diff --git a/python/sglang/srt/models/qwen3.py b/python/sglang/srt/models/qwen3.py
index 2a18e2adb..ce8c98d5b 100644
--- a/python/sglang/srt/models/qwen3.py
+++ b/python/sglang/srt/models/qwen3.py
@@ -674,8 +674,10 @@ class Qwen3ForCausalLM(nn.Module):
         return self.model.embed_tokens.weight, self.lm_head.weight
 
     def set_embed_and_head(self, embed, head):
-        del self.model.embed_tokens.weight
-        del self.lm_head.weight
+        if hasattr(self.model.embed_tokens, "weight"):
+            del self.model.embed_tokens.weight
+        if hasattr(self.lm_head, "weight"):
+            del self.lm_head.weight
         self.model.embed_tokens.weight = embed
         self.lm_head.weight = head
         torch.cuda.empty_cache()