[Fix] Support BoolTensor and LongTensor on Ascend NPU.

open-mmlab · Mar 24, 2023 · 67d513b · 67d513b
1 parent cbb6714
commit 67d513b
Showing 1 changed file with 15 additions and 10 deletions.
diff --git a/mmengine/structures/instance_data.py b/mmengine/structures/instance_data.py
@@ -1,16 +1,23 @@
 # Copyright (c) OpenMMLab. All rights reserved.
 import itertools
 from collections.abc import Sized
-from typing import List, Union
+from typing import Any, List, Union
 
 import numpy as np
 import torch
 
+from mmengine.device import is_npu_available
 from .base_data_element import BaseDataElement
 
-IndexType = Union[str, slice, int, list, torch.LongTensor,
-                  torch.cuda.LongTensor, torch.BoolTensor,
-                  torch.cuda.BoolTensor, np.ndarray]
+BoolTypeTensor: Union[Any] = Union[torch.BoolTensor, torch.cuda.BoolTensor,
+                                   torch.npu.BoolTensor] if \
+    is_npu_available() else Union[torch.BoolTensor, torch.cuda.BoolTensor]
+LongTypeTensor: Union[Any] = Union[torch.LongTensor, torch.cuda.LongTensor,
+                                   torch.npu.LongTensor] if \
+    is_npu_available() else Union[torch.LongTensor, torch.cuda.LongTensor]
+
+IndexType = Union[str, slice, int, list, LongTypeTensor, BoolTypeTensor,
+                  np.ndarray]
 
 
 # Modified from
@@ -165,9 +172,8 @@ def __getitem__(self, item: IndexType) -> 'InstanceData':
             # More details in https://github.com/numpy/numpy/issues/9464
             item = item.astype(np.int64) if item.dtype == np.int32 else item
             item = torch.from_numpy(item)
-        assert isinstance(
-            item, (str, slice, int, torch.LongTensor, torch.cuda.LongTensor,
-                   torch.BoolTensor, torch.cuda.BoolTensor))
+        assert isinstance(item, (str, slice, int, LongTypeTensor.__args__,
+                                 BoolTypeTensor.__args__))
 
         if isinstance(item, str):
             return getattr(self, item)
@@ -183,7 +189,7 @@ def __getitem__(self, item: IndexType) -> 'InstanceData':
         if isinstance(item, torch.Tensor):
             assert item.dim() == 1, 'Only support to get the' \
                                     ' values along the first dimension.'
-            if isinstance(item, (torch.BoolTensor, torch.cuda.BoolTensor)):
+            if isinstance(item, BoolTypeTensor.__args__):
                 assert len(item) == len(self), 'The shape of the ' \
                                                'input(BoolTensor) ' \
                                                f'{len(item)} ' \
@@ -202,8 +208,7 @@ def __getitem__(self, item: IndexType) -> 'InstanceData':
                         v, (str, list, tuple)) or (hasattr(v, '__getitem__')
                                                    and hasattr(v, 'cat')):
                     # convert to indexes from BoolTensor
-                    if isinstance(item,
-                                  (torch.BoolTensor, torch.cuda.BoolTensor)):
+                    if isinstance(item, BoolTypeTensor.__args__):
                         indexes = torch.nonzero(item).view(
                             -1).cpu().numpy().tolist()
                     else: