Skip to content

armgit / laya

Public Imported from GitHub
0

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.

ci(evals): run the offline research suites no workflow executes

thread3db5c260bfbc88
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index c97f733d..52d423cc 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -268,6 +268,19 @@ jobs:
- name: Eval ONNX runner tests
run: python tests/test_evals_onnx.py
+ # The offline research suites. Each claims a CI consumer and had none: `test_laya_eval.py`
+ # says in its own docstring that it "runs in CI next to the other suites", while
+ # `test_presentation_checks.py` and `test_act_head_eval.py` guard the eval harness's own
+ # presentation and act-head math. They need only `laya` and numpy, so they belong here
+ # rather than in the weekly weight lane. `test_metamorphic` is invoked as a module: run as a
+ # script its `research` import fails (`No module named 'research'`).
+ - name: Offline research eval suites
+ run: |
+ python research/eval/test_laya_eval.py
+ python research/eval/test_presentation_checks.py
+ python -m research.eval.test_metamorphic
+ python research/evals/test_act_head_eval.py
+
- name: Fixture dataset validates
run: laya-evals validate research/evals/fixture.jsonl