Update app.py

app.py

@@ -97,8 +97,10 @@ async def main(message: cl.Message):
     # decode output from tokenized output to str output
     decoded_output = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)
 
+    print(decoded_output)
+
     # return only the generated response (not the prompt) as output
-    response = decoded_output[0].split("<|end_header_id|>")[-1]
+    response = decoded_output[0].split("<|end_of_text|>")[-1]
 
     msg = cl.Message(content=response)