[SpecDecode] Support EAGLE for Qwen3 MoE

Signed-off-by: seven-mile <i@7li.moe>

Author: seven-mile

vllm/model_executor/models/qwen3_moe.py

@@ -64,7 +64,7 @@
 from vllm.model_executor.models.utils import sequence_parallel_chunk
 from vllm.sequence import IntermediateTensors
 
-from .interfaces import MixtureOfExperts, SupportsLoRA, SupportsPP
+from .interfaces import MixtureOfExperts, SupportsEagle3, SupportsLoRA, SupportsPP
 from .utils import (
     AutoWeightsLoader,
     PPMissingLayer,
@@ -422,6 +422,7 @@ def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         self.make_empty_intermediate_tensors = make_empty_intermediate_tensors_factory(
             ["hidden_states", "residual"], config.hidden_size
         )
+        self.aux_hidden_state_layers = tuple[int, ...]()
 
     def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
         return self.embed_tokens(input_ids)
@@ -443,13 +444,25 @@ def forward(
             assert intermediate_tensors is not None
             hidden_states = intermediate_tensors["hidden_states"]
             residual = intermediate_tensors["residual"]
-        for layer in islice(self.layers, self.start_layer, self.end_layer):
+
+        aux_hidden_states = []
+        for idx, layer in enumerate(
+            islice(self.layers, self.start_layer, self.end_layer)
+        ):
+            if idx in self.aux_hidden_state_layers:
+                aux_hidden_states.append(hidden_states + residual)
             hidden_states, residual = layer(positions, hidden_states, residual)
+
         if not get_pp_group().is_last_rank:
             return IntermediateTensors(
                 {"hidden_states": hidden_states, "residual": residual}
             )
+
         hidden_states, _ = self.norm(hidden_states, residual)
+
+        if len(aux_hidden_states) > 0:
+            return hidden_states, aux_hidden_states
+
         return hidden_states
 
     def get_expert_mapping(self) -> list[tuple[str, str, int, str]]:
@@ -606,7 +619,9 @@ def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]) -> set[str]:
         return loaded_params
 
 
-class Qwen3MoeForCausalLM(nn.Module, SupportsPP, SupportsLoRA, MixtureOfExperts):
+class Qwen3MoeForCausalLM(
+    nn.Module, SupportsPP, SupportsLoRA, SupportsEagle3, MixtureOfExperts
+):
     packed_modules_mapping = {
         "qkv_proj": [
             "q_proj",
@@ -669,6 +684,13 @@ def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         self.num_routed_experts = example_layer.n_routed_experts
         self.num_redundant_experts = example_layer.n_redundant_experts
 
+    def set_aux_hidden_state_layers(self, layers: tuple[int, ...]) -> None:
+        self.model.aux_hidden_state_layers = layers
+
+    def get_eagle3_aux_hidden_state_layers(self) -> tuple[int, ...]:
+        num_layers = len(self.model.layers)
+        return (2, num_layers // 2, num_layers - 3)
+
     def set_eplb_state(
         self,
         expert_load_view: torch.Tensor,