get postags and lemmas of sentences all zero when running prediction

[zjxs1997]

I run python -m sesame.targetid --mode predict --model_name fn1.7-pretrained-targetid --raw_input sentences.txt and find the ouput in logs/fn1.7-pretrained-targetid/predicted-targets.conll is total blank. It turned out that instance.postags and instance.lemmas are all lists of zero on line 435, sesame/targetid.py, which is a result of line 28&30, sesame/conll09.py. I wonder how can I solve this problem.