Update fine_tune_inference_test.py

fine_tune_inference_test.py

@@ -97,8 +97,11 @@ def chat(msg: Message):
                 max_new_tokens=200,
                 do_sample=True,
                 temperature=0.7,
+                top_k=50,
+                top_p=0.95,
                 return_dict_in_generate=True,
-                output_scores=True
+                output_scores=True,
+                suppress_tokens=[tokenizer.pad_token_id] if tokenizer.pad_token_id is not None else None
             )
 
         generated_ids = output.sequences[0]