Best model for summarization

Hey everybody!

I’d like to set up a text summarization pipeline in my local environment, to run summarization on .docs, .pdfs and text files.

Could someone please recommend an Open Source pre trained model. Would prefer to run on my laptop or consumer gaming rig, or I can run it inside a VPC in AWS but I need it to not leak any PII anywhere I can’t control

I got suggestions to use Flan and Bert. Any thoughts would be super welcome. Thank you wise community!