diff --git a/aten/src/ATen/CMakeLists.txt b/aten/src/ATen/CMakeLists.txt
index 40b526ebacc162..f6d296dfe79e45 100644
--- a/aten/src/ATen/CMakeLists.txt
+++ b/aten/src/ATen/CMakeLists.txt
@@ -247,7 +247,7 @@ IF(USE_CUDA AND NOT USE_ROCM)
 ENDIF()
 
 IF(USE_ROCM)
- ### Link in the ROCm libraries BLAS / RNG .
+ ### Link in the ROCm libraries BLAS / RNG.
  FIND_LIBRARY(ROCBLAS_LIBRARY rocblas HINTS ${ROCBLAS_PATH}/lib)
  FIND_LIBRARY(HIPRAND_LIBRARY hiprand HINTS ${HIPRAND_PATH}/lib)
 
diff --git a/test/common.py b/test/common.py
index 13a41822ac5254..9cd0a3c8a773e5 100644
--- a/test/common.py
+++ b/test/common.py
@@ -107,6 +107,7 @@ def wrapper(*args, **kwargs):
             fn(*args, **kwargs)
     return wrapper
 
+
 def skipIfNoLapack(fn):
     @wraps(fn)
     def wrapper(*args, **kwargs):
diff --git a/test/common_nn.py b/test/common_nn.py
index c8f34af22a6413..f159fe659672c4 100644
--- a/test/common_nn.py
+++ b/test/common_nn.py
@@ -41,7 +41,7 @@ def get_weight(m):
         constructor_args=(10, 8),
         input_size=(4, 10),
         reference_fn=lambda i, p: torch.mm(i, p[0].t()) + p[1].view(1, -1).expand(4, 8),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Linear',
@@ -103,20 +103,20 @@ def get_weight(m):
         constructor_args=(1,),
         input_size=(10, 20),
         reference_fn=lambda i, _: torch.exp(i).div(torch.exp(i).sum(1, True).expand(10, 20)),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Softmax2d',
         input_size=(1, 3, 10, 20),
         reference_fn=lambda i, _: torch.exp(i).div(torch.exp(i).sum(1, False)),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LogSoftmax',
         constructor_args=(1,),
         input_size=(10, 20),
         reference_fn=lambda i, _: torch.exp(i).div_(torch.exp(i).sum(1, True).expand(10, 20)).log_(),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LogSoftmax',
@@ -124,7 +124,7 @@ def get_weight(m):
         input_size=(1, 3, 10, 20),
         reference_fn=lambda i, _: torch.exp(i).div_(torch.exp(i).sum(1, False)).log_(),
         desc='multiparam',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='ELU',
@@ -204,14 +204,13 @@ def get_weight(m):
         input_size=(2, 3, 4),
         desc='1d_multiparam',
         reference_fn=lambda i, p: torch.clamp(i, min=0) + torch.clamp(i, max=0) * p[0][0],
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PReLU',
         input_size=(2, 3, 4, 5),
         desc='2d',
         reference_fn=lambda i, p: torch.clamp(i, min=0) + torch.clamp(i, max=0) * p[0][0],
-        #test_cuda = (not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PReLU',
@@ -219,14 +218,13 @@ def get_weight(m):
         input_size=(2, 3, 4, 5),
         desc='2d_multiparam',
         reference_fn=lambda i, p: torch.clamp(i, min=0) + torch.clamp(i, max=0) * p[0][0],
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PReLU',
         input_size=(2, 3, 4, 5, 6),
         reference_fn=lambda i, p: torch.clamp(i, min=0) + torch.clamp(i, max=0) * p[0][0],
         desc='3d',
-        #test_cuda = (not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PReLU',
@@ -234,31 +232,31 @@ def get_weight(m):
         input_size=(2, 3, 4, 5, 6),
         desc='3d_multiparam',
         reference_fn=lambda i, p: torch.clamp(i, min=0) + torch.clamp(i, max=0) * p[0][0],
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Softsign',
         input_size=(3, 2, 5),
         reference_fn=lambda i, _: i.div(1 + torch.abs(i)),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Softmin',
         constructor_args=(1,),
         input_size=(10, 20),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Softmin',
         constructor_args=(1,),
         input_size=(2, 3, 5, 10),
         desc='multidim',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Tanhshrink',
         input_size=(2, 3, 4, 5),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
 ]
 
