Pretrained Model | Model | Dataset | F1 |
---|---|---|---|
--- | SOTA | ConLL2003 | 94.6 |
bert-large-cased | Biaffine NER | ConLL2003 | 93.5 |
bert-base-cased(Our) (wo doc-context) | Deep Biaffine | ConLL2003 | 92.31 |
bert-large-cased(Our) (wo doc-context) | Deep Biaffine | ConLL2003 | 92.52 |
------------- | ------------- | ------------- | ------------- |
--- | SOTA | Ontonotes | 92.07 |
bert-large-cased (wo doc-context) | SynLSTM | Ontonotes | 90.85 |
bert-large-cased (wo doc-context) | AELGCN | Ontonotes | 91.16 |
bert-large-cased | Biaffine NER | Ontonotes | 91.3 |
bert-base-cased(Our) (wo doc-context) | Synlstm + Deep Biaffine | Ontonotes | 90.07 |
bert-large-cased(Our) (wo doc-context) | Synlstm + Deep Biaffine Synlstm | Ontonotes | 90.66 |
Conll03 bert-base-cased, plm_lr: 2e-05, other lr:0.002, lstm layer: 1, epoch: 100, batchsize: 48, Micro F1: 92.31
bert-base-cased, plm_lr: 2e-05, other lr:0.001, lstm layer: 3, epoch: 100, batchsize: 48,Micro F1: 92.30 + pos
bert-base-cased, plm_lr: 2e-05, other lr:0.003, lstm layer: 1, epoch: 100, batchsize: 48,Micro F1: 92.31 + pos(48)
bert-large-cased, plm_lr: 2e-05, other lr:0.003, lstm layer: 1, epoch: 100, batchsize: 48,Micro F1: 92.52 + pos(48)
Ontonotes bert-base-cased, plm_lr: 2e-05, other lr:0.001, synlstm, epoch: 100, batchsize: 48, Micro F1: 90.07
bert-base-cased, plm_lr: 2e-05, other lr:0.001, synlstm, epoch: 100, batchsize: 48, Micro F1: 90.07 + pos(48)+ dep(48)
bert-base-cased, plm_lr: 2e-05, other lr:0.001, synlstm, embedder dropout 0.5, epoch: 100, batchsize: 48, 90.00, 90.25, Micro F1: 90.13 + pos(48)+ dep(48)