Update app.py

app.py

@@ -3,19 +3,15 @@ from huggingface_hub import InferenceClient
 
 client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf")
 
-def format_alpaca_prompt(user_input, system_prompt):
-    """Formats input in Alpaca/LLaMA style"""
-    prompt = f"""{system_prompt}
-
-### Instruction:
-{user_input}
-
-### Response:
-"""
+# Store chat history
+def format_alpaca_prompt(user_input, history, system_prompt):
+    """Formats input in Alpaca/LLaMA style with history"""
+    history_text = "\n".join([f"User: {h[0]}\nAssistant: {h[1]}" for h in history])
+    prompt = f"""{system_prompt}\n{history_text}\nUser: {user_input}\nAssistant:"""
     return prompt
 
 def respond(message, history, system_message, max_tokens, temperature, top_p):
-    formatted_prompt = format_alpaca_prompt(message, system_message)
+    formatted_prompt = format_alpaca_prompt(message, history, system_message)
 
     response = client.text_generation(
         formatted_prompt,
@@ -25,8 +21,9 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
     )
 
     # ✅ Extract only the response
-    cleaned_response = response.split("### Response:")[-1].strip()
+    cleaned_response = response.strip()
     
+    history.append((message, cleaned_response))  # ✅ Store conversation history
     yield cleaned_response  # ✅ Output only the answer
 
 demo = gr.ChatInterface(