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
ci(cpp): tokenizer goldens and checkpoints in the lane, raise MIN_TESTS to 239
Adeshina Lasisi7a843100346a
diff --git a/.github/workflows/cpp.yml b/.github/workflows/cpp.ymlindex 27f0f022..e9aac0cc 100644--- a/.github/workflows/cpp.yml+++ b/.github/workflows/cpp.yml@@ -45,6 +45,7 @@ env:MIN_TESTS: "213"GOLDENS: ${{ github.workspace }}/.cpp-goldensLAYA_GOLDENS: ${{ github.workspace }}/.cpp-goldens+ LAYA_CHECKPOINTS: ${{ github.workspace }}/laya-java/.work/checkpoints # the tokenizer tests need both shipped checkpointsLAYA_REQUIRE_GOLDENS: "1" # a missing golden is a failure, never a skipjobs:@@ -92,7 +93,7 @@ jobs:LAYA_FIXTURE_CHECKPOINTS: ${{ github.workspace }}/laya-java/.work/checkpointsPYTHONPATH: ${{ github.workspace }}run: |- python laya-java/scripts/gen_fixtures.py --strict --out "$GOLDENS" --only lang_tables.json --only lang_detect.json --only python_json.json+ python laya-java/scripts/gen_fixtures.py --strict --out "$GOLDENS" --only lang_tables.json --only lang_detect.json --only python_json.json --only tokenizer_ids.json --only decode_text.jsonpython laya-go/scripts/gen_go_fixtures.py --out "$GOLDENS"- name: Compiled C++ tables match CPythonenv: