transition-amr-parser
transition-amr-parser copied to clipboard
Where does entity_rules.json get generated?
I trained a model according to the instructions, and it seems like there was a file entity_rules.json that was supposed to be generated. It looks like these rules are necessary for post-processing, but I am really confused about where in the preprocessing/training code the file gets created.
this should not be needed for the recent parsers, which version are you using?
closed due to lack of activity