@@ -575,6 +573,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         reference_fn=lambda i, t, m:
             kldivloss_reference(i, t, get_reduction(m)),
         check_sum_reduction=True,
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='MSELoss',
@@ -591,7 +590,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         reference_fn=lambda i, t, m: -(t * i.log() + (1 - t) * (1 - i).log()).sum() /
             (i.numel() if get_reduction(m) else 1),
         check_gradgrad=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='BCELoss',
@@ -602,7 +601,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
             (i.numel() if get_reduction(m) else 1),
         desc='weights',
         check_gradgrad=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='CrossEntropyLoss',
@@ -623,7 +622,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         reference_fn=lambda i, t, m:
             hingeembeddingloss_reference(i, t, reduction=get_reduction(m)),
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='HingeEmbeddingLoss',
@@ -634,7 +633,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
             hingeembeddingloss_reference(i, t, margin=0.5, reduction=get_reduction(m)),
         desc='margin',
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='MultiLabelMarginLoss',
@@ -661,7 +660,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         target_fn=lambda: torch.rand(5, 10).mul(2).floor(),
         reference_fn=lambda i, t, m: -(t * i.sigmoid().log() + (1 - t) * (-i).sigmoid().log()).sum() / i.numel(),
         check_gradgrad=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='MultiMarginLoss',
@@ -740,7 +739,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         reference_fn=lambda i, t, m:
             cosineembeddingloss_reference(i[0], i[1], t, reduction=get_reduction(m)),
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='CosineEmbeddingLoss',
@@ -751,7 +750,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
             cosineembeddingloss_reference(i[0], i[1], t, margin=0.7, reduction=get_reduction(m)),
         desc='margin',
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='MarginRankingLoss',
@@ -760,7 +759,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
         reference_fn=lambda i, t, m:
             marginrankingloss_reference(i[0], i[1], t, reduction=get_reduction(m)),
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='MarginRankingLoss',
@@ -771,7 +770,7 @@ def ctcloss_reference(log_probs, targets, input_lengths, target_lengths, blank=0
             marginrankingloss_reference(i[0], i[1], t, margin=0.5, reduction=get_reduction(m)),
         desc='margin',
         check_sum_reduction=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
 ]
 
diff --git a/test/test_jit.py b/test/test_jit.py
index 8ff0a1ddee5518..66724e0bccbfd6 100644
--- a/test/test_jit.py
+++ b/test/test_jit.py
@@ -6324,6 +6324,7 @@ class Config:
 
         self.checkTrace(SNLIClassifier(Config()), (premise, hypothesis), inputs_require_grads=False)
 
+    @skipIfRocm
     def test_super_resolution(self):
         import torch.nn.init as init
 
