
+	"postCreateCommand": "sudo chown -R vscode:users ${containerWorkspaceFolder}/.cache",
 	"features": {
 		"ghcr.io/nils-geistmann/devcontainers-features/create-remote-user:0": {
 			"passwordLessSudo": true

.devcontainer/docker-compose.devcontainer.yml

@@ -0,0 +1,10 @@
+version: "3.9"
+
+services:
+  lcars:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    user: vscode
+    volumes:
+      - ./.cache:/home/vscode/.cache # Override default cache location for devcontainer only 

.env

@@ -3,4 +3,5 @@ DEFAULT_IMAGE_MODEL="qwen2.5vl:3b"
 
 OLLAMA_URL="http://ollama:11434"
 DB_PATH="/data/enmemoryalpha_db"
-COLLECTION_NAME="memoryalpha"
+TEXT_COLLECTION_NAME="memoryalpha_text"
+IMAGE_COLLECTION_NAME="memoryalpha_images"