Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
Python calibrationclassificationdecision-modelhuggingfacejevmodernbertmultilingualnlppythonpytorchroutingtyped-decisionszero-shot
fix(lang): leave ki out of the Azerbaijani stopwords
Nandakishorab87e4792eab
diff --git a/laya/lang.py b/laya/lang.pyindex 93388752..a0e84637 100644--- a/laya/lang.py+++ b/laya/lang.py@@ -138,8 +138,9 @@ _STOP = {"shomossa", "somossa", "dhonnobad", "bhai", "shob", "keu", "kichu", "bolte", "bolun","parben", "asbe", "jabe", "pabo", "ferot", "dorkar", "hoye", "geche", "gese"},- # "her", "ne", "men", "de" collide with English, French and Romanian, so they are left out.- "az": {"və", "ve", "bir", "bu", "ki", "üçün", "ucun", "ilə", "ile", "olan", "olub", "olmasa",+ # "her", "ne", "men", "de" collide with English, French and Romanian, and "ki" with the+ # romanized Bangla list, so they are left out.+ "az": {"və", "ve", "bir", "bu", "üçün", "ucun", "ilə", "ile", "olan", "olub", "olmasa","var", "yox", "yoxdur", "mən", "sən", "biz", "siz", "onlar", "daha", "çox", "cox","hər", "nə", "kimi", "görə", "sonra", "əgər", "eger", "deyil", "lakin", "amma","ancaq", "artıq", "artiq", "də", "isə", "həm", "yalnız", "yalniz"},