Consists of 1.3 million records of U.S. patent documents along with human written abstractive summaries.
Source: BIGPATENT: A Large-Scale Dataset for Abstractive and Coherent Summarization
Variants: BigPatent, big_patent
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Text Summarization | LongT5 | LongT5: Efficient Text-To-Text Transformer for … | 2021-12-15 |
| Text Summarization | BigBird-Pegasus | Big Bird: Transformers for Longer … | 2020-07-28 |
Recent papers with results on this dataset: