[hotfix] Fix deepep w4a8 bug (#12642)

2025-11-04 13:55:59 -08:00
parent 211f4070e5
commit 42889acbd0
1 changed files with 3 additions and 3 deletions
--- a/python/sglang/srt/layers/moe/token_dispatcher/deepep.py
+++ b/python/sglang/srt/layers/moe/token_dispatcher/deepep.py
@@ -357,9 +357,9 @@ class _DeepEPDispatcherImplNormal(_DeepEPDispatcherImplBase):
    ):
        topk_weights, topk_ids = topk_output.topk_weights, topk_output.topk_ids
        topk_ids = topk_ids.to(torch.int64)
-        if deep_gemm_wrapper.ENABLE_JIT_DEEPGEMM and not (
-            get_moe_runner_backend().is_cutlass()
-            and self.quant_config.get_name() == "w4afp8"
+        if (
+            deep_gemm_wrapper.ENABLE_JIT_DEEPGEMM
+            and not get_moe_runner_backend().is_cutlass()
        ):
            # TODO hard code 128 block quant,use fp8 communication
            hidden_states = sglang_per_token_group_quant_fp8(