diff --git a/monai/apps/auto3dseg/auto_runner.py b/monai/apps/auto3dseg/auto_runner.py
index 8296973f6e..12566c0d34 100644
--- a/monai/apps/auto3dseg/auto_runner.py
+++ b/monai/apps/auto3dseg/auto_runner.py
@@ -36,7 +36,7 @@
 from monai.auto3dseg.utils import algo_to_pickle
 from monai.bundle import ConfigParser
 from monai.transforms import SaveImage
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 from monai.utils.module import look_up_option, optional_import
 
 logger = get_logger(module_name=__name__)
@@ -636,11 +636,11 @@ def _train_algo_in_sequence(self, history: list[dict[str, Any]]) -> None:
             progress.yaml, accuracies in CSV and a pickle file of the Algo object.
         """
         for algo_dict in history:
-            algo = algo_dict[AlgoEnsembleKeys.ALGO]
+            algo = algo_dict[AlgoKeys.ALGO]
             algo.train(self.train_params)
             acc = algo.get_score()
 
-            algo_meta_data = {str(AlgoEnsembleKeys.SCORE): acc}
+            algo_meta_data = {str(AlgoKeys.SCORE): acc}
             algo_to_pickle(algo, template_path=algo.template_path, **algo_meta_data)
 
     def _train_algo_in_nni(self, history: list[dict[str, Any]]) -> None:
@@ -675,8 +675,8 @@ def _train_algo_in_nni(self, history: list[dict[str, Any]]) -> None:
         last_total_tasks = len(import_bundle_algo_history(self.work_dir, only_trained=True))
         mode_dry_run = self.hpo_params.pop("nni_dry_run", False)
         for algo_dict in history:
-            name = algo_dict[AlgoEnsembleKeys.ID]
-            algo = algo_dict[AlgoEnsembleKeys.ALGO]
+            name = algo_dict[AlgoKeys.ID]
+            algo = algo_dict[AlgoKeys.ALGO]
             nni_gen = NNIGen(algo=algo, params=self.hpo_params)
             obj_filename = nni_gen.get_obj_filename()
             nni_config = deepcopy(default_nni_config)
@@ -772,13 +772,13 @@ def run(self):
                 )
 
             if auto_train_choice:
-                skip_algos = [h[AlgoEnsembleKeys.ID] for h in history if h["is_trained"]]
+                skip_algos = [h[AlgoKeys.ID] for h in history if h[AlgoKeys.IS_TRAINED]]
                 if len(skip_algos) > 0:
                     logger.info(
                         f"Skipping already trained algos {skip_algos}."
                         "Set option train=True to always retrain all algos."
                     )
-                    history = [h for h in history if not h["is_trained"]]
+                    history = [h for h in history if not h[AlgoKeys.IS_TRAINED]]
 
             if len(history) > 0:
                 if not self.hpo:
@@ -794,13 +794,13 @@ def run(self):
         if self.ensemble:
             history = import_bundle_algo_history(self.work_dir, only_trained=False)
 
-            history_untrained = [h for h in history if not h["is_trained"]]
+            history_untrained = [h for h in history if not h[AlgoKeys.IS_TRAINED]]
             if len(history_untrained) > 0:
                 warnings.warn(
                     f"Ensembling step will skip {[h['name'] for h in history_untrained]} untrained algos."
                     "Generally it means these algos did not complete training."
                 )
-                history = [h for h in history if h["is_trained"]]
+                history = [h for h in history if h[AlgoKeys.IS_TRAINED]]
 
             if len(history) == 0:
                 raise ValueError(
@@ -816,7 +816,7 @@ def run(self):
             if len(preds) > 0:
                 logger.info("Auto3Dseg picked the following networks to ensemble:")
                 for algo in ensembler.get_algo_ensemble():
-                    logger.info(algo[AlgoEnsembleKeys.ID])
+                    logger.info(algo[AlgoKeys.ID])
 
                 for pred in preds:
                     self.save_image(pred)
diff --git a/monai/apps/auto3dseg/bundle_gen.py b/monai/apps/auto3dseg/bundle_gen.py
index 3f34e4453f..51fcf684be 100644
--- a/monai/apps/auto3dseg/bundle_gen.py
+++ b/monai/apps/auto3dseg/bundle_gen.py
@@ -33,7 +33,7 @@
 from monai.auto3dseg.utils import algo_to_pickle
 from monai.bundle.config_parser import ConfigParser
 from monai.utils import ensure_tuple
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 
 logger = get_logger(module_name=__name__)
 ALGO_HASH = os.environ.get("MONAI_ALGO_HASH", "7758ad1")
@@ -539,5 +539,5 @@ def generate(
 
