[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00
parent c8d3a402c1
commit 6c88f6c8d9
38 changed files with 342 additions and 299 deletions
--- a/test/srt/test_moe_deepep.py
+++ b/test/srt/test_moe_deepep.py
@@ -27,7 +27,8 @@ class TestPureTP(CustomTestCase):
                "--trust-remote-code",
                "--tp",
                "2",
-                "--enable-deepep-moe",
+                "--moe-a2a-backend",
+                "deepep",
                "--disable-cuda-graph",
            ],
        )
@@ -65,7 +66,8 @@ class TestDPAttn(unittest.TestCase):
                "--dp",
                "2",
                "--enable-dp-attention",
-                "--enable-deepep-moe",
+                "--moe-a2a-backend",
+                "deepep",
                "--deepep-mode",
                "normal",
                "--disable-cuda-graph",