diff --git a/test/test_legacy_nn.py b/test/test_legacy_nn.py
index f37b34cd43603f..b446920c4fec65 100644
--- a/test/test_legacy_nn.py
+++ b/test/test_legacy_nn.py
@@ -67,26 +67,26 @@ def _do_test(self, test_case, module, input):
                   input_size=(3, 5, 4),
                   reference_fn=lambda i, _: i + 3.5,
                   check_inplace=True,
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.BatchNormalization,
                   constructor_args=(10,),
                   input_size=(4, 10),
                   desc='affine',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.BatchNormalization,
                   constructor_args=(10, 1e-3, 0.3, False),
                   input_size=(4, 10),
                   desc='not_affine',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialBatchNormalization,
                   constructor_args=(3,),
                   input_size=(2, 3, 6, 6),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialBatchNormalization,
                   constructor_args=(3, 1e-3, 0.8),
                   input_size=(2, 3, 6, 6),
                   desc='momentum',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialBatchNormalization,
                   constructor_args=(3, 1e-3, 0.8, False),
                   input_size=(2, 3, 6, 6),
@@ -94,12 +94,12 @@ def _do_test(self, test_case, module, input):
     OldModuleTest(nn.VolumetricBatchNormalization,
                   constructor_args=(3,),
                   input_size=(2, 3, 4, 4, 4),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.VolumetricBatchNormalization,
                   constructor_args=(3, 1e-3, 0.7),
                   input_size=(2, 3, 4, 4, 4),
                   desc='momentum',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.VolumetricBatchNormalization,
                   constructor_args=(3, 1e-3, 0.7, False),
                   input_size=(2, 3, 4, 4, 4),
@@ -108,25 +108,25 @@ def _do_test(self, test_case, module, input):
                   constructor_args=(5, 6),
                   input_size=(10, 5, 6),
                   desc='3D',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CMul,
                   constructor_args=(50, 4),
                   input_size=(1, 50, 4),
                   desc='3D_single_example',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CMul,
                   constructor_args=(1, 5),
                   input_fn=lambda: torch.randn(10, 3, 5)[:, 1],
                   desc='3D_noncontiguous',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Exp,
                   input_size=(2, 3, 4),
                   reference_fn=lambda i, _: i.exp(),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Log,
                   input_fn=lambda: torch.rand(2, 3, 2) + 0.1,
                   reference_fn=lambda i, _: i.log(),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Clamp,
                   constructor_args=(-2., 5.),
                   input_fn=lambda: torch.randn(3, 2, 50) * 6,
@@ -134,40 +134,40 @@ def _do_test(self, test_case, module, input):
     OldModuleTest(nn.Abs,
                   input_size=(3, 20, 5),
                   reference_fn=lambda i, _: i.abs(),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Bilinear,
                   constructor_args=(2, 3, 10),
                   input_size=[(4, 2), (4, 3)],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Bilinear,
                   constructor_args=(5, 4, 2),
                   input_size=[(2, 5), (2, 4)],
                   desc='small_output',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Euclidean,
                   constructor_args=(5, 7),
                   input_size=(10, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.WeightedEuclidean,
                   constructor_args=(5, 7),
                   input_size=(10, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Cosine,
                   constructor_args=(5, 7),
                   input_size=(10, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CAddTable,
                   input_size=[(5, 7), (5, 7)],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CSubTable,
                   input_size=[(5, 7), (5, 7)],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CDivTable,
                   input_fn=lambda: [torch.randn(1, 7), torch.rand(1, 7) + 0.1],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.CMulTable,
                   input_size=[(5, 7), (5, 7)],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Square,
                   input_size=(10, 2, 4),
                   reference_fn=lambda i, _: i.mul(i)),
@@ -215,22 +215,22 @@ def _do_test(self, test_case, module, input):
                   constructor_args=(1,),
                   input_size=(2, 4, 5),
                   reference_fn=lambda i, _: i.sum(1, keepdim=False),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Sum,
                   constructor_args=(1, True),
                   input_size=(2, 4, 5),
                   reference_fn=lambda i, _: i.sum(1, keepdim=False).div(i.size(1)),
                   desc='sizeAverage',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Mean,
                   constructor_args=(1,),
                   input_size=(2, 4, 5),
                   reference_fn=lambda i, _: torch.mean(i, 1, keepdim=False),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(lambda: nn.Sequential().add(nn.GradientReversal()).add(nn.GradientReversal()),
                   input_size=(4, 3, 2, 2),
                   fullname='GradientReversal',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Identity,
                   input_size=(4, 3, 2, 4),
