diff --git a/optimum/intel/openvino/modeling.py b/optimum/intel/openvino/modeling.py
index 8cea5eb7b6..ceb547678b 100644
--- a/optimum/intel/openvino/modeling.py
+++ b/optimum/intel/openvino/modeling.py
@@ -129,7 +129,7 @@ def to(self, device: str):
         Use the specified `device` for inference. For example: "cpu" or "gpu". `device` can
         be in upper or lower case. To speed up first inference, call `.compile()` after `.to()`.
         """
-        self._device = device.upper()
+        self._device = str(device).upper()
         self.request = None
         return self
 
diff --git a/optimum/intel/openvino/modeling_diffusion.py b/optimum/intel/openvino/modeling_diffusion.py
index 4443381cd6..8595e33719 100644
--- a/optimum/intel/openvino/modeling_diffusion.py
+++ b/optimum/intel/openvino/modeling_diffusion.py
@@ -88,7 +88,7 @@ def __init__(
         **kwargs,
     ):
         self._internal_dict = config
-        self._device = device.upper()
+        self._device = str(device).upper()
         self.is_dynamic = dynamic_shapes
         self.ov_config = ov_config if ov_config is not None else {}
         self._model_save_dir = (
diff --git a/optimum/intel/openvino/modeling_seq2seq.py b/optimum/intel/openvino/modeling_seq2seq.py
index d43dbf3427..48df3b68bf 100644
--- a/optimum/intel/openvino/modeling_seq2seq.py
+++ b/optimum/intel/openvino/modeling_seq2seq.py
@@ -303,7 +303,7 @@ def __init__(
             pass
 
     def to(self, device: str):
-        self._device = device.upper()
+        self._device = str(device).upper()
         self.encoder._device = self._device
         self.decoder._device = self._device
         if self.use_cache: