diff --git a/vllm/model_executor/models/llama.py b/vllm/model_executor/models/llama.py
index 676a51ce67f9..d659d0a3f112 100644
--- a/vllm/model_executor/models/llama.py
+++ b/vllm/model_executor/models/llama.py
@@ -517,6 +517,8 @@ def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
                 weight_loader = getattr(param, "weight_loader",
                                         default_weight_loader)
                 weight_loader(param, loaded_weight)
+            if current_platform.is_hpu():
+                torch.hpu.synchronize()
 
     # If this function is called, it should always initialize KV cache scale
     # factors (or else raise an exception). Thus, handled exceptions should