fix: typo in llama modeling file

aphrodite/modeling/models/llama.py

@@ -276,7 +276,7 @@ class LlamaForCausalLM(nn.Module):
 
 
             param = state_dict[name]
-            load_tensor_parallel_weights(param, loaded_weights, name,
+            load_tensor_parallel_weights(param, loaded_weight, name,
                                         self._column_parallel_weights,
                                         self._row_parallel_weights,
                                         tensor_model_parallel_rank)