As someone whose extremely interested in this domain. I am excited to see these models productized.

Are the pretrained extractive models able to create extractive summaries at the word level? Or are they (like most other extractive models) only selecting and ranking sentences?

It is at the token level. The length of the results can be changed.

It looks like according to the notebook recepies that it's actually working at the sentence level (despite your claim). The way you train these is to mark a sentence as "1" or "0" based on if it's selected.

Oh well, the world waits longer for a word-level (willing to skip around in sentences) and grammatically correct extractive summarizer. You got my hopes up though.

https://github.com/microsoft/nlp-recipes/blob/master/example...

My (failed) attempt to build what I was actually looking for:

https://github.com/Hellisotherpeople/CX_DB8