add and test finish 4 some models

vllm/model_executor/models/gpt_bigcode.py

@@ -25,6 +25,7 @@
 from transformers import GPTBigCodeConfig
 
 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.activation import get_act_fn
@@ -187,6 +188,7 @@ def forward(
         return hidden_states
 
 
+@support_torch_compile
 class GPTBigCodeModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/gpt_j.py

@@ -23,6 +23,7 @@
 from transformers import GPTJConfig
 
 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.activation import get_act_fn
@@ -174,6 +175,7 @@ def forward(
         return hidden_states
 
 
+@support_torch_compile
 class GPTJModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/gpt_neox.py

@@ -23,6 +23,7 @@
 from transformers import GPTNeoXConfig
 
 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.activation import get_act_fn
@@ -187,6 +188,7 @@ def forward(
         return hidden_states
 
 
+@support_torch_compile
 class GPTNeoXModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/granite.py

@@ -28,6 +28,7 @@
 from transformers import GraniteConfig
 
 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_rank,
                               get_tensor_model_parallel_world_size)
@@ -254,6 +255,7 @@ def forward(
         return hidden_states
 
 
+@support_torch_compile
 class GraniteModel(nn.Module):
 
     def __init__(

vllm/model_executor/models/internlm2.py

@@ -7,6 +7,7 @@
 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_rank,
                               get_tensor_model_parallel_world_size,
@@ -230,6 +231,7 @@ def forward(
         return hidden_states, residual
 
 
+@support_torch_compile
 class InternLM2Model(nn.Module):
 
     def __init__(