Grammar models require a lot of memory. This is a problem for each individual celery worker. We have some options like hosting an external service.
|Open||None||T152710 Politeness as signal/model|
|Open||None||T144636 [Epic] Implement PCFG features for editquality and draftquality|
|Open||None||T120170 [Epic] Paid editing (COI) detection model|
|Open||None||T157041 Address memory usage issues for deploying PCFG-based features|
|Open||None||T155111 [Spike] Investigate use of Apertium LTtoolbox API in labs/production|