Skip to Main Content
Table 5: 
Test set performance on GLUE tasks. MRPC: F1/accuracy, STS-B: Pearson/Spearmanr correlation, QQP: F1/accuracy, MNLI: matched/mistached accuracies, and accuracy for all the other tasks. WNLI (not shown) is always set to majority class (65.1% accuracy) and included in the average.
CoLASST-2MRPCSTS-BQQPMNLIQNLIRTE(Avg)
Google BERT 59.3 95.2 88.5/84.3 86.4/88.0 71.2/89.0 86.1/85.7 93.0 71.1 80.4 
Our BERT 58.6 93.9 90.1/86.6 88.4/89.1 71.8/89.3 87.2/86.6 93.0 74.7 81.1 
Our BERT-1seq 63.5 94.8 91.2/87.8 89.0/88.4 72.1/89.5 88.0/87.4 93.0 72.1 81.7 
SpanBERT 64.3 94.8 90.9/87.9 89.9/89.1 71.9/89.5 88.1/87.7 94.3 79.0 82.8 
CoLASST-2MRPCSTS-BQQPMNLIQNLIRTE(Avg)
Google BERT 59.3 95.2 88.5/84.3 86.4/88.0 71.2/89.0 86.1/85.7 93.0 71.1 80.4 
Our BERT 58.6 93.9 90.1/86.6 88.4/89.1 71.8/89.3 87.2/86.6 93.0 74.7 81.1 
Our BERT-1seq 63.5 94.8 91.2/87.8 89.0/88.4 72.1/89.5 88.0/87.4 93.0 72.1 81.7 
SpanBERT 64.3 94.8 90.9/87.9 89.9/89.1 71.9/89.5 88.1/87.7 94.3 79.0 82.8 
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