What are the best transformers for summarization of big texts in pt-BR?

Hi !
After some years doing ml using only sckit-learn for text classification, I have now a opportunity to do something a little bit more advanced: text summarization

So, I would like to create a small proof-of-concept using (already extracted in txt files) ┬▒ 4.000 legal text divided in:

  • 2.000 initial petitions / complaints *.txt
  • 2.000 summaries of each inital petition

PS.: all text files are in brazilian portuguese

So how can I use these txt files to train a new transformer able to generate new summaries?