                   reference_fn=lambda i, _: i),
@@ -238,13 +238,13 @@ def _do_test(self, test_case, module, input):
                   input_size=[(10, 4), (10, 4)],
                   reference_fn=lambda i, _: torch.Tensor(list(
                       a.dot(b) for a, b in zip(i[0], i[1]))),
-                  test_cuda = (not TEST_WITH_ROCM)
+                  test_cuda=(not TEST_WITH_ROCM)
                   ),
     OldModuleTest(nn.CosineDistance,
                   input_size=[(10, 4), (10, 4)],
                   reference_fn=lambda i, _: torch.Tensor(list(
                       a.dot(b) / (a.norm(2) * b.norm(2)) for a, b in zip(i[0], i[1]))),
-                  test_cuda = (not TEST_WITH_ROCM)
+                  test_cuda=(not TEST_WITH_ROCM)
                   ),
     OldModuleTest(nn.JoinTable,
                   constructor_args=(0,),
@@ -285,22 +285,22 @@ def _do_test(self, test_case, module, input):
                   desc='with_dimension'),
     OldModuleTest(nn.MixtureTable,
                   input_size=[(5, 3), (5, 3, 6)],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.LookupTable,
                   constructor_args=(4, 3),
                   input_fn=lambda: torch.randperm(2).repeat(1, 2),
                   jacobian_input=False,
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Mul,
                   input_size=(2, 3, 4, 2),
                   reference_fn=lambda i, p: i * p[0][0],
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.MulConstant,
                   constructor_args=(4,),
                   input_size=(2, 3, 4, 2),
                   reference_fn=lambda i, _: i * 4,
                   check_inplace=True,
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Narrow,
                   constructor_args=(0, 0),
                   input_size=(2, 3, 4, 2),
@@ -324,7 +324,7 @@ def _do_test(self, test_case, module, input):
                   constructor_args=(2, 1),
                   input_size=(10, 3, 4, 5),
                   reference_fn=lambda i, _: i.view(10, 1, 3, 4, 5).expand(10, 2, 3, 4, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Padding,
                   constructor_args=(0, 2, -10),
                   input_size=(2, 3, 4, 5)),
@@ -339,20 +339,20 @@ def _do_test(self, test_case, module, input):
     OldModuleTest(nn.PartialLinear,
                   constructor_args=(5, 6),
                   input_size=(4, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(lambda: nn.PartialLinear(5, 6).setPartition(torch.Tensor((2, 4))),
                   input_size=(4, 5),
                   fullname='PartialLinear_setPartition',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Power,
                   constructor_args=(2,),
                   input_size=(2, 3, 4, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Power,
                   constructor_args=(1.5,),
                   input_fn=lambda: torch.rand(3, 4, 5),
                   desc='fractional',
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.Reshape,
                   constructor_args=(4, 5),
                   input_size=(3, 4 * 5),
@@ -413,11 +413,11 @@ def _do_test(self, test_case, module, input):
     OldModuleTest(nn.SpatialDivisiveNormalization,
                   constructor_args=(3,),
                   input_size=(2, 3, 8, 8),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialContrastiveNormalization,
                   constructor_args=(3,),
                   input_size=(2, 3, 8, 8),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialDilatedConvolution,
                   constructor_args=(3, 2, 3, 3, 2, 2, 1, 1, 2, 2),
                   input_size=(2, 3, 8, 8)),
@@ -476,14 +476,14 @@ def _do_test(self, test_case, module, input):
     OldModuleTest(nn.SpatialLPPooling,
                   constructor_args=(3, 2, 2, 2, 2, 2),
                   input_size=(1, 3, 7, 7),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialSubSampling,
                   constructor_args=(3, 3, 3, 2, 2),
                   input_size=(1, 3, 7, 7)),
     OldModuleTest(nn.SpatialSubtractiveNormalization,
                   constructor_args=(3,),
                   input_size=(1, 3, 7, 7),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     OldModuleTest(nn.SpatialSubtractiveNormalization,
                   constructor_args=(3, torch.rand(3)),
                   input_size=(1, 3, 7, 7),
@@ -563,7 +563,7 @@ def _do_test(self, test_case, module, input):
                   constructor_args_fn=lambda: (torch.rand(3, 4, 5),),
                   input_size=(2, 3, 4, 5),
                   target_size=(2, 3, 4, 5),
-                  test_cuda = (not TEST_WITH_ROCM)),
+                  test_cuda=(not TEST_WITH_ROCM)),
     CriterionTest(nn.MarginCriterion,
                   input_size=(5, 10),
                   target_fn=lambda: torch.randn(5, 10).sign()),
@@ -587,7 +587,7 @@ def _do_test(self, test_case, module, input):
                       # Eh, we need to use p as a default, so it's passed by value
                       reference_fn=lambda i, _, p=p: i.div(i.norm(p, 1, True).expand_as(i)),
                       desc=str(p),
-                      test_cuda = (not TEST_WITH_ROCM)),
+                      test_cuda=(not TEST_WITH_ROCM)),
     )
 for p in range(1, 4 + 1):
     tests.append(
@@ -595,7 +595,7 @@ def _do_test(self, test_case, module, input):
                       constructor_args=(p,),
                       input_size=[(4, 10), (4, 10)],
                       desc=str(p),
-                      test_cuda = (not TEST_WITH_ROCM))
+                      test_cuda=(not TEST_WITH_ROCM))
     )
 
