Update app.py

app.py

@@ -64,8 +64,13 @@ def chat_llama3_8b(message: str, history: list, temperature: float, max_new_toke
     decoded = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
     prompt_end_index = decoded[0].find(message) + len(message)
     final_response = decoded[0][prompt_end_index:] if prompt_end_index != -1 else decoded[0]
+    
+    final_response = final_response.strip("assistant")
+    if final_response.startswith(':'):
+        final_response = final_response.lstrip(':').strip()
 
-    return final_response.strip("assistant")
+    return final_response
+    
 
 
 # Gradio block