Skip to content

Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

C

fix(test): skip the CUDA=1 recipe assertions where CUDA=1 is rejected

JustVugg73da98f2dd47
diff --git a/c/tests/test_makefile_cuda_scope.py b/c/tests/test_makefile_cuda_scope.py
index c8325f03..1e95593a 100644
--- a/c/tests/test_makefile_cuda_scope.py
+++ b/c/tests/test_makefile_cuda_scope.py
@@ -31,7 +31,16 @@ def recipe(target, *variables):
return result.stdout + result.stderr
+def cuda_flag_is_accepted():
+ """CUDA=1 is a hard error outside Linux ("supported only on Linux"), so on
+ macOS and Windows there is no recipe to inspect at all — the assertions
+ below would fail on an absence rather than a regression."""
+ return "only on Linux" not in recipe("colibri", "CUDA=1")
+
+
@unittest.skipUnless(shutil.which("make"), "make is not installed")
[email protected](shutil.which("make") and cuda_flag_is_accepted(),
+ "this toolchain rejects CUDA=1 before any recipe is emitted")
class MakefileCudaScopeTest(unittest.TestCase):
def test_kimi_k3_is_not_built_with_cuda_flags(self):
out = recipe("kimi_k3", "CUDA=1")