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

python/sglang/test/runners.py

@@ -499,7 +499,6 @@ class SRTRunner:
         chunked_prefill_size: Optional[int] = None,
         dp_size: int = 1,
         tokenizer_path: Optional[str] = None,
-        enable_ep_moe: bool = False,
         mem_fraction_static: float = 0.65,
         trust_remote_code: bool = False,
         speculative_draft_model_path: Optional[str] = None,
@@ -550,7 +549,6 @@ class SRTRunner:
             enable_dp_attention=enable_dp_attention,
             dp_size=dp_size,
             tokenizer_path=tokenizer_path,
-            enable_ep_moe=enable_ep_moe,
             disable_overlap_schedule=disable_overlap_schedule,
             cuda_graph_max_bs=cuda_graph_max_bs,
             disable_custom_all_reduce=disable_custom_all_reduce,