Update app.py

app.py

@@ -155,7 +155,8 @@ async def stream_text(model, tokenizer, input_text, generation_config, stop_sequ
             num_return_sequences=generation_config.num_return_sequences,
             stopping_criteria=stopping_criteria,
             output_scores=True,
-            return_dict_in_generate=True
+            return_dict_in_generate=True,
+            pad_token_id=tokenizer.pad_token_id
         )
         new_token_ids = outputs.sequences[0][encoded_input.input_ids.shape[-1]:]