Learning A Unified Named Entity Tagger From Multiple Partially Annotated Corpora For Efficient Adaptation
https://arxiv.org/abs/1909.11535
The paper was accepted as a long paper in CoNLL 2019.
I'm using:
-Python 3.6
-Pytorch 0.3.1
-tqdm
Please download word embedding into /embeddings/
from http://bio.nlplab.org/ (wikipedia-pubmed-and-PMC-w2v.bin)
To train the proposed models, run commands in
submit_train.sh
To evaluate the models, run commands in
submit_eval.sh