8587 test erros on pytorch release 2508 on series 50

monai/apps/auto3dseg/bundle_gen.py

@@ -264,8 +264,7 @@ def _run_cmd(self, cmd: str, devices_info: str = "") -> subprocess.CompletedProc
                 look_up_option(self.device_setting["MN_START_METHOD"], ["bcprun"])
             except ValueError as err:
                 raise NotImplementedError(
-                    f"{self.device_setting['MN_START_METHOD']} is not supported yet."
-                    "Try modify BundleAlgo._run_cmd for your cluster."
+                    f"{self.device_setting['MN_START_METHOD']} is not supported yet.Try modify BundleAlgo._run_cmd for your cluster."
                 ) from err
 
             return _run_cmd_bcprun(cmd, n=self.device_setting["NUM_NODES"], p=self.device_setting["n_devices"])

monai/apps/detection/networks/retinanet_detector.py

@@ -342,8 +342,7 @@ def set_regular_matcher(
         """
         if fg_iou_thresh < bg_iou_thresh:
             raise ValueError(
-                "Require fg_iou_thresh >= bg_iou_thresh. "
-                f"Got fg_iou_thresh={fg_iou_thresh}, bg_iou_thresh={bg_iou_thresh}."
+                f"Require fg_iou_thresh >= bg_iou_thresh. Got fg_iou_thresh={fg_iou_thresh}, bg_iou_thresh={bg_iou_thresh}."
             )
         self.proposal_matcher = Matcher(
             fg_iou_thresh, bg_iou_thresh, allow_low_quality_matches=allow_low_quality_matches
@@ -519,7 +518,7 @@ def forward(
         else:
             if self.inferer is None:
                 raise ValueError(
-                    "`self.inferer` is not defined." "Please refer to function self.set_sliding_window_inferer(*)."
+                    "`self.inferer` is not defined.Please refer to function self.set_sliding_window_inferer(*)."
                 )
             head_outputs = predict_with_inferer(
                 images, self.network, keys=[self.cls_key, self.box_reg_key], inferer=self.inferer

monai/auto3dseg/analyzer.py

@@ -952,8 +952,7 @@ def __call__(self, data: dict) -> dict:
             self.hist_range = nr_channels * self.hist_range
         if len(self.hist_range) != nr_channels:
             raise ValueError(
-                f"There is a mismatch between the number of channels ({nr_channels}) "
-                f"and histogram ranges ({len(self.hist_range)})."
+                f"There is a mismatch between the number of channels ({nr_channels}) and histogram ranges ({len(self.hist_range)})."
             )
 
         # perform calculation

monai/data/wsi_reader.py

@@ -416,8 +416,7 @@ def get_data(
             # Check if there are three color channels for RGB
             elif mode in "RGB" and patch.shape[self.channel_dim] != 3:
                 raise ValueError(
-                    f"The image is expected to have three color channels in '{mode}' mode but has "
-                    f"{patch.shape[self.channel_dim]}. "
+                    f"The image is expected to have three color channels in '{mode}' mode but has {patch.shape[self.channel_dim]}. "
                 )
             # Get patch-related metadata
             metadata: dict = self._get_metadata(wsi=each_wsi, patch=patch, location=location, size=size, level=level)

monai/transforms/regularization/array.py

@@ -24,7 +24,6 @@
 
 
 class Mixer(RandomizableTransform):
-
     def __init__(self, batch_size: int, alpha: float = 1.0) -> None:
         """
         Mixer is a base class providing the basic logic for the mixup-class of