                 algo_to_pickle(gen_algo, template_path=algo.template_path)
                 self.history.append(
-                    {AlgoEnsembleKeys.ID: name, AlgoEnsembleKeys.ALGO: gen_algo}
+                    {AlgoKeys.ID: name, AlgoKeys.ALGO: gen_algo}
                 )  # track the previous, may create a persistent history
diff --git a/monai/apps/auto3dseg/ensemble_builder.py b/monai/apps/auto3dseg/ensemble_builder.py
index dffadb818d..3a77fd9c05 100644
--- a/monai/apps/auto3dseg/ensemble_builder.py
+++ b/monai/apps/auto3dseg/ensemble_builder.py
@@ -27,7 +27,7 @@
 from monai.auto3dseg.utils import datafold_read
 from monai.bundle import ConfigParser
 from monai.transforms import MeanEnsemble, VoteEnsemble
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 from monai.utils.misc import prob2class
 from monai.utils.module import look_up_option
 
@@ -59,7 +59,7 @@ def get_algo(self, identifier):
             identifier: the name of the bundleAlgo
         """
         for algo in self.algos:
-            if identifier == algo[AlgoEnsembleKeys.ID]:
+            if identifier == algo[AlgoKeys.ID]:
                 return algo
 
     def get_algo_ensemble(self):
@@ -160,7 +160,7 @@ def __call__(self, pred_param: dict[str, Any] | None = None) -> list[torch.Tenso
             print(i)
             preds = []
             for algo in self.algo_ensemble:
-                infer_instance = algo[AlgoEnsembleKeys.ALGO]
+                infer_instance = algo[AlgoKeys.ALGO]
                 pred = infer_instance.predict(predict_files=[file], predict_params=param)
                 preds.append(pred[0])
             outputs.append(self.ensemble_pred(preds, sigmoid=sigmoid))
@@ -187,7 +187,7 @@ def sort_score(self):
         """
         Sort the best_metrics
         """
-        scores = concat_val_to_np(self.algos, [AlgoEnsembleKeys.SCORE])
+        scores = concat_val_to_np(self.algos, [AlgoKeys.SCORE])
         return np.argsort(scores).tolist()
 
     def collect_algos(self, n_best: int = -1) -> None:
@@ -238,14 +238,14 @@ def collect_algos(self) -> None:
             best_model: BundleAlgo | None = None
             for algo in self.algos:
                 # algorithm folder: {net}_{fold_index}_{other}
-                identifier = algo[AlgoEnsembleKeys.ID].split("_")[1]
+                identifier = algo[AlgoKeys.ID].split("_")[1]
                 try:
                     algo_id = int(identifier)
                 except ValueError as err:
                     raise ValueError(f"model identifier {identifier} is not number.") from err
-                if algo_id == f_idx and algo[AlgoEnsembleKeys.SCORE] > best_score:
+                if algo_id == f_idx and algo[AlgoKeys.SCORE] > best_score:
                     best_model = algo
-                    best_score = algo[AlgoEnsembleKeys.SCORE]
+                    best_score = algo[AlgoKeys.SCORE]
             self.algo_ensemble.append(best_model)
 
 
@@ -268,7 +268,7 @@ class AlgoEnsembleBuilder:
     """
 
     def __init__(self, history: Sequence[dict[str, Any]], data_src_cfg_filename: str | None = None):
-        self.infer_algos: list[dict[AlgoEnsembleKeys, Any]] = []
+        self.infer_algos: list[dict[AlgoKeys, Any]] = []
         self.ensemble: AlgoEnsemble
         self.data_src_cfg = ConfigParser(globals=False)
 
@@ -278,8 +278,8 @@ def __init__(self, history: Sequence[dict[str, Any]], data_src_cfg_filename: str
         for algo_dict in history:
             # load inference_config_paths
 
