Merge branch 'dev' into 8267-fix-normalize-intensity

monai/networks/nets/swin_unetr.py

@@ -782,9 +782,9 @@ def forward(self, x):
         x1 = x[:, 1::2, 0::2, 0::2, :]
         x2 = x[:, 0::2, 1::2, 0::2, :]
         x3 = x[:, 0::2, 0::2, 1::2, :]
-        x4 = x[:, 1::2, 0::2, 1::2, :]
-        x5 = x[:, 0::2, 1::2, 0::2, :]
-        x6 = x[:, 0::2, 0::2, 1::2, :]
+        x4 = x[:, 1::2, 1::2, 0::2, :]
+        x5 = x[:, 1::2, 0::2, 1::2, :]
+        x6 = x[:, 0::2, 1::2, 1::2, :]
         x7 = x[:, 1::2, 1::2, 1::2, :]
         x = torch.cat([x0, x1, x2, x3, x4, x5, x6, x7], -1)
         x = self.norm(x)

tests/test_load_image.py

@@ -217,7 +217,12 @@ def test_nibabel_reader(self, input_param, filenames, expected_shape):
     @SkipIfNoModule("kvikio")
     @parameterized.expand([TEST_CASE_GPU_1, TEST_CASE_GPU_2, TEST_CASE_GPU_3, TEST_CASE_GPU_4])
     def test_nibabel_reader_gpu(self, input_param, filenames, expected_shape):
-        test_image = np.random.rand(128, 128, 128)
+        if torch.__version__.endswith("nv24.8"):
+            # related issue: https://github.com/Project-MONAI/MONAI/issues/8274
+            # for this version, use randint test case to avoid the issue
+            test_image = torch.randint(0, 256, (128, 128, 128), dtype=torch.uint8).numpy()
+        else:
+            test_image = np.random.rand(128, 128, 128)
         with tempfile.TemporaryDirectory() as tempdir:
             for i, name in enumerate(filenames):
                 filenames[i] = os.path.join(tempdir, name)

tests/test_zarr_avg_merger.py

@@ -19,11 +19,18 @@
 from torch.nn.functional import pad
 
 from monai.inferers import ZarrAvgMerger
-from monai.utils import optional_import
+from monai.utils import get_package_version, optional_import, version_geq
 from tests.utils import assert_allclose
 
 np.seterr(divide="ignore", invalid="ignore")
 zarr, has_zarr = optional_import("zarr")
+if has_zarr:
+    if version_geq(get_package_version("zarr"), "3.0.0"):
+        directory_store = zarr.storage.LocalStore("test.zarr")
+    else:
+        directory_store = zarr.storage.DirectoryStore("test.zarr")
+else:
+    directory_store = None
 numcodecs, has_numcodecs = optional_import("numcodecs")
 
 TENSOR_4x4 = torch.randint(low=0, high=255, size=(2, 3, 4, 4), dtype=torch.float32)
@@ -154,7 +161,7 @@
 
 # explicit directory store
 TEST_CASE_10_DIRECTORY_STORE = [
-    dict(merged_shape=TENSOR_4x4.shape, store=zarr.storage.DirectoryStore("test.zarr")),
+    dict(merged_shape=TENSOR_4x4.shape, store=directory_store),
     [
         (TENSOR_4x4[..., :2, :2], (0, 0)),
         (TENSOR_4x4[..., :2, 2:], (0, 2)),