Update app.py

app.py

@@ -5,9 +5,18 @@ from collections import defaultdict
 # Initialize the model client
 client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf")
 
-# Store user preferences & history
-user_preferences = defaultdict(int)  # Tracks keywords & topics
-session_histories = defaultdict(list)  # Stores conversation history per session
+# Store user preferences & chat history
+user_preferences = defaultdict(int)  # Tracks user interests
+session_histories = defaultdict(list)  # Stores chat history per session
+
+def format_chat_history(history, system_message):
+    """Formats history into a single string in Alpaca/LLaMA style."""
+    chat_str = f"{system_message}\n\n"  # Start with system message
+
+    for user_msg, bot_response in history:
+        chat_str += f"### Instruction:\n{user_msg}\n\n### Response:\n{bot_response}\n\n"
+
+    return chat_str  # Return formatted conversation history
 
 def extract_keywords(text):
     """Extracts simple keywords from user input."""
@@ -18,46 +27,41 @@ def extract_keywords(text):
 def respond(message, history, system_message, max_tokens, temperature, top_p):
     session_id = id(history)  # Unique ID for each session
     session_history = session_histories[session_id]  # Retrieve session history
-    
+
     # Extract keywords & update preferences
     keywords = extract_keywords(message)
     for kw in keywords:
         user_preferences[kw] += 1
 
-    # Add past conversation to message history
-    messages = [{"role": "system", "content": system_message}]
-    for user_msg, bot_response in session_history:
-        messages.append({"role": "user", "content": user_msg})
-        messages.append({"role": "assistant", "content": bot_response})
-
-    # Append current user message
-    messages.append({"role": "user", "content": message})
+    # Format full conversation as a single string
+    formatted_input = format_chat_history(session_history, system_message) + f"### Instruction:\n{message}\n\n### Response:\n"
 
-    # Generate response (non-streaming fix)
-    response_data = client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
+    # Send request (fix: ensure input is a single string)
+    response = client.text_generation(
+        formatted_input,
+        max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
     )
 
-    response = response_data["choices"][0]["message"]["content"]
+    # ✅ Extract only the response
+    cleaned_response = response.split("### Response:")[-1].strip()
 
     # Save to session history
-    session_history.append((message, response))
+    session_history.append((message, cleaned_response))
 
-    # Optionally, adapt responses based on learned preferences
+    # Adapt response based on learning
     most_asked = max(user_preferences, key=user_preferences.get, default=None)
     if most_asked and most_asked in message.lower():
-        response += f"\n\nI see you're interested in {most_asked} a lot! Want to explore more details?"
-    
-    return response  # Return final response instead of streaming
+        cleaned_response += f"\n\nNaona unapenda mada ya '{most_asked}' sana! Unataka kujua zaidi?"
+
+    return cleaned_response  # ✅ Fixed: Returns only the final response
 
 # Create Chat Interface
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
-        gr.Textbox(value="You are a friendly chatbot that learns user interests.", label="System message"),
+        gr.Textbox(value="Wewe ni msaidizi wa kirafiki anayejifunza upendeleo wa mtumiaji.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
         gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),