Lynten / stanford-corenlp

Python wrapper for Stanford CoreNLP.
MIT License
919 stars 200 forks source link

Program never ends #62

Open Sannndy0000 opened 5 years ago

Sannndy0000 commented 5 years ago

I found a quite weird situation that when I use several annotators in some certain orders, it will somehow stuck halfway and never ends. At the same time my whole computer including other processes will be extremely slow, seems to be effected. However, once I got it fixed simply by changing the order of the annotators. For example, pos_list = nlp.pos_tag(sentence) print("1") dependency = nlp.dependency_parse(sentence) print("2") ner = nlp.ner(sentence) print("3") This will stuck after printing 2. If I delete the 2nd step, then it can print 3 and successfully end.

Anyone encountered this before?