Fix num_qps_per_rank computation when providing custom DeepEP configuration (#6468)

python/sglang/srt/layers/moe/ep_moe/token_dispatcher.py

@@ -67,9 +67,9 @@ class DeepEPBuffer:
         if deepep_mode.enable_normal():
             hidden_bytes = hidden_size * param_bytes
             for config in (
-                _DeepEPConfig.get_instance().normal_dispatch_config
+                DeepEPConfig.get_instance().normal_dispatch_config
                 or Buffer.get_dispatch_config(group.size()),
-                _DeepEPConfig.get_instance().normal_combine_config
+                DeepEPConfig.get_instance().normal_combine_config
                 or Buffer.get_combine_config(group.size()),
             ):
                 num_nvl_bytes = max(
@@ -97,7 +97,12 @@ class DeepEPBuffer:
             num_nvl_bytes,
             num_rdma_bytes,
             low_latency_mode=deepep_mode.enable_low_latency(),
-            num_qps_per_rank=(max(num_experts // group.size(), Buffer.num_sms // 2)),
+            num_qps_per_rank=(
+                max(
+                    num_experts // group.size(),
+                    DeepEPConfig.get_instance().num_sms // 2,
+                )
+            ),
         )
         return cls._buffer
 
@@ -122,7 +127,7 @@ class DeepEPBuffer:
         cls._dispatch_mode = DeepEPDispatchMode.LOW_LATENCY
 
 
-class _DeepEPConfig:
+class DeepEPConfig:
     _instance = None
 
     def __init__(self):
@@ -131,16 +136,23 @@ class _DeepEPConfig:
             config_parsed = load_json_config(config_str)
             if torch.distributed.get_rank() == 0:
                 logger.info(f"Use DeepEP Config: {config_parsed}")
-            self.normal_dispatch_config = Config(**config_parsed["normal_dispatch"])
-            self.normal_combine_config = Config(**config_parsed["normal_combine"])
+            config_dispatch = config_parsed["normal_dispatch"]
+            config_combine = config_parsed["normal_combine"]
+
+            self.normal_dispatch_config = Config(**config_dispatch)
+            self.normal_combine_config = Config(**config_combine)
+
+            assert config_dispatch["num_sms"] == config_combine["num_sms"]
+            self.num_sms = config_dispatch["num_sms"]
         else:
             self.normal_dispatch_config = None
             self.normal_combine_config = None
+            self.num_sms = Buffer.num_sms
 
     @classmethod
     def get_instance(cls):
         if cls._instance is None:
-            cls._instance = _DeepEPConfig()
+            cls._instance = DeepEPConfig()
         return cls._instance
 
 
@@ -326,7 +338,7 @@ class _DeepEPDispatcherImplNormal(_DeepEPDispatcherImplBase):
             async_finish=self.async_finish,
             allocate_on_comm_stream=(previous_event is not None) and self.async_finish,
             expert_alignment=128 if _ENABLE_JIT_DEEPGEMM else 1,
-            config=_DeepEPConfig.get_instance().normal_dispatch_config,
+            config=DeepEPConfig.get_instance().normal_dispatch_config,
         )
 
         get_global_expert_distribution_recorder().on_deepep_dispatch_normal(
@@ -433,7 +445,7 @@ class _DeepEPDispatcherImplNormal(_DeepEPDispatcherImplBase):
             async_finish=self.async_finish,
             previous_event=previous_event,
             allocate_on_comm_stream=previous_event is not None,
-            config=_DeepEPConfig.get_instance().normal_combine_config,
+            config=DeepEPConfig.get_instance().normal_combine_config,
         )
         return combined_x, event