Update app.py

app.py

@@ -109,7 +109,13 @@ def chat_with_model(user_input, chat_history=[]):
     )
     
     outputs = model.generate(input_ids, max_new_tokens=1200, do_sample=True, top_k=50, temperature=0.7)
-    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+   # response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    # Decode and clean the response to remove unwanted repetitions of "assistant"
+    response = self.tokenizer.decode(outputs[0][input_ids.shape[-1]:], skip_special_tokens=True).strip()
+    
+    # Remove potential repeated assistant text
+    response = response.replace("<|assistant|>", "").strip()
     
     # End model generation span
     generation_span.end(