[V0.7.3][Qwen3] make v0.7.3 support qwen3

Signed-off-by: paulyu <paulyu0307@gmail.com>

Author: paulyu

vllm_ascend/models/qwen3.py

@@ -399,6 +399,16 @@ class Qwen3ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
         ],
     }
 
+    # LoRA specific attributes
+    supported_lora_modules = [
+        "qkv_proj",
+        "o_proj",
+        "gate_up_proj",
+        "down_proj",
+    ]
+    embedding_modules = {}
+    embedding_padding_modules = []
+
     def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         super().__init__()
         config = vllm_config.model_config.hf_config