Update neura_self_hosted.py

neura_self_hosted.py

@@ -23,19 +23,17 @@ WEATHER_API_KEY = os.getenv("WEATHER_API_KEY", "") # For the weather tool
 
     # This is a great starting model. It will be downloaded automatically the first time.
     # It's specifically logging.info("Loading self-hosted generative model. This may take a moment...")
-# In neura_self_hosted.py
 
-# ... (you don't need the BitsAndBytesConfig import for this option)
-
-# --- 🧠 LOAD THE LOCAL GENERATIVE AI MODEL (CPU Version) ---
 logging.info("Loading self-hosted generative model for CPU...")
 try:
     MODEL_NAME = "microsoft/Phi-3-mini-4k-instruct"
     
     tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+    
+    # Force the model to load on the CPU and remove all GPU/quantization code
     model = AutoModelForCausalLM.from_pretrained(
         MODEL_NAME,
-        device_map="cpu", # <-- Force the model onto the CPU
+        device_map="cpu",
         trust_remote_code=True,
     )
 
@@ -44,11 +42,13 @@ try:
         model=model,
         tokenizer=tokenizer,
     )
-    logging.info(f"✅ Model '{MODEL_NAME}' loaded successfully on CPU. Expect slow performance.")
+    logging.info(f"✅ Model '{MODEL_NAME}' loaded on CPU. Expect very slow performance.")
 except Exception as e:
     logging.error(f"❌ Failed to load local AI model: {e}")
     local_ai_pipeline = None
 
+# ... (The rest of your Python code remains the same) ...
+
 # --- AI Personas & System Prompt ---
 ANIME_PERSONAS = {
     "default": "You are a versatile and intelligent AI assistant.",