File size: 117 Bytes
9f570f1
1
2
hello! this is the pretrained BART. The dataset used for pretraining is nonsense summary corpus with output as diff.