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
docs(readme): point the long-document section at router.predict_long
Aashishd6a7d5f22398
diff --git a/README.md b/README.mdindex 45215ba7..97e5db49 100644--- a/README.md+++ b/README.md@@ -581,13 +581,6 @@ A smaller `window` isolates a short deciding span better (it becomes a larger frwindow); the default (`max_len - head_max_len`) favors context and throughput. Output shape matches`predict`, with `usage["windows"]` added.-The same scan is available from the Router, which routes first and then windows the checkpoint it-chose — same hints (`model=`, `task=`, `lang=`), same hooks, same `routing` key:--```python-result = router.predict_long(state, questions, model="multilingual")-```-The returned probability is the deciding window's, **not a calibrated number for the wholedocument** — a `noul` max drifts up with the window count even with no signal, and `choice` can landon a confidently-neutral window when nothing is decisive. Each answer carries `answer["window"]`@@ -599,6 +592,13 @@ r = agent.predict_long(state, questions)r["answers"]["refund"]["window"] # {'index': 13, 'token_start': 4680, 'token_end': 5432, 'count': 14}```+The same scan is reachable from the Router, which routes first and then windows the checkpoint it+picked — the same `model=`/`task=`/`lang=` hints, hooks and `routing` key as `predict`:++```python+result = router.predict_long(state, questions, model="multilingual")+```+---## GPU Fast Path (TileLang)