Additional information regarding the parse_xfm_bart_large model

bjascob / amrlib

A python library that makes AMR parsing, generation and visualization simple.

MIT License

216 stars 33 forks source link

Additional information regarding the parse_xfm_bart_large model #62

Closed helena-balabin closed 5 months ago

helena-balabin commented 6 months ago

If I understand correctly, this is a BART-large based model that was trained on AMR-3 (LDC2020T02) to get graphs from sentences. If so, what does the XFM part stand for? I couldn't find any info in the READMEs and code.

bjascob commented 6 months ago

XFM = transformer. Be sure to look at the read-the-docs pages for how to use it.