diff --git a/modelopt/torch/quantization/model_calib.py b/modelopt/torch/quantization/model_calib.py
index 9d95a0651..61e934f0e 100644
--- a/modelopt/torch/quantization/model_calib.py
+++ b/modelopt/torch/quantization/model_calib.py
@@ -1038,6 +1038,8 @@ def get_scale(x_max, w_max, alpha, tensor_parallel_group=None):
     def update_loss(self, out, out_actual, alpha):
         out_actual = out_actual[0] if isinstance(out_actual, tuple) else out_actual
         out = out[0] if isinstance(out, tuple) else out
+        out = out.to_local() if hasattr(out, "to_local") else out
+        out_actual = out_actual.to_local() if hasattr(out_actual, "to_local") else out_actual
         loss = (out - out_actual).float().pow(2).mean()
         self.awq_lite.loss[alpha] += loss.to(self.awq_lite.loss[alpha].device)