monai/transforms/spatial/array.py

@@ -34,6 +34,7 @@
 from monai.transforms.croppad.array import CenterSpatialCrop, ResizeWithPadOrCrop
 from monai.transforms.inverse import InvertibleTransform
 from monai.transforms.spatial.functional import (
+    _compiled_unsupported,
     affine_func,
     convert_box_to_points,
     convert_points_to_box,
@@ -2104,14 +2105,15 @@ def __call__(
         _align_corners = self.align_corners if align_corners is None else align_corners
         img_t, *_ = convert_data_type(img, torch.Tensor, dtype=_dtype, device=_device)
         sr = min(len(img_t.peek_pending_shape() if isinstance(img_t, MetaTensor) else img_t.shape[1:]), 3)
+        _use_compiled = USE_COMPILED and not _compiled_unsupported(img_t.device)
         backend, _interp_mode, _padding_mode, _ = resolves_modes(
             self.mode if mode is None else mode,
             self.padding_mode if padding_mode is None else padding_mode,
             backend=None,
-            use_compiled=USE_COMPILED,
+            use_compiled=_use_compiled,
         )
 
-        if USE_COMPILED or backend == TransformBackends.NUMPY:
+        if _use_compiled or backend == TransformBackends.NUMPY:
             grid_t, *_ = convert_to_dst_type(grid[:sr], img_t, dtype=grid.dtype, wrap_sequence=True)
             if isinstance(grid, torch.Tensor) and grid_t.data_ptr() == grid.data_ptr():
                 grid_t = grid_t.clone(memory_format=torch.contiguous_format)
@@ -2122,7 +2124,7 @@ def __call__(
                     grid_t[i] = ((_dim - 1) / _dim) * grid_t[i] + t if _align_corners else grid_t[i] + t
                 elif _align_corners:
                     grid_t[i] = ((_dim - 1) / _dim) * (grid_t[i] + 0.5)
-            if USE_COMPILED and backend == TransformBackends.TORCH:  # compiled is using torch backend param name
+            if _use_compiled and backend == TransformBackends.TORCH:  # compiled is using torch backend param name
                 grid_t = moveaxis(grid_t, 0, -1)  # type: ignore
                 out = grid_pull(
                     img_t.unsqueeze(0),

monai/transforms/spatial/functional.py

@@ -54,6 +54,27 @@
 __all__ = ["spatial_resample", "orientation", "flip", "resize", "rotate", "zoom", "rotate90", "affine_func"]
 
 
+def _compiled_unsupported(device: torch.device) -> bool:
+    """
+    Return True if ``monai._C`` (the compiled C extension providing ``grid_pull``) is not
+    compiled with support for the given CUDA device's compute capability.
+
+    ``monai._C`` is built at install time against a fixed set of CUDA architectures.
+    NVIDIA Blackwell GPUs (sm_120, compute capability 12.x) and newer were not included in
+    the default ``TORCH_CUDA_ARCH_LIST`` when the MONAI slim image was originally built,
+    so executing ``grid_pull`` on those devices produces incorrect results.  Falling back to
+    the PyTorch-native ``affine_grid`` + ``grid_sample`` path (``USE_COMPILED=False``) gives
+    correct output on all architectures.
+
+    The threshold (``major >= 12``) matches the first architecture family (Blackwell, sm_120)
+    that shipped after the highest sm supported in the current default build list (sm_90,
+    Hopper).  Adjust this constant when ``monai._C`` is rebuilt with sm_120+ support.
+    """
+    if device.type != "cuda":
+        return False
+    return torch.cuda.get_device_properties(device).major >= 12
+
+
 def _maybe_new_metatensor(img, dtype=None, device=None):
     """create a metatensor with fresh metadata if track_meta is True otherwise convert img into a torch tensor"""
     return convert_to_tensor(
@@ -158,7 +179,8 @@ def spatial_resample(
         xform_shape = [-1] + in_sp_size
         img = img.reshape(xform_shape)
     img = img.to(dtype_pt)
-    if isinstance(mode, int) or USE_COMPILED:
+    _use_compiled = USE_COMPILED and not _compiled_unsupported(img.device)
+    if isinstance(mode, int) or _use_compiled:
         dst_xform = create_translate(spatial_rank, [float(d - 1) / 2 for d in spatial_size])
         xform = xform @ convert_to_dst_type(dst_xform, xform)[0]
         affine_xform = monai.transforms.Affine(