Added forward pass of test_inference_image_classification_head with torch.no_grad() (#14777)

tests/test_modeling_vit.py

@@ -347,6 +347,7 @@ class ViTModelIntegrationTest(unittest.TestCase):
         inputs = feature_extractor(images=image, return_tensors="pt").to(torch_device)
 
         # forward pass
+        with torch.no_grad():
             outputs = model(**inputs)
 
         # verify the logits