[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/docs/references/disaggregation/lws-examples/p.yaml
+++ b/docs/references/disaggregation/lws-examples/p.yaml
@@ -38,9 +38,8 @@ spec:
          - --dp-size
          - "16"
          - --disable-radix-cache
-          - --enable-deepep-moe
-          - --deepep-mode
-          - normal
+          - --moe-a2a-backend
+          - deepep
          - --disaggregation-mode
          - prefill
          - --mem-fraction-static
@@ -184,9 +183,8 @@ spec:
          - --dp-size
          - "16"
          - --disable-radix-cache
-          - --enable-deepep-moe
-          - --deepep-mode
-          - normal
+          - --moe-a2a-backend
+          - deepep
          - --disaggregation-mode
          - prefill
          - --mem-fraction-static