Update app.py

app.py

@@ -3,7 +3,7 @@ import time
 import requests
 from streamlit.components.v1 import html
 import os
-
+from dotenv import load_dotenv
 
 # Import transformers and cache the help agent for performance
 @st.cache_resource
@@ -236,37 +236,28 @@ def ask_llama(conversation_history, category, is_final_guess=False):
 
 def ask_help_agent(query):
     try:
-        import ollama
-        import requests
-
-        # (1) Check if Ollama server is running
-        try:
-            requests.get("http://localhost:11434", timeout=5)
-        except:
-            return "🛑 **Ollama is not running!**\n\nPlease:\n1. [Download Ollama](https://ollama.com)\n2. Run `ollama serve` in terminal\n3. Pull a model (`ollama pull llama3`)"
-
-        # (2) Build chat history
-        messages = [{"role": "system", "content": "You are a helpful AI assistant."}]
+        API_URL = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.1"
+        headers = {"Authorization": "Bearer hf_wm5eLl09b9I9cOxR3E9n5rrRr1CRQQjn"}  # ⚠️ Replace with your token
         
-        if "help_conversation" in st.session_state:
-            for msg in st.session_state.help_conversation:
-                if msg.get("query"):
-                    messages.append({"role": "user", "content": msg["query"]})
-                if msg.get("response"):
-                    messages.append({"role": "assistant", "content": msg["response"]})
+        prompt = f"""<s>[INST] <<SYS>>
+        You are a helpful assistant. Keep answers short.
+        <</SYS>>
+        {query} [/INST]"""
         
-        messages.append({"role": "user", "content": query})
-
-        # (3) Get response
-        response = ollama.chat(
-            model="llama3",  # or "mistral" for lighter model
-            messages=messages,
-            options={"temperature": 0.7}
+        response = requests.post(
+            API_URL,
+            headers=headers,
+            json={"inputs": prompt},
+            timeout=10
         )
-        return response['message']['content']
-
+        
+        if response.status_code == 200:
+            return response.json()[0]['generated_text'].split("[/INST]")[-1].strip()
+        else:
+            return f"Error: {response.text}"
+            
     except Exception as e:
-        return f"⚠️ **Assistant Error**\n\n{str(e)}\n\nPlease ensure Ollama is installed and running."
+        return f"Assistant busy. Try later. ({str(e)})"
 
 # Main game logic with enhanced UI
 def main():