Adding "torch compile" annotations to moe models (vllm-project#9758)

vllm/model_executor/models/arctic.py

@@ -5,6 +5,7 @@
 from torch import nn
 
 from vllm.attention import Attention, AttentionMetadata
+from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig
 from vllm.distributed import (get_pp_group, get_tensor_model_parallel_rank,
                               get_tensor_model_parallel_world_size,
@@ -360,6 +361,7 @@ def forward(
         return hidden_states
 
 
+@support_torch_compile
 class ArcticModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/mixtral.py

@@ -28,6 +28,7 @@
 from transformers import MixtralConfig
 
 from vllm.attention import Attention, AttentionMetadata
+from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, LoRAConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.fused_moe import FusedMoE
@@ -245,6 +246,7 @@ def forward(
         return hidden_states, residual
 
 
+@support_torch_compile
 class MixtralModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/olmoe.py

@@ -17,6 +17,7 @@
 from transformers import PretrainedConfig
 
 from vllm.attention import Attention, AttentionMetadata
+from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.fused_moe import FusedMoE
@@ -239,6 +240,7 @@ def forward(
         return hidden_states, residual
 
 
+@support_torch_compile
 class OlmoeModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/phimoe.py

@@ -28,6 +28,7 @@
 from transformers.configuration_utils import PretrainedConfig
 
 from vllm.attention import Attention, AttentionMetadata
+from vllm.compilation.decorators import support_torch_compile
 from vllm.config import CacheConfig, LoRAConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
 from vllm.model_executor.layers.fused_moe import FusedMoE
@@ -429,6 +430,7 @@ def forward(
         return hidden_states, residual
 
 
+@support_torch_compile
 class PhiMoEModel(nn.Module):
 
     def __init__(