Open cparjaszewski opened 5 years ago
~e.x is probably not supported in smatch right now. Is there any recent AMR guideline referring to this? If so, can you send the link?
This is the output of the ISI aligner. I take it that smatch expects the plain AMR, without automatic alignments to words of the sentence. (The AMR data releases include both kinds.)
Support for ISI alignments would be very welcome! as e.g. Penman does
https://penman.readthedocs.io/en/latest/api/penman.surface.html
Hello,
I've the following AMR format, but it raises a warning and error:
These are the errors:
Is the AMR in the correct format? Is something missing within Smatch?