The JSON file for PRECISE 2.0 is over 100 MB, which makes GitHub reject it by default. I've added some handling for load_json_model and save_to_json to work with ZIP archives. Specifically, save_to_json gets a compress option, and load_json_model automatically detects if the provided file to load is a .zip file.
When creating/loading the ZIP file, I'm doing it in a fairly dumb way involving the intermediate step of having the bare JSON file existing too; for now, I just delete that after the fact. A bit hacky, but we can iterate and figure out how to directly write JSON to a ZIP archive later perhaps?
The JSON file for PRECISE 2.0 is over 100 MB, which makes GitHub reject it by default. I've added some handling for
load_json_model
andsave_to_json
to work with ZIP archives. Specifically,save_to_json
gets acompress
option, andload_json_model
automatically detects if the provided file to load is a.zip
file.When creating/loading the ZIP file, I'm doing it in a fairly dumb way involving the intermediate step of having the bare JSON file existing too; for now, I just delete that after the fact. A bit hacky, but we can iterate and figure out how to directly write JSON to a ZIP archive later perhaps?