-            name = algo_dict[AlgoEnsembleKeys.ID]
-            gen_algo = algo_dict[AlgoEnsembleKeys.ALGO]
+            name = algo_dict[AlgoKeys.ID]
+            gen_algo = algo_dict[AlgoKeys.ALGO]
 
             best_metric = gen_algo.get_score()
             algo_path = gen_algo.output_path
@@ -306,7 +306,7 @@ def add_inferer(self, identifier: str, gen_algo: BundleAlgo, best_metric: float
         if best_metric is None:
             raise ValueError("Feature to re-validate is to be implemented")
 
-        algo = {AlgoEnsembleKeys.ID: identifier, AlgoEnsembleKeys.ALGO: gen_algo, AlgoEnsembleKeys.SCORE: best_metric}
+        algo = {AlgoKeys.ID: identifier, AlgoKeys.ALGO: gen_algo, AlgoKeys.SCORE: best_metric}
         self.infer_algos.append(algo)
 
     def set_ensemble_method(self, ensemble: AlgoEnsemble, *args: Any, **kwargs: Any) -> None:
diff --git a/monai/apps/auto3dseg/hpo_gen.py b/monai/apps/auto3dseg/hpo_gen.py
index 8acf56e7c6..87ba3fa84f 100644
--- a/monai/apps/auto3dseg/hpo_gen.py
+++ b/monai/apps/auto3dseg/hpo_gen.py
@@ -23,7 +23,7 @@
 from monai.bundle.config_parser import ConfigParser
 from monai.config import PathLike
 from monai.utils import optional_import
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 
 nni, has_nni = optional_import("nni")
 optuna, has_optuna = optional_import("optuna")
@@ -99,8 +99,8 @@ class NNIGen(HPOGen):
             # Bundle Algorithms are already generated by BundleGen in work_dir
             import_bundle_algo_history(work_dir, only_trained=False)
             algo_dict = self.history[0]  # pick the first algorithm
-            algo_name = algo_dict[AlgoEnsembleKeys.ID]
-            onealgo = algo_dict[AlgoEnsembleKeys.ALGO]
+            algo_name = algo_dict[AlgoKeys.ID]
+            onealgo = algo_dict[AlgoKeys.ALGO]
             nni_gen = NNIGen(algo=onealgo)
             nni_gen.print_bundle_algo_instruction()
 
@@ -238,7 +238,7 @@ def run_algo(self, obj_filename: str, output_folder: str = ".", template_path: P
         self.algo.train(self.params)
         # step 4 report validation acc to controller
         acc = self.algo.get_score()
-        algo_meta_data = {str(AlgoEnsembleKeys.SCORE): acc}
+        algo_meta_data = {str(AlgoKeys.SCORE): acc}
 
         if isinstance(self.algo, BundleAlgo):
             algo_to_pickle(self.algo, template_path=self.algo.template_path, **algo_meta_data)
@@ -411,7 +411,7 @@ def run_algo(self, obj_filename: str, output_folder: str = ".", template_path: P
         self.algo.train(self.params)
         # step 4 report validation acc to controller
         acc = self.algo.get_score()
-        algo_meta_data = {str(AlgoEnsembleKeys.SCORE): acc}
+        algo_meta_data = {str(AlgoKeys.SCORE): acc}
         if isinstance(self.algo, BundleAlgo):
             algo_to_pickle(self.algo, template_path=self.algo.template_path, **algo_meta_data)
         else:
diff --git a/monai/apps/auto3dseg/utils.py b/monai/apps/auto3dseg/utils.py
index 9382aff569..90de5e8f75 100644
--- a/monai/apps/auto3dseg/utils.py
+++ b/monai/apps/auto3dseg/utils.py
@@ -15,7 +15,7 @@
 
 from monai.apps.auto3dseg.bundle_gen import BundleAlgo
 from monai.auto3dseg import algo_from_pickle, algo_to_pickle
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 
 
 def import_bundle_algo_history(
@@ -49,17 +49,12 @@ def import_bundle_algo_history(
         if isinstance(algo, BundleAlgo):  # algo's template path needs override
             algo.template_path = algo_meta_data["template_path"]
 
