Closed seanswyi closed 2 years ago
I checked this GitHub Issue on the AMR parsing GitHub that you used and apparently they haven't made their pretrained model public.
Did you separately train your own AMR model using the AMR data?
Yes. We used IBM's code and retrained the AMR parser using AMR 3.0 data (LDC2019E81).
I checked this GitHub Issue on the AMR parsing GitHub that you used and apparently they haven't made their pretrained model public.
Did you separately train your own AMR model using the AMR data?