Update app.py

app.py

@@ -1,10 +1,11 @@
 
-model_name = 'redael/model_udc'
 import os
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer
 import torch
 
+# Load your model and tokenizer from Hugging Face
+model_name = 'redael/model_udc'
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 model = AutoModelForCausalLM.from_pretrained(model_name)
 
@@ -41,17 +42,16 @@ def generate_response(message, history, system_message, max_tokens, temperature,
     )
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     
-    # Post-process the response
+    # Clean up the response
     response = response.split("Assistant:")[-1].strip()
     response_lines = response.split('\n')
     clean_response = []
     for line in response_lines:
         if "User:" not in line and "Assistant:" not in line:
             clean_response.append(line)
-    response = ' '.join(clean_response).strip()
+    response = ' '.join(clean_response)
     
-    history.append((message, response))
-    return history, history
+    return [(message, response)]
 
 # Create the Gradio chat interface
 demo = gr.ChatInterface(