 
@@ -658,7 +658,8 @@ def add_test(test):
     }
     for test in tests:
         name = test.get_name()
-        if ((name == "test_Max" or name == "test_Min" or name == "test_Max_with_dimension" or name == "test_Min_with_dimension") and TEST_WITH_ROCM):
+        if ((name == "test_Max" or name == "test_Min" or name == "test_Max_with_dimension" or
+           name == "test_Min_with_dimension") and TEST_WITH_ROCM):
             continue
         add_test(test)
     for test_params in module_tests:
diff --git a/test/test_nn.py b/test/test_nn.py
index e5ab6303cdb3c8..0a8d6fee034bab 100644
--- a/test/test_nn.py
+++ b/test/test_nn.py
@@ -50,7 +50,7 @@
                    torch.double]
 
 if not TEST_WITH_ROCM:
-  ALL_TENSORTYPES.append(torch.half)
+    ALL_TENSORTYPES.append(torch.half)
 
 NO_HALF_TENSORTYPES = [torch.float,
                        torch.double]
@@ -361,22 +361,22 @@ def _do_test(self, test_case, module, input):
                         test_case.assertEqual(p.get_device(), 1)
 
                 if not self.skip_double:
-                  # test double()
-                  input = input.double().cuda()
-                  module.double().cuda()
-                  module(input)
-                  for p in module.parameters():
-                      test_case.assertIsInstance(p, torch.cuda.DoubleTensor)
-                      test_case.assertEqual(p.get_device(), 0)
+                    # test double()
+                    input = input.double().cuda()
+                    module.double().cuda()
+                    module(input)
+                    for p in module.parameters():
+                        test_case.assertIsInstance(p, torch.cuda.DoubleTensor)
+                        test_case.assertEqual(p.get_device(), 0)
 
                 if not TEST_WITH_ROCM:
-                  # test half()
-                  input = input.half().cuda()
-                  module.half().cuda()
-                  module(input)
-                  for p in module.parameters():
-                      test_case.assertIsInstance(p, torch.cuda.HalfTensor)
-                      test_case.assertEqual(p.get_device(), 0)
+                    # test half()
+                    input = input.half().cuda()
+                    module.half().cuda()
+                    module(input)
+                    for p in module.parameters():
+                        test_case.assertIsInstance(p, torch.cuda.HalfTensor)
+                        test_case.assertEqual(p.get_device(), 0)
 
     def _get_target(self):
         return self._get_arg('target', False)
@@ -5030,6 +5030,7 @@ def test_grid_sample_error_checking(self):
         if TEST_CUDA:
             with self.assertRaisesRegex(RuntimeError, "expected input and grid to be on same device"):
                 F.grid_sample(input.cuda(), grid)
+
     @skipIfRocm
     def test_grid_sample(self):
         def test(N, C, H, W, mode, padding_mode):
@@ -5177,6 +5178,7 @@ def test_shape(N, C, IH, IW, H, W, mode, padding_mode):
                 if TEST_CUDNN:
                     with cudnn.flags(enabled=False):
                         test(N, C, H, W, mode, padding_mode)
+
     @skipIfRocm
     def test_grid_sample_3d(self):
         def test(N, C, D, H, W, mode, padding_mode):
@@ -5290,6 +5292,7 @@ def test_shape(N, C, ID, IH, IW, D, H, W, mode, padding_mode):
                     (input, grid)))
 
