metadata
language:
- en
tags:
- summarization
- diff generation
datasets:
- nonsense corpus
metrics:
- rouge
hello! this is the pretrained BART. The dataset used for pretraining is nonsense summary corpus with output as diff.