diff --git a/lang_sam/models/__init__.py b/lang_sam/models/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/lang_sam/models/gdino.py b/lang_sam/models/gdino.py
index e9b6afe..4522a90 100644
--- a/lang_sam/models/gdino.py
+++ b/lang_sam/models/gdino.py
@@ -2,8 +2,11 @@
 import torch
 from PIL import Image
 from transformers import AutoModelForZeroShotObjectDetection, AutoProcessor
+from lang_sam.models.utils import get_device_type
+
+device_type = get_device_type()
+DEVICE = torch.device(device_type)
 
-DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 if torch.cuda.is_available():
     torch.autocast(device_type="cuda", dtype=torch.bfloat16).__enter__()
     if torch.cuda.get_device_properties(0).major >= 8:
@@ -18,7 +21,9 @@ def __init__(self):
     def build_model(self, ckpt_path: str | None = None):
         model_id = "IDEA-Research/grounding-dino-base"
         self.processor = AutoProcessor.from_pretrained(model_id)
-        self.model = AutoModelForZeroShotObjectDetection.from_pretrained(model_id).to(DEVICE)
+        self.model = AutoModelForZeroShotObjectDetection.from_pretrained(model_id).to(
+            DEVICE
+        )
 
     def predict(
         self,
diff --git a/lang_sam/models/sam.py b/lang_sam/models/sam.py
index e013b4e..220352e 100644
--- a/lang_sam/models/sam.py
+++ b/lang_sam/models/sam.py
@@ -5,8 +5,10 @@
 from omegaconf import OmegaConf
 from sam2.automatic_mask_generator import SAM2AutomaticMaskGenerator
 from sam2.sam2_image_predictor import SAM2ImagePredictor
+from lang_sam.models.utils import get_device_type
+
+DEVICE = torch.device(get_device_type())
 
-DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 if torch.cuda.is_available():
     torch.autocast(device_type="cuda", dtype=torch.bfloat16).__enter__()
     if torch.cuda.get_device_properties(0).major >= 8:
diff --git a/lang_sam/models/utils.py b/lang_sam/models/utils.py
new file mode 100644
index 0000000..bb4f003
--- /dev/null
+++ b/lang_sam/models/utils.py
@@ -0,0 +1,12 @@
+import logging
+import torch
+
+
+def get_device_type() -> str:
+    if torch.backends.mps.is_available():
+        return "mps"
+    elif torch.cuda.is_available():
+        return "cuda"
+    else:
+        logging.warning("No GPU found, using CPU instead")
+        return "cpu"