Add pyupgrade pre-commit hook (#1937)

* add pyupgrade

* add options for pyupgrade

* minor refinement

tests/test_ops/test_box_iou_rotated.py

@@ -4,7 +4,7 @@ import pytest
 import torch
 
 
-class TestBoxIoURotated(object):
+class TestBoxIoURotated:
 
     def test_box_iou_rotated_cpu(self):
         from mmcv.ops import box_iou_rotated

tests/test_ops/test_carafe.py

@@ -3,7 +3,7 @@ import torch
 from torch.autograd import gradcheck
 
 
-class TestCarafe(object):
+class TestCarafe:
 
     def test_carafe_naive_gradcheck(self):
         if not torch.cuda.is_available():

tests/test_ops/test_cc_attention.py

@@ -15,7 +15,7 @@ class Loss(nn.Module):
         return torch.mean(input - target)
 
 
-class TestCrissCrossAttention(object):
+class TestCrissCrossAttention:
 
     def test_cc_attention(self):
         device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')

tests/test_ops/test_deform_conv.py

@@ -35,7 +35,7 @@ gt_offset_bias_grad = [1.44, -0.72, 0., 0., -0.10, -0.08, -0.54, -0.54],
 gt_deform_weight_grad = [[[[3.62, 0.], [0.40, 0.18]]]]
 
 
-class TestDeformconv(object):
+class TestDeformconv:
 
     def _test_deformconv(self,
                          dtype=torch.float,

tests/test_ops/test_deform_roi_pool.py

@@ -35,7 +35,7 @@ outputs = [([[[[1, 1.25], [1.5, 1.75]]]], [[[[3.0625, 0.4375],
                                 0.00390625]]]])]
 
 
-class TestDeformRoIPool(object):
+class TestDeformRoIPool:
 
     def test_deform_roi_pool_gradcheck(self):
         if not torch.cuda.is_available():

tests/test_ops/test_focal_loss.py

@@ -37,7 +37,7 @@ sigmoid_outputs = [(0.13562961, [[-0.00657264, 0.11185755],
                                  [-0.02462499, 0.08277918, 0.18050370]])]
 
 
-class Testfocalloss(object):
+class Testfocalloss:
 
     def _test_softmax(self, dtype=torch.float):
         if not torch.cuda.is_available():

tests/test_ops/test_fused_bias_leakyrelu.py

@@ -10,7 +10,7 @@ except ImportError:
     _USING_PARROTS = False
 
 
-class TestFusedBiasLeakyReLU(object):
+class TestFusedBiasLeakyReLU:
 
     @classmethod
     def setup_class(cls):

tests/test_ops/test_info.py

@@ -2,7 +2,7 @@
 import torch
 
 
-class TestInfo(object):
+class TestInfo:
 
     def test_info(self):
         if not torch.cuda.is_available():

tests/test_ops/test_masked_conv2d.py

@@ -2,7 +2,7 @@
 import torch
 
 
-class TestMaskedConv2d(object):
+class TestMaskedConv2d:
 
     def test_masked_conv2d(self):
         if not torch.cuda.is_available():

tests/test_ops/test_modulated_deform_conv.py

@@ -37,7 +37,7 @@ dcn_offset_b_grad = [
 ]
 
 
-class TestMdconv(object):
+class TestMdconv:
 
     def _test_mdconv(self, dtype=torch.float, device='cuda'):
         if not torch.cuda.is_available() and device == 'cuda':

tests/test_ops/test_ms_deformable_attn.py

@@ -55,7 +55,7 @@ def test_forward_multi_scale_deformable_attn_pytorch():
     N, M, D = 1, 2, 2
     Lq, L, P = 2, 2, 2
     shapes = torch.as_tensor([(6, 4), (3, 2)], dtype=torch.long)
-    S = sum([(H * W).item() for H, W in shapes])
+    S = sum((H * W).item() for H, W in shapes)
 
     torch.manual_seed(3)
     value = torch.rand(N, S, M, D) * 0.01
@@ -78,7 +78,7 @@ def test_forward_equal_with_pytorch_double():
     shapes = torch.as_tensor([(6, 4), (3, 2)], dtype=torch.long).cuda()
     level_start_index = torch.cat((shapes.new_zeros(
         (1, )), shapes.prod(1).cumsum(0)[:-1]))
-    S = sum([(H * W).item() for H, W in shapes])
+    S = sum((H * W).item() for H, W in shapes)
 
     torch.manual_seed(3)
     value = torch.rand(N, S, M, D).cuda() * 0.01
@@ -111,7 +111,7 @@ def test_forward_equal_with_pytorch_float():
     shapes = torch.as_tensor([(6, 4), (3, 2)], dtype=torch.long).cuda()
     level_start_index = torch.cat((shapes.new_zeros(
         (1, )), shapes.prod(1).cumsum(0)[:-1]))
-    S = sum([(H * W).item() for H, W in shapes])
+    S = sum((H * W).item() for H, W in shapes)
 
     torch.manual_seed(3)
     value = torch.rand(N, S, M, D).cuda() * 0.01
@@ -155,7 +155,7 @@ def test_gradient_numerical(channels,
     shapes = torch.as_tensor([(3, 2), (2, 1)], dtype=torch.long).cuda()
     level_start_index = torch.cat((shapes.new_zeros(
         (1, )), shapes.prod(1).cumsum(0)[:-1]))
-    S = sum([(H * W).item() for H, W in shapes])
+    S = sum((H * W).item() for H, W in shapes)
 
     value = torch.rand(N, S, M, channels).cuda() * 0.01
     sampling_locations = torch.rand(N, Lq, M, L, P, 2).cuda()

tests/test_ops/test_nms.py

@@ -6,7 +6,7 @@ import torch
 from mmcv.utils import IS_CUDA_AVAILABLE, IS_MLU_AVAILABLE
 
