PEGASUS 2B + SLiC

Reported on 3 benchmarks across 1 task · 1 paper · 2 SOTA

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

Knowledge Base3 results

Text SummarizationonReddit TIFU
ROUGE-1· 2022-09-30
32.03
SOTA
Calibrating Sequence likelihood Improves Conditional Language Generation arXiv:2210.00045
Text SummarizationonReddit TIFU
ROUGE-L· 2022-09-30
25.51
SOTA
Calibrating Sequence likelihood Improves Conditional Language Generation arXiv:2210.00045
Text SummarizationonReddit TIFU
ROUGE-2· 2022-09-30
11.13
best: 11.25 (MUPPET BART Large)
Calibrating Sequence likelihood Improves Conditional Language Generation arXiv:2210.00045