Fix BeitForMaskedImageModeling (#13275)

* First pass

* Fix docs of bool_masked_pos

* Add integration script

* Fix docstring

* Add integration test for BeitForMaskedImageModeling

* Remove file

* Fix docs

tests/test_modeling_beit.py

@@ -379,6 +379,31 @@ class BeitModelIntegrationTest(unittest.TestCase):
             BeitFeatureExtractor.from_pretrained("microsoft/beit-base-patch16-224") if is_vision_available() else None
         )
 
+    @slow
+    def test_inference_masked_image_modeling_head(self):
+        model = BeitForMaskedImageModeling.from_pretrained("microsoft/beit-base-patch16-224-pt22k").to(torch_device)
+
+        feature_extractor = self.default_feature_extractor
+        image = prepare_img()
+        pixel_values = feature_extractor(images=image, return_tensors="pt").pixel_values.to(torch_device)
+
+        # prepare bool_masked_pos
+        bool_masked_pos = torch.ones((1, 196), dtype=torch.bool).to(torch_device)
+
+        # forward pass
+        outputs = model(pixel_values=pixel_values, bool_masked_pos=bool_masked_pos)
+        logits = outputs.logits
+
+        # verify the logits
+        expected_shape = torch.Size((1, 196, 8192))
+        self.assertEqual(logits.shape, expected_shape)
+
+        expected_slice = torch.tensor(
+            [[-3.2437, 0.5072, -13.9174], [-3.2456, 0.4948, -13.9401], [-3.2033, 0.5121, -13.8550]]
+        ).to(torch_device)
+
+        self.assertTrue(torch.allclose(logits[bool_masked_pos][:3, :3], expected_slice, atol=1e-2))
+
     @slow
     def test_inference_image_classification_head_imagenet_1k(self):
         model = BeitForImageClassification.from_pretrained("microsoft/beit-base-patch16-224").to(torch_device)