UKPLab / coling2018-graph-neural-networks-question-answering

Accompanying code for our COLING 2018 paper "Modeling Semantics with Gated Graph Neural Networks for Knowledge Base Question Answering"
Apache License 2.0
173 stars 29 forks source link

Using the pre-trained model get error #6

Closed Mystery-000 closed 4 years ago

Mystery-000 commented 5 years ago

Hi,when I Run python -m questionanswering.evaluate_on_test [model_file_path] configs/webqsp_eval_config.yaml [random_seed] [GPU_id] command,get the exception error:Exception: Check whether you have started the CoreNLP server e.g. $ cd stanford-corenlp-full-2015-12-09/ $ java -mx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer However,I have had started the nlp server,eg:[main] INFO CoreNLP - Starting server... [main] INFO CoreNLP - StanfordCoreNLPServer listening at /0:0:0:0:0:0:0:0:9000 Could help me solve this problem? Thank you for your attention~

daniilsorokin commented 5 years ago

Hi,

yes, you need a running instance of the Stanford CoreNLP server (https://stanfordnlp.github.io/CoreNLP/corenlp-server.html) for tokenisation and NE recognition.

Best, Daniil

SRL94 commented 5 years ago

Hi Daniil,

I have the same issue. I have started the nlp server and run the code again, but I still got the same issue: Exception: Check whether you have started the CoreNLP server.

Best, SR

SRL94 commented 5 years ago

Also, can I know what version of CoreNLP did you use in your project? Thank you for your help.

SRL94 commented 5 years ago

Hi Daniil,

I have the same issue. I have started the nlp server and run the code again, but I still got the same issue: Exception: Check whether you have started the CoreNLP server.

Best, SR

I replace corenlp = StanfordCoreNLP('http://semanticparsing:9000') bycorenlp = StanfordCoreNLP('http://localhost:9000'). The CoreNLP connection issue is solved, but I get a new exception:

properties={corenlp_properties, corenlp_caseless} if caseless else corenlp_properties AttributeError: 'str' object has no attribute 'get'

I guess it is because of CoreNLP version. I am using

stanford-corenlp-full-2018-10-05

hsaghir commented 4 years ago

@SRL94 you need to also download CoreNLP models for your specific language from here: https://stanfordnlp.github.io/CoreNLP/#human-languages-supported

daniilsorokin commented 4 years ago

Yes, exactly, do not forget the CoreNLP models as well. Thanks, @hsaghir!