[misc] Use --cuda-graph-max-bs-decode in tests, examples, and docs (#29591)

2026-06-28 18:38:28 -07:00
parent 3217410cf6
commit 909123ddb8
140 changed files with 304 additions and 290 deletions
@@ -21,7 +21,13 @@ class TestDeepseekTP2(CustomTestCase):
        other_args = ["--trust-remote-code"]
        if torch.cuda.is_available() and torch.version.cuda:
            other_args.extend(
-                ["--tp", "2", "--enable-torch-compile", "--cuda-graph-max-bs", "2"]
+                [
+                    "--tp",
+                    "2",
+                    "--enable-torch-compile",
+                    "--cuda-graph-max-bs-decode",
+                    "2",
+                ]
            )
        cls.process = popen_launch_server(
            cls.model,