Support custom DeepEP tuning config (#6257)

python/sglang/srt/server_args.py

@@ -169,6 +169,7 @@ class ServerArgs:
     enable_ep_moe: bool = False
     enable_deepep_moe: bool = False
     deepep_mode: Optional[Literal["auto", "normal", "low_latency"]] = "auto"
+    deepep_config: Optional[str] = None
     enable_torch_compile: bool = False
     torch_compile_max_bs: int = 32
     cuda_graph_max_bs: Optional[int] = None
@@ -1249,6 +1250,12 @@ class ServerArgs:
             default="auto",
             help="Select the mode when enable DeepEP MoE, could be `normal`, `low_latency` or `auto`. Default is `auto`, which means `low_latency` for decode batch and `normal` for prefill batch.",
         )
+        parser.add_argument(
+            "--deepep-config",
+            type=str,
+            default=ServerArgs.deepep_config,
+            help="Tuned DeepEP config suitable for your own cluster.",
+        )
 
         parser.add_argument(
             "--n-share-experts-fusion",