Decoupled_NER

We will update the code, data and readme soon. The temporary code can be seen in https://github.com/huyun-cs/Decoupled_NER/blob/main/1240_file_SupMat__Software.zip

If you have read the code and need data, please send an email to huyuniot@163.com .

Train:

nohup python -u run_3_loss_share.py --model bert_3_loss_share >ner_train_dropout.log 2>&1 &

Test:

nohup python -u eval_3_loss.py --model bert_3_loss_share >ner_test_dropout.log 2>&1 &

dataset:

OntoNotes： https://catalog.ldc.upenn.edu/LDC2011T03

MSRA: http://download.fastnlp.top/dataset/MSRA_NER.zip

Weibo : https://github.com/hltcoe/golden-horse

Dictionary:

from: https://shurufa.baidu.com/dict

from: https://pinyin.sogou.com/dict/

Distantly supervised data:

from: http://paper.people.com.cn

from: https://weibo.com/

Detail

We will show the detail of the process.

Annotation data

Pretained data

Pretrain process

put data in the ./deNER