                 test(N, C, D, H, W, mode, padding_mode)
+
     @skipIfRocm
     def test_affine_grid(self):
         # test known input on CPU
@@ -6525,9 +6528,10 @@ def add(test_name, fn):
             test=test, kwargs=kwargs: test.test_cuda(self, dtype=torch.float, **kwargs))
         add(cuda_test_name + '_double', lambda self,
             test=test, kwargs=kwargs: test.test_cuda(self, dtype=torch.double, **kwargs))
+
         @skipIfRocm
         def test_half(self, test=test, kwargs=kwargs):
-          test.test_cuda(self, dtype=torch.half, **kwargs)
+            test.test_cuda(self, dtype=torch.half, **kwargs)
         if getattr(test, 'check_half', True):
             add(cuda_test_name + '_half', test_half)
     else:
@@ -6571,7 +6575,7 @@ def forward(self, *args):
             loss_reference_fns['NLLLossNd'](i, t, reduction=get_reduction(m)),
         check_sum_reduction=True,
         desc='2d',
-        test_cuda = (not TEST_WITH_ROCM),
+        test_cuda=(not TEST_WITH_ROCM),
     ),
     dict(
         module_name='NLLLoss',
@@ -6581,7 +6585,7 @@ def forward(self, *args):
         reference_fn=lambda i, t, m:
             loss_reference_fns['NLLLossNd'](i, t, weight=get_weight(m)),
         desc='2d_weights',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='NLLLoss',
@@ -6591,7 +6595,7 @@ def forward(self, *args):
         reference_fn=lambda i, t, m:
             loss_reference_fns['NLLLossNd'](i, t, ignore_index=1),
         desc='2d_ignore_index',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='NLLLoss',
@@ -6601,7 +6605,7 @@ def forward(self, *args):
             loss_reference_fns['NLLLossNd'](i, t, reduction=get_reduction(m)),
         check_sum_reduction=True,
         desc='higher_dim',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='NLLLoss',
@@ -6611,14 +6615,14 @@ def forward(self, *args):
             loss_reference_fns['NLLLossNd'](i, t, reduction=get_reduction(m)),
         check_sum_reduction=True,
         desc='dim_is_3',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PoissonNLLLoss',
         input_size=(2, 3, 4, 5),
         target_fn=lambda: torch.randn(2, 3, 4, 5).floor_().abs_(),
         desc='no_full_loss',  # without sterling approx
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='PoissonNLLLoss',
@@ -6626,7 +6630,7 @@ def forward(self, *args):
         input_fn=lambda: torch.randn(2, 3, 4, 5).abs_().add_(0.001),
         target_fn=lambda: torch.randn(2, 3, 4, 5).floor_().abs_(),
         desc='full_loss',  # with sterling approx
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='L1Loss',
@@ -6699,7 +6703,7 @@ def forward(self, *args):
         desc='weights',
         check_sum_reduction=True,
         check_gradgrad=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='CTCLoss',
@@ -6768,7 +6772,7 @@ def poissonnllloss_no_reduce_test():
         constructor=wrap_functional(
             lambda i: F.poisson_nll_loss(i, t.type_as(i), reduction='none')),
         input_fn=lambda: torch.rand(10, 10),
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def bceloss_no_reduce_test():
@@ -6780,7 +6784,7 @@ def bceloss_no_reduce_test():
         input_fn=lambda: torch.rand(15, 10).clamp_(2.8e-2, 1 - 2.8e-2),
         reference_fn=lambda i, m: -(t * i.log() + (1 - t) * (1 - i).log()),
         check_gradgrad=False,
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def bceloss_no_reduce_scalar_test():
@@ -6806,7 +6810,7 @@ def bceloss_weights_no_reduce_test():
         input_fn=lambda: torch.rand(15, 10).clamp_(2.8e-2, 1 - 2.8e-2),
         reference_fn=lambda i, m: -(t * i.log() + (1 - t) * (1 - i).log()) * weights,
         check_gradgrad=False,
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def bceloss_weights_no_reduce_scalar_test():
@@ -6860,7 +6864,7 @@ def kldivloss_with_target_no_reduce_test():
         input_fn=lambda: torch.rand(10, 10),
         reference_fn=lambda t, _:
             loss_reference_fns['KLDivLoss'](i.type_as(t), t, reduction='none'),
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def kldivloss_no_reduce_test():
@@ -6872,7 +6876,8 @@ def kldivloss_no_reduce_test():
         input_fn=lambda: torch.rand(10, 10).log(),
         reference_fn=lambda i, _:
             loss_reference_fns['KLDivLoss'](i, t.type_as(i), reduction='none'),
-        pickle=False)
+        pickle=False,
+        decorator=skipIfRocm)
 
