bert-base-gl-cased / README.md
marcosgg's picture
Update README.md
89fde9b
metadata
language:
  - gl
  - pt
widget:
  - text: A mesa estaba feita de [MASK].
license: agpl-3.0

BERT for Galician (Base)

This is a base pre-trained BERT model (12 layers, cased) for Galician (ILG/RAG spelling). It was evaluated on lexical semantics tasks, using a dataset to identify homonymy and synonymy in context, and presented at ACL 2021.

There is also a small version (6 layers, cased): marcosgg/bert-small-gl-cased

Citation

If you use this model, please cite the following paper:

@inproceedings{garcia-2021-exploring,
    title = "Exploring the Representation of Word Meanings in Context: {A} Case Study on Homonymy and Synonymy",
    author = "Garcia, Marcos",
    booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)",
    year = "2021",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.acl-long.281",
    doi = "10.18653/v1/2021.acl-long.281",
    pages = "3625--3640"
}