-        best_metric = algo_meta_data.get(AlgoEnsembleKeys.SCORE, None)
+        best_metric = algo_meta_data.get(AlgoKeys.SCORE, None)
         is_trained = best_metric is not None
 
         if (only_trained and is_trained) or not only_trained:
             history.append(
-                {
-                    AlgoEnsembleKeys.ID: name,
-                    AlgoEnsembleKeys.ALGO: algo,
-                    AlgoEnsembleKeys.SCORE: best_metric,
-                    "is_trained": is_trained,
-                }
+                {AlgoKeys.ID: name, AlgoKeys.ALGO: algo, AlgoKeys.SCORE: best_metric, AlgoKeys.IS_TRAINED: is_trained}
             )
 
     return history
@@ -73,5 +68,5 @@ def export_bundle_algo_history(history: list[dict[str, BundleAlgo]]) -> None:
         history: a List of Bundle. Typically, the history can be obtained from BundleGen get_history method
     """
     for algo_dict in history:
-        algo = algo_dict[AlgoEnsembleKeys.ALGO]
+        algo = algo_dict[AlgoKeys.ALGO]
         algo_to_pickle(algo, template_path=algo.template_path)
diff --git a/monai/data/meta_tensor.py b/monai/data/meta_tensor.py
index 48b9320f99..e3aacb95ee 100644
--- a/monai/data/meta_tensor.py
+++ b/monai/data/meta_tensor.py
@@ -510,9 +510,16 @@ def new_empty(self, size, dtype=None, device=None, requires_grad=False):
             self.as_tensor().new_empty(size=size, dtype=dtype, device=device, requires_grad=requires_grad)
         )
 
-    def clone(self):
-        """returns a copy of the MetaTensor instance."""
-        new_inst = MetaTensor(self.as_tensor().clone())
+    def clone(self, **kwargs):
+        """
+        Returns a copy of the MetaTensor instance.
+
+        Args:
+            kwargs: additional keyword arguments to `torch.clone`.
+
+        See also: https://pytorch.org/docs/stable/generated/torch.clone.html
+        """
+        new_inst = MetaTensor(self.as_tensor().clone(**kwargs))
         new_inst.__dict__ = deepcopy(self.__dict__)
         return new_inst
 
diff --git a/monai/inferers/utils.py b/monai/inferers/utils.py
index 0e1f0f4a42..b171d20ebb 100644
--- a/monai/inferers/utils.py
+++ b/monai/inferers/utils.py
@@ -41,7 +41,7 @@
 
 
 def sliding_window_inference(
-    inputs: torch.Tensor,
+    inputs: torch.Tensor | MetaTensor,
     roi_size: Sequence[int] | int,
     sw_batch_size: int,
     predictor: Callable[..., torch.Tensor | Sequence[torch.Tensor] | dict[Any, torch.Tensor]],
@@ -307,9 +307,11 @@ def sliding_window_inference(
             output_image_list[ss] = output_i[(slice(None), slice(None), *final_slicing)]
 
     final_output = _pack_struct(output_image_list, dict_keys)
-    final_output = convert_to_dst_type(final_output, inputs, device=device)[0]  # type: ignore
     if temp_meta is not None:
-        final_output = MetaTensor(final_output).copy_meta_from(temp_meta)
+        final_output = convert_to_dst_type(final_output, temp_meta, device=device)[0]  # type: ignore
+    else:
+        final_output = convert_to_dst_type(final_output, inputs, device=device)[0]
+
     return final_output  # type: ignore
 
 
@@ -322,7 +324,7 @@ def _create_buffered_slices(slices, batch_size, sw_batch_size, buffer_dim, buffe
 
     _, _, _b_lens = np.unique(slices_np[:, 0], return_counts=True, return_index=True)
     b_ends = np.cumsum(_b_lens).tolist()  # possible buffer flush boundaries
-    x = [0, *b_ends][:: min(len(b_ends), int(buffer_steps))]  # type: ignore
+    x = [0, *b_ends][:: min(len(b_ends), int(buffer_steps))]
     if x[-1] < b_ends[-1]:
         x.append(b_ends[-1])
     n_per_batch = len(x) - 1
@@ -385,7 +387,7 @@ def _flatten_struct(seg_out):
         dict_keys = sorted(seg_out.keys())  # track predictor's output keys
         seg_probs = tuple(seg_out[k] for k in dict_keys)
     else:
-        seg_probs = ensure_tuple(seg_out)  # type: ignore
+        seg_probs = ensure_tuple(seg_out)
     return dict_keys, seg_probs
 
