[torch.compile] Adding torch compile annotations to some models (#9614)

vllm/model_executor/models/bloom.py

@@ -24,6 +24,7 @@ from torch import nn
 from transformers import BloomConfig
 
 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)
@@ -218,6 +219,7 @@ class BloomBlock(nn.Module):
         return output
 
 
+@support_torch_compile
 class BloomModel(nn.Module):
 
     def __init__(