hyesunyun's picture
Update from yun.hy
9f570f1
|
raw
history blame
117 Bytes

hello! this is the pretrained BART. The dataset used for pretraining is nonsense summary corpus with output as diff.