 
 def kldivloss_no_reduce_scalar_test():
@@ -6895,7 +6900,7 @@ def l1loss_no_reduce_test():
             lambda i: F.l1_loss(i, t.type_as(i), reduction='none')),
         input_fn=lambda: torch.randn(2, 3, 4),
         reference_fn=lambda i, m: (i - t.type_as(i)).abs(),
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def l1loss_no_reduce_scalar_test():
@@ -7107,7 +7112,7 @@ def smoothl1loss_no_reduce_test():
         input_fn=lambda: torch.randn(2, 3, 4),
         reference_fn=lambda i, _:
             loss_reference_fns['SmoothL1Loss'](i, t.type_as(i), reduction='none'),
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def smoothl1loss_no_reduce_scalar_test():
@@ -7199,7 +7204,7 @@ def softmarginloss_no_reduce_test():
         input_fn=lambda: torch.randn(5, 5),
         reference_fn=lambda i, _:
             loss_reference_fns['SoftMarginLoss'](i, t.type_as(i), reduction='none'),
-        pickle=False, test_cuda = (not TEST_WITH_ROCM))
+        pickle=False, test_cuda=(not TEST_WITH_ROCM))
 
 
 def multilabelsoftmarginloss_no_reduce_test():
@@ -7212,7 +7217,7 @@ def multilabelsoftmarginloss_no_reduce_test():
         reference_fn=lambda i, m:
             (-(t * i.sigmoid().log() + (1 - t) * (-i).sigmoid().log())).sum(dim=1) / i.size(1),
         check_gradgrad=False,
-        pickle=False, decorator = skipIfRocm)
+        pickle=False, decorator=skipIfRocm)
 
 
 def multilabelsoftmarginloss_weights_no_reduce_test():
@@ -7228,7 +7233,7 @@ def multilabelsoftmarginloss_weights_no_reduce_test():
             (-(t * i.sigmoid().log() + (1 - t) * (-i).sigmoid().log()) * weights).sum(dim=1) / i.size(1),
         check_sum_reduction=True,
         check_gradgrad=False,
-        pickle=False, decorator = skipIfRocm)
+        pickle=False, decorator=skipIfRocm)
 
