[BugFix] Fix combination of MTP and --n-share-experts-fusionwith R1

python/sglang/srt/models/deepseek_nextn.py

@@ -13,12 +13,14 @@
 # ==============================================================================
 
 """Inference-only DeepSeek NextN Speculative Decoding."""
+import logging
 from typing import Iterable, Optional, Tuple
 
 import torch
 from torch import nn
 from transformers import PretrainedConfig
 
+from sglang.srt.distributed import get_tensor_model_parallel_world_size
 from sglang.srt.layers.layernorm import RMSNorm
 from sglang.srt.layers.linear import ReplicatedLinear
 from sglang.srt.layers.logits_processor import LogitsProcessor
@@ -51,6 +53,9 @@
     from vllm._custom_ops import awq_dequantize
 
 
+logger = logging.getLogger(__name__)
+
+
 class DeepseekModelNextN(nn.Module):
     def __init__(
         self,
@@ -134,7 +139,9 @@ def __init__(
     ) -> None:
         nn.Module.__init__(self)
         self.config = config
+        self.tp_size = get_tensor_model_parallel_world_size()
         self.quant_config = quant_config
+        self.determine_n_share_experts_fusion("DeepseekV3ForCausalLMNextN")
 
         self.model = DeepseekModelNextN(
             config, quant_config, prefix=add_prefix("model", prefix)
@@ -182,6 +189,48 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
             ("gate_up_proj", "gate_proj", 0),
             ("gate_up_proj", "up_proj", 1),
         ]
+        if self.n_share_experts_fusion > 0:
+            logger.info(
+                f"Cloning {self.n_share_experts_fusion} "
+                "replicas of the shared expert into MoE for DeepseekV3ForCausalLMNextN"
+            )
+            weights_list = list(weights)
+            weights_dict = dict(weights_list)
+            if self.quant_config is None or self.quant_config.get_name() == "w8a8_int8":
+                suffix_list = [
+                    "down_proj.weight",
+                    "down_proj.weight_scale",
+                    "gate_proj.weight",
+                    "gate_proj.weight_scale",
+                    "up_proj.weight",
+                    "up_proj.weight_scale",
+                ]
+            else:
+                suffix_list = [
+                    "down_proj.weight",
+                    "down_proj.weight_scale_inv",
+                    "gate_proj.weight",
+                    "gate_proj.weight_scale_inv",
+                    "up_proj.weight",
+                    "up_proj.weight_scale_inv",
+                ]
+            names_to_remove = []
+            for num_repeat in range(self.n_share_experts_fusion):
+                for suffix in suffix_list:
+                    shared_expert_weight_name = (
+                        f"model.layers.0.mlp.shared_experts.{suffix}"
+                    )
+                    weights_list.append(
+                        (
+                            f"model.layers.0."
+                            f"mlp.experts."
+                            f"{self.config.n_routed_experts + num_repeat}"
+                            f".{suffix}",
+                            weights_dict[shared_expert_weight_name],
+                        )
+                    )
+                    names_to_remove += [shared_expert_weight_name]
+            weights = [w for w in weights_list if w[0] not in names_to_remove]
 
         # Params for weights, fp8 weight scales, fp8 activation scales
         # (param_name, weight_name, expert_id, shard_id)
@@ -190,7 +239,7 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
             ckpt_gate_proj_name="gate_proj",
             ckpt_down_proj_name="down_proj",
             ckpt_up_proj_name="up_proj",
-            num_experts=self.config.n_routed_experts,
+            num_experts=self.config.n_routed_experts + self.n_share_experts_fusion,
         )
 
         nextn_layer_prefix = "model.layers.0"

python/sglang/srt/models/deepseek_v2.py

@@ -1431,11 +1431,27 @@ def __init__(
         self.config = config
         self.tp_size = get_tensor_model_parallel_world_size()
         self.quant_config = quant_config
+        self.determine_n_share_experts_fusion()
+        self.model = DeepseekV2Model(
+            config, quant_config, prefix=add_prefix("model", prefix)
+        )
+        self.lm_head = ParallelLMHead(
+            config.vocab_size,
+            config.hidden_size,
+            quant_config=quant_config,
+            prefix=add_prefix("lm_head", prefix),
+        )
+        self.logits_processor = LogitsProcessor(config)
+        self.dp_size = get_attention_dp_size()
+
+    def determine_n_share_experts_fusion(
+        self, architecture: str = "DeepseekV3ForCausalLM"
+    ):
         self.n_share_experts_fusion = global_server_args_dict["n_share_experts_fusion"]
         if self.n_share_experts_fusion > 0:
             # Only Deepseek V3/R1 can use shared experts fusion optimization now.
             if (
-                self.config.architectures[0] != "DeepseekV3ForCausalLM"
+                self.config.architectures[0] != architecture
                 or self.config.n_routed_experts != 256
             ):
                 self.n_share_experts_fusion = 0
@@ -1450,7 +1466,7 @@ def __init__(
         elif self.n_share_experts_fusion == 0:
             if (
                 torch.cuda.get_device_capability("cuda") >= (9, 0)
-                and self.config.architectures[0] == "DeepseekV3ForCausalLM"
+                and self.config.architectures[0] == architecture
                 and self.config.n_routed_experts == 256
                 and (not global_server_args_dict["enable_deepep_moe"])
             ):
@@ -1460,18 +1476,6 @@ def __init__(
                     "Deepseek V3/R1 with fp8 can use shared experts fusion optimization when SM version >=90. Shared experts fusion optimization is enabled."
                 )
 
-        self.model = DeepseekV2Model(
-            config, quant_config, prefix=add_prefix("model", prefix)
-        )
-        self.lm_head = ParallelLMHead(
-            config.vocab_size,
-            config.hidden_size,
-            quant_config=quant_config,
-            prefix=add_prefix("lm_head", prefix),
-        )
-        self.logits_processor = LogitsProcessor(config)
-        self.dp_size = get_attention_dp_size()
-
     def get_input_embeddings(self) -> nn.Embedding:
         return self.model.embed_tokens