[BugFix][LoRA] TritonExperts is ModularMoEPath for FP8 models (#33393)

Signed-off-by: Danielle Robinson <dmmaddix@amazon.com> Co-authored-by: Danielle Robinson <dmmaddix@amazon.com>
2026-01-30 07:27:42 -08:00
parent 8f5d51203b
commit 74898a7015
1 changed files with 1 additions and 3 deletions
--- a/vllm/lora/layers/fused_moe.py
+++ b/vllm/lora/layers/fused_moe.py
@@ -143,9 +143,7 @@ class FusedMoEWithLoRA(BaseLayerWithLoRA):
                m_fused_moe_fn.fused_experts, (MarlinExperts, UnfusedOAITritonExperts)
            )
        else:
-            assert isinstance(
-                m_fused_moe_fn.fused_experts, (MarlinExperts, TritonExperts)
-            )
+            assert isinstance(m_fused_moe_fn.fused_experts, TritonExperts)

        def fwd_decorator(layer, func):
            def wrapper(*args, **kwargs):