bug fix for empty audio

funasr/bin/asr_inference_launch.py

@@ -370,7 +370,7 @@ def inference_paraformer(
             results = speech2text(**batch)
             if len(results) < 1:
                 hyp = Hypothesis(score=0.0, scores={}, states={}, yseq=[])
-                results = [[" ", ["sil"], [2], hyp, 10, 6]] * nbest
+                results = [[" ", ["sil"], [2], hyp, 10, 6, []]] * nbest
             time_end = time.time()
             forward_time = time_end - time_beg
             lfr_factor = results[0][-1]