[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
@@ -31,7 +31,7 @@ class TestMLADeepseekV3ChannelInt8(CustomTestCase):
        if torch.cuda.is_available() and torch.version.cuda:
            other_args.extend(
                [
-                    "--cuda-graph-max-bs",
+                    "--cuda-graph-max-bs-decode",
                    "16",
                    "--enable-torch-compile",
                    "--torch-compile-max-bs",
@@ -75,7 +75,7 @@ class TestMLADeepseekV3BlockInt8(CustomTestCase):
        if torch.cuda.is_available() and torch.version.cuda:
            other_args.extend(
                [
-                    "--cuda-graph-max-bs",
+                    "--cuda-graph-max-bs-decode",
                    "16",
                    "--enable-torch-compile",
                    "--torch-compile-max-bs",