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

This commit is contained in:
Cheng Wan
2025-08-01 01:20:03 -07:00
committed by GitHub
parent c8d3a402c1
commit 6c88f6c8d9
38 changed files with 342 additions and 299 deletions

View File

@@ -33,7 +33,8 @@ class TestTwoBatchOverlap(unittest.TestCase):
"--dp",
"2",
"--enable-dp-attention",
"--enable-deepep-moe",
"--moe-a2a-backend",
"deepep",
"--deepep-mode",
"normal",
"--disable-cuda-graph", # DeepEP normal does not support CUDA Graph
@@ -122,7 +123,8 @@ class TestQwen3TwoBatchOverlap(TestTwoBatchOverlap):
"--dp",
"2",
"--enable-dp-attention",
"--enable-deepep-moe",
"--moe-a2a-backend",
"deepep",
"--deepep-mode",
"normal",
"--disable-cuda-graph", # DeepEP normal does not support CUDA Graph