 
diff --git a/monai/networks/utils.py b/monai/networks/utils.py
index 962db956c7..d264c22e28 100644
--- a/monai/networks/utils.py
+++ b/monai/networks/utils.py
@@ -672,7 +672,7 @@ def convert_to_onnx(
         set_determinism(seed=None)
         # compare onnx/ort and PyTorch results
         for r1, r2 in zip(torch_out, onnx_out):
-            torch.testing.assert_allclose(r1.cpu(), r2, rtol=rtol, atol=atol)  # type: ignore
+            torch.testing.assert_allclose(r1.cpu(), r2, rtol=rtol, atol=atol)
 
     return onnx_model
 
diff --git a/monai/transforms/spatial/array.py b/monai/transforms/spatial/array.py
index 2b61b1f61a..f263e89152 100644
--- a/monai/transforms/spatial/array.py
+++ b/monai/transforms/spatial/array.py
@@ -230,6 +230,7 @@ def inverse(self, data: torch.Tensor) -> torch.Tensor:
         with self.trace_transform(False):
             # we can't use `self.__call__` in case a child class calls this inverse.
             out: torch.Tensor = SpatialResample.__call__(self, data, **kw_args)
+        kw_args["src_affine"] = kw_args.get("dst_affine")
         return out
 
 
diff --git a/monai/utils/enums.py b/monai/utils/enums.py
index 6b01e43b47..a7ea9e29a8 100644
--- a/monai/utils/enums.py
+++ b/monai/utils/enums.py
@@ -14,6 +14,8 @@
 import random
 from enum import Enum
 
+from monai.utils import deprecated
+
 __all__ = [
     "StrEnum",
     "NumpyPadMode",
@@ -56,6 +58,7 @@
     "LazyAttr",
     "BundleProperty",
     "BundlePropertyConfig",
+    "AlgoKeys",
 ]
 
 
@@ -592,6 +595,7 @@ class LabelStatsKeys(StrEnum):
     LABEL_NCOMP = "ncomponents"
 
 
+@deprecated(since="1.2", msg_suffix="please use `AlgoKeys` instead.")
 class AlgoEnsembleKeys(StrEnum):
     """
     Default keys for Mixed Ensemble
@@ -664,3 +668,18 @@ class BundlePropertyConfig(StrEnum):
 
     ID = "id"
     REF_ID = "refer_id"
+
+
+class AlgoKeys(StrEnum):
+    """
+    Default keys for templated Auto3DSeg Algo.
+    `ID` is the identifier of the algorithm. The string has the format of <name>_<idx>_<other>.
+    `ALGO` is the Auto3DSeg Algo instance.
+    `IS_TRAINED` is the status that shows if the Algo has been trained.
+    `SCORE` is the score the Algo has achieved after training.
+    """
+
+    ID = "identifier"
+    ALGO = "algo_instance"
+    IS_TRAINED = "is_trained"
+    SCORE = "best_metric"
diff --git a/monai/utils/type_conversion.py b/monai/utils/type_conversion.py
index c5dd3a797c..734d8a2b17 100644
--- a/monai/utils/type_conversion.py
+++ b/monai/utils/type_conversion.py
@@ -45,7 +45,7 @@
 
 def get_numpy_dtype_from_string(dtype: str) -> np.dtype:
     """Get a numpy dtype (e.g., `np.float32`) from its string (e.g., `"float32"`)."""
-    return np.empty([], dtype=dtype).dtype
+    return np.empty([], dtype=str(dtype).split(".")[-1]).dtype
 
 
 def get_torch_dtype_from_string(dtype: str) -> torch.dtype:
diff --git a/tests/test_auto3dseg_ensemble.py b/tests/test_auto3dseg_ensemble.py
index f2170f1dfc..2c8d615c58 100644
--- a/tests/test_auto3dseg_ensemble.py
+++ b/tests/test_auto3dseg_ensemble.py
@@ -23,7 +23,7 @@
 from monai.bundle.config_parser import ConfigParser
 from monai.data import create_test_image_3d
 from monai.utils import optional_import, set_determinism
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 from tests.utils import (
     SkipIfBeforePyTorchVersion,
     get_testing_algo_template_path,
@@ -135,8 +135,8 @@ def test_ensemble(self) -> None:
         history = bundle_generator.get_history()
 
