Piecewise Cuda Graph set default (#16331)

2026-03-02 23:18:07 +08:00
parent 468e3dc56b
commit c64274c746
34 changed files with 505 additions and 143 deletions
@@ -140,7 +140,7 @@ class TestVLMViTCudaGraph(CustomTestCase):
                other_args=[
                    "--mm-attention-backend",
                    "fa3",
-                    "--enable-piecewise-cuda-graph",
+                    "--enforce-piecewise-cuda-graph",
                    "--piecewise-cuda-graph-max-tokens",
                    "8192",
                    "--chunked-prefill-size",