update lora args

setup.py

@@ -528,7 +528,7 @@ def get_version_add(sha: Optional[str] = None) -> str:
 try:
     __version__ = "0.10.2"
     __version_tuple__ = (0, 10, 2)
-    __hcu_version__ = f'0.10.2.rc1+{version}' 
+    __hcu_version__ = f'0.10.2+{version}' 
     
     from vllm.version import __version__, __version_tuple__, __hcu_version__
 except Exception as e:

vllm/config/lora.py

@@ -2,7 +2,7 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 import hashlib
-from typing import TYPE_CHECKING, Any, ClassVar, Literal, Optional, Union
+from typing import TYPE_CHECKING, Any, ClassVar, Literal, Optional, Union, List
 
 import torch
 from pydantic import ConfigDict
@@ -42,6 +42,10 @@ class LoRAConfig:
     max_cpu_loras: Optional[int] = None
     """Maximum number of LoRAs to store in CPU memory. Must be >= than
     `max_loras`."""
+    lora_target_modules: Optional[List[str]] = None
+    """List of lora module name, If not specified, 
+    modules will be chosen according to the model architecture.
+    """
     lora_dtype: Union[torch.dtype, LoRADType] = "auto"
     """Data type for LoRA. If auto, will default to base model dtype."""
     lora_extra_vocab_size: int = 256