         for algo_dict in history:
-            name = algo_dict[AlgoEnsembleKeys.ID]
-            algo = algo_dict[AlgoEnsembleKeys.ALGO]
+            name = algo_dict[AlgoKeys.ID]
+            algo = algo_dict[AlgoKeys.ALGO]
             _train_param = train_param.copy()
             if name.startswith("segresnet"):
                 _train_param["network#init_filters"] = 8
@@ -148,7 +148,7 @@ def test_ensemble(self) -> None:
         builder = AlgoEnsembleBuilder(history, data_src_cfg)
         builder.set_ensemble_method(AlgoEnsembleBestN(n_best=1))
         ensemble = builder.get_ensemble()
-        name = ensemble.get_algo_ensemble()[0][AlgoEnsembleKeys.ID]
+        name = ensemble.get_algo_ensemble()[0][AlgoKeys.ID]
         if name.startswith("segresnet"):
             pred_param["network#init_filters"] = 8
         elif name.startswith("swinunetr"):
@@ -159,7 +159,7 @@ def test_ensemble(self) -> None:
         builder.set_ensemble_method(AlgoEnsembleBestByFold(1))
         ensemble = builder.get_ensemble()
         for algo in ensemble.get_algo_ensemble():
-            print(algo[AlgoEnsembleKeys.ID])
+            print(algo[AlgoKeys.ID])
 
     def tearDown(self) -> None:
         set_determinism(None)
diff --git a/tests/test_auto3dseg_hpo.py b/tests/test_auto3dseg_hpo.py
index 00b13adc35..60298ae254 100644
--- a/tests/test_auto3dseg_hpo.py
+++ b/tests/test_auto3dseg_hpo.py
@@ -24,7 +24,7 @@
 from monai.bundle.config_parser import ConfigParser
 from monai.data import create_test_image_3d
 from monai.utils import optional_import
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 from tests.utils import (
     SkipIfBeforePyTorchVersion,
     get_testing_algo_template_path,
@@ -140,7 +140,7 @@ def setUp(self) -> None:
     @skip_if_no_cuda
     def test_run_algo(self) -> None:
         algo_dict = self.history[0]
-        algo = algo_dict[AlgoEnsembleKeys.ALGO]
+        algo = algo_dict[AlgoKeys.ALGO]
         nni_gen = NNIGen(algo=algo, params=override_param)
         obj_filename = nni_gen.get_obj_filename()
         # this function will be used in HPO via Python Fire
@@ -150,7 +150,7 @@ def test_run_algo(self) -> None:
     @skip_if_no_optuna
     def test_run_optuna(self) -> None:
         algo_dict = self.history[0]
-        algo = algo_dict[AlgoEnsembleKeys.ALGO]
+        algo = algo_dict[AlgoKeys.ALGO]
 
         class OptunaGenLearningRate(OptunaGen):
             def get_hyperparameters(self):
@@ -172,7 +172,7 @@ def get_hyperparameters(self):
     @skip_if_no_cuda
     def test_get_history(self) -> None:
         algo_dict = self.history[0]
-        algo = algo_dict[AlgoEnsembleKeys.ALGO]
+        algo = algo_dict[AlgoKeys.ALGO]
         nni_gen = NNIGen(algo=algo, params=override_param)
         obj_filename = nni_gen.get_obj_filename()
 
diff --git a/tests/test_get_equivalent_dtype.py b/tests/test_get_equivalent_dtype.py
index 01f8adca73..299a3963b7 100644
--- a/tests/test_get_equivalent_dtype.py
+++ b/tests/test_get_equivalent_dtype.py
@@ -42,7 +42,15 @@ def test_native_type(self):
                 out_dtype = get_equivalent_dtype(n, type(im_dtype))
                 self.assertEqual(out_dtype, n)
 
-    @parameterized.expand([["float", np.float64], ["float32", np.float32], ["float64", np.float64]])
+    @parameterized.expand(
+        [
+            ["float", np.float64],
+            ["float32", np.float32],
+            ["np.float32", np.float32],
+            ["float64", np.float64],
+            ["torch.float64", np.float64],
+        ]
+    )
     def test_from_string(self, dtype_str, expected_np):
         expected_pt = get_equivalent_dtype(expected_np, torch.Tensor)
         # numpy
diff --git a/tests/test_integration_classification_2d.py b/tests/test_integration_classification_2d.py
index 9aef9cbe18..91ce331a4a 100644
--- a/tests/test_integration_classification_2d.py
+++ b/tests/test_integration_classification_2d.py
@@ -266,7 +266,7 @@ def test_training(self):
             repeated.append(results)
         np.testing.assert_allclose(repeated[0], repeated[1])
 
