diff --git a/examples/arm/image_classification_example_ethos_u/model_export/README.md b/examples/arm/image_classification_example_ethos_u/model_export/README.md
index 541edd75904..649737c54f3 100644
--- a/examples/arm/image_classification_example_ethos_u/model_export/README.md
+++ b/examples/arm/image_classification_example_ethos_u/model_export/README.md
@@ -30,6 +30,9 @@ Running this script achieves a test set accuracy of 86.10% in FP32.
 
 ## Export and quantize
 
+The exporter requires a safetensors checkpoint. The training script saves
+this format by default. Legacy PyTorch `.bin` checkpoints are not accepted.
+
 The `export_deit.py` script can be run as follows:
 
 ```bash
diff --git a/examples/arm/image_classification_example_ethos_u/model_export/export_deit.py b/examples/arm/image_classification_example_ethos_u/model_export/export_deit.py
index cd8557588b7..60632a85858 100644
--- a/examples/arm/image_classification_example_ethos_u/model_export/export_deit.py
+++ b/examples/arm/image_classification_example_ethos_u/model_export/export_deit.py
@@ -96,7 +96,7 @@ def measure_accuracy(quantized_model, test_set):
         type=str,
         default="./deit-tiny-oxford-pet/final_model",
         required=True,
-        help="Path to the fine-tuned ViT model.",
+        help="Path to the fine-tuned ViT model saved in safetensors format.",
     )
     argparser.add_argument(
         "--output-path",
@@ -123,6 +123,7 @@ def measure_accuracy(quantized_model, test_set):
         num_labels=37,
         ignore_mismatched_sizes=True,
         revision="b3428f18dcc7b543470d07f14b4a4157815d1880",
+        use_safetensors=True,
     ).eval()
     image_preprocessor = AutoImageProcessor.from_pretrained(
         "facebook/deit-tiny-patch16-224",