DeBERTalarge

Reported on 4 benchmarks across 3 tasks · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing4 results

Common Sense ReasoningonSWAG
Test· 2020-06-05
90.8
SOTA
DeBERTa: Decoding-enhanced BERT with Disentangled Attention arXiv:2006.03654
Reading ComprehensiononRACE
Accuracy· 2020-06-05
86.8
best: 91.4 (ALBERT (Ensemble))
DeBERTa: Decoding-enhanced BERT with Disentangled Attention arXiv:2006.03654
Question AnsweringonSQuAD2.0
EM· 2020-06-05
88
best: 90.939 (IE-Net (ensemble))
DeBERTa: Decoding-enhanced BERT with Disentangled Attention arXiv:2006.03654
Question AnsweringonSQuAD2.0
F1· 2020-06-05
90.7
best: 93.214 (IE-Net (ensemble))
DeBERTa: Decoding-enhanced BERT with Disentangled Attention arXiv:2006.03654