Which of the sshleifer/* models can be used as-is for text summarization?

Hi @sshleifer,

I would like to ask which of the following models can be used for text summarization:

• shleifer/distilbart-cnn-12-6
• sshleifer/distilbart-xsum-12-6
• sshleifer/distilbart-cnn-6-6
• sshleifer/distilbart-xsum-12-3
• sshleifer/distilbart-xsum-9-6
• sshleifer/distilbart-xsum-12-1
• sshleifer/distilbart-xsum-6-6
• sshleifer/distilbart-xsum-1-1
• sshleifer/distilbart-cnn-12-3

I would also like to know about:

• sshleifer/mbart-large-cc25

Thank you!


You can use any of these models as-is, you can use cnn checkpoints if you want longer summaries or xsum if you want short summaries,

you can also use facebook/bart-large-cnn and facebook/bart-large-xsum

Thank you very much valhalla!

:slight_smile: A great Valhalla moment: https://www.youtube.com/watch?v=qQekqWha7fg

1 Like

Hi @valhalla,

Even sshleifer/mbart-large-cc25, right? I can use it also for text summarization?


sshleifer/mbart-large-cc25 is not fine-tuned on summarization, you’ll need to fine-tune it.

Thank you!