Skip to Main Content
Table 2: 

Impact of training methods on BERT-large models fine-tuned on CoNLL or OntoNotes.

MethodCoNLLOntoNotes
TestnrbwtsTestnrbwts
BERT-lrg 92.8 75.6 98.6 89.9 75.4 95.1 
 
 +mask 92.9 82.9 98.4 89.8 77.3 96.5 
 +freeze 92.7 83.1 98.4 89.9 79.8 96.0 
 +adv 92.7 86.1 98.3 90.1 85.8 95.2 
 
 +f&m 92.8 85.5 97.8 89.9 80.6 95.9 
 +a&m 92.8 87.7 98.1 89.7 87.6 95.9 
 +a&f 92.7 88.4 98.2 90.0 88.1 95.7 
 
 +a&m&f 92.8 89.7 97.9 89.9 88.8 95.6 
MethodCoNLLOntoNotes
TestnrbwtsTestnrbwts
BERT-lrg 92.8 75.6 98.6 89.9 75.4 95.1 
 
 +mask 92.9 82.9 98.4 89.8 77.3 96.5 
 +freeze 92.7 83.1 98.4 89.9 79.8 96.0 
 +adv 92.7 86.1 98.3 90.1 85.8 95.2 
 
 +f&m 92.8 85.5 97.8 89.9 80.6 95.9 
 +a&m 92.8 87.7 98.1 89.7 87.6 95.9 
 +a&f 92.7 88.4 98.2 90.0 88.1 95.7 
 
 +a&m&f 92.8 89.7 97.9 89.9 88.8 95.6 
View Large
Close Modal
This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal
This site uses cookies. By continuing to use our website, you are agreeing to our privacy policy.
Accept