-    @TimedCall(seconds=1000, skip_timing=not torch.cuda.is_available(), force_quit=False, daemon=False)
+    @TimedCall(seconds=2000, skip_timing=not torch.cuda.is_available(), force_quit=False, daemon=False)
     def test_timing(self):
         self.train_and_infer()
 
diff --git a/tests/test_integration_gpu_customization.py b/tests/test_integration_gpu_customization.py
index 1f871f246e..4a7490e529 100644
--- a/tests/test_integration_gpu_customization.py
+++ b/tests/test_integration_gpu_customization.py
@@ -23,7 +23,7 @@
 from monai.bundle.config_parser import ConfigParser
 from monai.data import create_test_image_3d
 from monai.utils import optional_import
-from monai.utils.enums import AlgoEnsembleKeys
+from monai.utils.enums import AlgoKeys
 from tests.utils import (
     SkipIfBeforePyTorchVersion,
     get_testing_algo_template_path,
@@ -139,7 +139,7 @@ def test_ensemble_gpu_customization(self) -> None:
         history = bundle_generator.get_history()
 
         for algo_dict in history:
-            algo = algo_dict[AlgoEnsembleKeys.ALGO]
+            algo = algo_dict[AlgoKeys.ALGO]
             algo.train(train_param)
 
         builder = AlgoEnsembleBuilder(history, data_src_cfg)
@@ -151,7 +151,7 @@ def test_ensemble_gpu_customization(self) -> None:
         builder.set_ensemble_method(AlgoEnsembleBestByFold(1))
         ensemble = builder.get_ensemble()
         for algo in ensemble.get_algo_ensemble():
-            print(algo[AlgoEnsembleKeys.ID])
+            print(algo[AlgoKeys.ID])
 
     def tearDown(self) -> None:
         self.test_dir.cleanup()
diff --git a/tests/test_meta_tensor.py b/tests/test_meta_tensor.py
index a6607a3ccd..e547675a0e 100644
--- a/tests/test_meta_tensor.py
+++ b/tests/test_meta_tensor.py
@@ -177,6 +177,7 @@ def test_copy(self, device, dtype):
         a = deepcopy(m)
         self.check(a, m, ids=False)
         # clone
+        a = m.clone(memory_format=torch.preserve_format)
         a = m.clone()
         self.check(a, m, ids=False)
         a = MetaTensor([[]], device=device, dtype=dtype)
diff --git a/tests/test_sliding_window_hovernet_inference.py b/tests/test_sliding_window_hovernet_inference.py
index 8f7f8346cc..b17e8525ec 100644
--- a/tests/test_sliding_window_hovernet_inference.py
+++ b/tests/test_sliding_window_hovernet_inference.py
@@ -18,6 +18,7 @@
 from parameterized import parameterized
 
 from monai.apps.pathology.inferers import SlidingWindowHoVerNetInferer
+from monai.data import MetaTensor
 from monai.inferers import sliding_window_inference
 from monai.utils import optional_import
 from tests.test_sliding_window_inference import TEST_CASES
@@ -31,6 +32,8 @@
     ["hover", (1, 3, 16, 8), (4, 4), 7, 0.5, "constant", torch.device("cpu:0"), (1,) * 4],
 ]
 
+TEST_CASES_MULTIOUTPUT = [[torch.ones((1, 6, 20, 20))], [MetaTensor(torch.ones((1, 6, 20, 20)))]]
+
 
 class TestSlidingWindowHoVerNetInference(unittest.TestCase):
     @parameterized.expand(TEST_CASES_PADDING)
@@ -245,9 +248,10 @@ def compute(data, test1, test2):
         )(inputs, compute, t1, test2=t2)
         np.testing.assert_allclose(result.cpu().numpy(), expected, rtol=1e-4)
 
-    def test_multioutput(self):
+    @parameterized.expand(TEST_CASES_MULTIOUTPUT)
+    def test_multioutput(self, inputs):
         device = "cuda" if torch.cuda.is_available() else "cpu:0"
-        inputs = torch.ones((1, 6, 20, 20)).to(device=device)
+        inputs = inputs.to(device=device)
         roi_shape = (8, 8)
         sw_batch_size = 10