 
-class Testnms(object):
+class Testnms:
 
     @pytest.mark.parametrize('device', [
         pytest.param(
@@ -129,8 +129,7 @@ class Testnms(object):
         scores = tensor_dets[:, 4]
         nms_keep_inds = nms(boxes.contiguous(), scores.contiguous(),
                             iou_thr)[1]
-        assert set([g[0].item()
-                    for g in np_groups]) == set(nms_keep_inds.tolist())
+        assert {g[0].item() for g in np_groups} == set(nms_keep_inds.tolist())
 
         # non empty tensor input
         tensor_dets = torch.from_numpy(np_dets)

tests/test_ops/test_onnx.py

@@ -33,7 +33,7 @@ def run_before_and_after_test():
 class WrapFunction(nn.Module):
 
     def __init__(self, wrapped_function):
-        super(WrapFunction, self).__init__()
+        super().__init__()
         self.wrapped_function = wrapped_function
 
     def forward(self, *args, **kwargs):
@@ -662,7 +662,7 @@ def test_cummax_cummin(key, opset=11):
     input_list = [
         # arbitrary shape, e.g. 1-D, 2-D, 3-D, ...
         torch.rand((2, 3, 4, 1, 5)),
-        torch.rand((1)),
+        torch.rand(1),
         torch.rand((2, 0, 1)),  # tensor.numel() is 0
         torch.FloatTensor(),  # empty tensor
     ]

tests/test_ops/test_psa_mask.py

@@ -15,7 +15,7 @@ class Loss(nn.Module):
         return torch.mean(input - target)
 
 
-class TestPSAMask(object):
+class TestPSAMask:
 
     def test_psa_mask_collect(self):
         if not torch.cuda.is_available():

tests/test_ops/test_roi_pool.py

@@ -29,7 +29,7 @@ outputs = [([[[[1., 2.], [3., 4.]]]], [[[[1., 1.], [1., 1.]]]]),
                                                               1.]]]])]
 
 
-class TestRoiPool(object):
+class TestRoiPool:
 
     def test_roipool_gradcheck(self):
         if not torch.cuda.is_available():

tests/test_ops/test_syncbn.py

@@ -14,7 +14,7 @@ else:
     import re
 
 
-class TestSyncBN(object):
+class TestSyncBN:
 
     def dist_init(self):
         rank = int(os.environ['SLURM_PROCID'])

tests/test_ops/test_tensorrt.py

@@ -30,7 +30,7 @@ if not is_tensorrt_plugin_loaded():
 class WrapFunction(nn.Module):
 
     def __init__(self, wrapped_function):
-        super(WrapFunction, self).__init__()
+        super().__init__()
         self.wrapped_function = wrapped_function
 
     def forward(self, *args, **kwargs):
@@ -576,7 +576,7 @@ def test_cummin_cummax(func: Callable):
     input_list = [
         # arbitrary shape, e.g. 1-D, 2-D, 3-D, ...
         torch.rand((2, 3, 4, 1, 5)).cuda(),
-        torch.rand((1)).cuda()
+        torch.rand(1).cuda()
     ]
 
     input_names = ['input']
@@ -756,7 +756,7 @@ def test_corner_pool(mode):
     class CornerPoolWrapper(CornerPool):
 
         def __init__(self, mode):
-            super(CornerPoolWrapper, self).__init__(mode)
+            super().__init__(mode)
 
         def forward(self, x):
             # no use `torch.cummax`, instead `corner_pool` is used

tests/test_ops/test_upfirdn2d.py

@@ -10,7 +10,7 @@ except ImportError:
     _USING_PARROTS = False
 
 
-class TestUpFirDn2d(object):
+class TestUpFirDn2d:
     """Unit test for UpFirDn2d.
 
     Here, we just test the basic case of upsample version. More gerneal tests

tests/test_ops/test_voxelization.py

@@ -96,8 +96,8 @@ def test_voxelization_nondeterministic():
     coors_all = dynamic_voxelization.forward(points)
     coors_all = coors_all.cpu().detach().numpy().tolist()
 
-    coors_set = set([tuple(c) for c in coors])
-    coors_all_set = set([tuple(c) for c in coors_all])
+    coors_set = {tuple(c) for c in coors}
+    coors_all_set = {tuple(c) for c in coors_all}
 
     assert len(coors_set) == len(coors)
     assert len(coors_set - coors_all_set) == 0
@@ -112,7 +112,7 @@ def test_voxelization_nondeterministic():
 
     for c, ps, n in zip(coors, voxels, num_points_per_voxel):
         ideal_voxel_points_set = coors_points_dict[tuple(c)]
-        voxel_points_set = set([tuple(p) for p in ps[:n]])
+        voxel_points_set = {tuple(p) for p in ps[:n]}
         assert len(voxel_points_set) == n
         if n < max_num_points:
             assert voxel_points_set == ideal_voxel_points_set
@@ -133,7 +133,7 @@ def test_voxelization_nondeterministic():
     voxels, coors, num_points_per_voxel = hard_voxelization.forward(points)
     coors = coors.cpu().detach().numpy().tolist()
 
-    coors_set = set([tuple(c) for c in coors])
-    coors_all_set = set([tuple(c) for c in coors_all])
+    coors_set = {tuple(c) for c in coors}
+    coors_all_set = {tuple(c) for c in coors_all}
 
     assert len(coors_set) == len(coors) == len(coors_all_set)

tests/test_parallel.py

@@ -63,7 +63,7 @@ def test_is_module_wrapper():
 
     # test module wrapper registry
     @MODULE_WRAPPERS.register_module()
-    class ModuleWrapper(object):
+    class ModuleWrapper:
 
         def __init__(self, module):
             self.module = module