Skip to content

FrameNetBrasil/span-finder

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

span-finder

Parse sentences by finding & labeling spans

Installation

Environment:

  • python >= 3.7
  • python-pip

To install the dependencies, execute

pip3 install -r requirements.txt

Optionally, you may install the package via

python3 setup.py install

and import span-finder with import sftp.

Demo

A demo (combined with Patrick's coref model) is here.

Inference Only

If you use SpanFinder only for inference, we provide a checkpoint that was trained on FrameNet v1.7. This example shows the basic API of SpanFinder. Note that the script will incur a checkpoint download everytime, so the best way is to download the checkpoint to local (~1.7GiB), or better extract it out, and then point the -m argument to the archived file or extracted folder.

Training

For training, you may need to read an overall document, the doc for data, and the doc for training.

Paper

Welcome to cite our work if you found it useful:

@inproceedings{xia-etal-2021-lome,
    title = "{LOME}: Large Ontology Multilingual Extraction",
    author = "Xia, Patrick  and
      Qin, Guanghui  and
      Vashishtha, Siddharth  and
      Chen, Yunmo  and
      Chen, Tongfei  and
      May, Chandler  and
      Harman, Craig  and
      Rawlins, Kyle  and
      White, Aaron Steven  and
      Van Durme, Benjamin",
    booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations",
    year = "2021",
    url = "https://www.aclweb.org/anthology/2021.eacl-demos.19",
    pages = "149--159",
}

About

Parse sentences by finding & labeling spans

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 86.7%
  • HTML 11.1%
  • Jsonnet 2.2%