 
 def multimarginloss_no_reduce_test():
@@ -7505,7 +7510,7 @@ def multimarginloss_weights_no_reduce_test():
         input_size=(4, 3, 15),
         cudnn=True,
         check_eval=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='InstanceNorm1d',
@@ -7514,7 +7519,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='tracking_stats',
-        decorator = skipIfRocm
+        decorator=skipIfRocm
     ),
     dict(
         module_name='InstanceNorm2d',
@@ -7522,7 +7527,7 @@ def multimarginloss_weights_no_reduce_test():
         input_size=(2, 3, 6, 6),
         cudnn=True,
         check_eval=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='InstanceNorm2d',
@@ -7531,7 +7536,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='tracking_stats',
-        decorator = skipIfRocm 
+        decorator=skipIfRocm
     ),
     dict(
         module_name='InstanceNorm3d',
@@ -7539,7 +7544,7 @@ def multimarginloss_weights_no_reduce_test():
         input_size=(2, 3, 4, 4, 4),
         cudnn=True,
         check_eval=True,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='InstanceNorm3d',
@@ -7548,7 +7553,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='tracking_stats',
-        decorator = skipIfRocm
+        decorator=skipIfRocm
     ),
     dict(
         module_name='LayerNorm',
@@ -7557,7 +7562,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='1d_elementwise_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LayerNorm',
@@ -7566,7 +7571,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='1d_no_elementwise_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LayerNorm',
@@ -7575,7 +7580,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='3d_elementwise_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LayerNorm',
@@ -7584,7 +7589,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='3d_no_elementwise_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7593,7 +7598,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='1d_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7602,7 +7607,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='1d_no_affine_IN',  # this setting is equivalent with InstanceNormi
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7611,7 +7616,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='1d_no_affine_LN',  # this setting is equivalent with LayerNorm
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7620,7 +7625,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='2d_affine',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7629,7 +7634,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='2d_no_affine_IN',  # this setting is equivalent with InstanceNorm
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GroupNorm',
@@ -7638,7 +7643,7 @@ def multimarginloss_weights_no_reduce_test():
         cudnn=True,
         check_eval=True,
         desc='2d_no_affine_LN',  # this setting is equivalent with LayerNorm
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='Conv1d',
@@ -7927,21 +7932,21 @@ def multimarginloss_weights_no_reduce_test():
         constructor_args=(3, ),
         input_size=(1, 5, 7),
         desc='1d',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LocalResponseNorm',
         constructor_args=(2, ),
         input_size=(1, 5, 7, 7),
         desc='2d_uneven_pad',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='LocalResponseNorm',
         constructor_args=(1, 1, 0.5, 2),
         input_size=(1, 5, 7, 7, 7),
         desc='3d_custom_params',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='ReflectionPad1d',
@@ -8131,8 +8136,8 @@ def multimarginloss_weights_no_reduce_test():
         jacobian_input=False,
         check_gradgrad=False,
         desc='mean',
-        test_cuda = (not TEST_WITH_ROCM),
-        decorator = skipIfRocm
+        test_cuda=(not TEST_WITH_ROCM),
+        decorator=skipIfRocm
     ),
     dict(
         module_name='EmbeddingBag',
@@ -8141,8 +8146,8 @@ def multimarginloss_weights_no_reduce_test():
         jacobian_input=False,
         check_gradgrad=False,
         desc='sum',
-        test_cuda = (not TEST_WITH_ROCM),
-        decorator = skipIfRocm
+        test_cuda=(not TEST_WITH_ROCM),
+        decorator=skipIfRocm
     ),
     dict(
         module_name='EmbeddingBag',
@@ -8158,8 +8163,8 @@ def multimarginloss_weights_no_reduce_test():
         input_fn=lambda: torch.randperm(2).repeat(1, 2),
         jacobian_input=False,
         check_gradgrad=False,
-        decorator = skipIfRocm,
-        test_cuda = (not TEST_WITH_ROCM)
+        decorator=skipIfRocm,
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=lambda: nn.Embedding(4, 3, sparse=True),
@@ -8167,7 +8172,7 @@ def multimarginloss_weights_no_reduce_test():
         jacobian_input=False,
         fullname='Embedding_sparse',
         check_gradgrad=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=lambda: nn.FractionalMaxPool2d(
@@ -8471,35 +8476,35 @@ def multimarginloss_weights_no_reduce_test():
     dict(
         module_name='GLU',
         input_size=(5, 6),
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='GLU',
         constructor_args=(1,),
         input_size=(5, 6, 7),
         desc='dim',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.softmax, dim=-1),
         input_size=(2, 128),  # trigger the last-dim algo in CUDA
         fullname='softmax_lastdim',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.softmax, dim=1),
         input_size=(2, 128, 2, 2),  # trigger special case of spatial CUDA algo
         fullname='softmax_spatial_special',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.softmax, dim=1),
         input_size=(2, 2, 4, 4),  # regular spatial algorithm
         fullname='softmax_spatial',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.softmax, dim=0),
@@ -8527,35 +8532,35 @@ def multimarginloss_weights_no_reduce_test():
         input_size=(2, 128),  # trigger the last-dim algo in CUDA
         fullname='log_softmax_lastdim',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.log_softmax, dim=1),
         input_size=(2, 128, 2, 2),  # trigger special case of spatial CUDA algo
         fullname='log_softmax_spatial_special',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.log_softmax, dim=1),
         input_size=(2, 2, 4, 4),  # regular spatial algorithm
         fullname='log_softmax_spatial',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.log_softmax, dim=0),
         input_size=(2, 3, 4, 5),
         fullname='log_softmax_dim0',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.log_softmax, dim=3),
         input_size=(2, 3, 4, 5),
         fullname='log_softmax_dim3',
         pickle=False,
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         constructor=wrap_functional(F.log_softmax, dim=0),
@@ -8647,7 +8652,7 @@ def multimarginloss_weights_no_reduce_test():
         input_size=(),
         reference_fn=lambda i, _: torch.exp(i).div_(torch.exp(i).sum(0, False)).log_(),
         desc='multiparam_scalar',
-        test_cuda = (not TEST_WITH_ROCM)
+        test_cuda=(not TEST_WITH_ROCM)
     ),
     dict(
         module_name='ELU',