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(docs): render Hindi and Japanese examples as text, not \uXXXX escapes
Nandakishor0ff667c1f789
diff --git a/README.md b/README.mdindex 7e7f0bf5..6ce8ac0c 100644--- a/README.md+++ b/README.md@@ -156,7 +156,7 @@ router = Router() # nothing is downloaded until a request needs itrouter.predict({"body": "I was charged twice, please refund."}, questions)# Hindi -> routed to the multilingual checkpoint automatically-router.predict({"body": "\u092e\u0941\u091d\u0938\u0947 \u0926\u094b \u092c\u093e\u0930 \u0936\u0941\u0932\u094d\u0915 \u0932\u093f\u092f\u093e \u0917\u092f\u093e"}, questions)+router.predict({"body": "मुझसे दो बार शुल्क लिया गया"}, questions)# explicit when you already knowrouter.predict(state, questions, model="typed-decisions")@@ -166,7 +166,7 @@ router.predict(state, questions, lang="de")Every result carries the decision that produced it:```python-result = router.predict({"body": "\u4e8c\u91cd\u306b\u8acb\u6c42\u3055\u308c\u307e\u3057\u305f"}, questions)+result = router.predict({"body": "二重に請求されました"}, questions)result["routing"]# {'model': 'multilingual',# 'repo': 'convaiinnovations/laya-multilingual',