BART + R-Drop

Reported on 6 benchmarks across 2 tasks · 1 paper

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

Knowledge Base3 results

Text SummarizationonCNN / Daily Mail
ROUGE-1· 2021-06-28
44.51
best: 48.18 (Scrambled code + broken (alter))
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448
Text SummarizationonCNN / Daily Mail
ROUGE-2· 2021-06-28
21.58
best: 24.02 (Pegasus)
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448
Text SummarizationonCNN / Daily Mail
ROUGE-L· 2021-06-28
41.24
best: 45.35 (Scrambled code + broken (alter))
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448

Natural Language Processing3 results

Abstractive Text SummarizationonCNN / Daily Mail
ROUGE-1· 2021-06-28
44.51
best: 48.18 (Scrambled code + broken (alter))
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448
Abstractive Text SummarizationonCNN / Daily Mail
ROUGE-2· 2021-06-28
21.58
best: 24.02 (Pegasus)
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448
Abstractive Text SummarizationonCNN / Daily Mail
ROUGE-L· 2021-06-28
41.24
best: 45.35 (Scrambled code + broken (alter))
R-Drop: Regularized Dropout for Neural Networks arXiv:2106.14448