Skip to content

reenable bfloat16 masked_select #36859

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 2 commits into from
Closed

reenable bfloat16 masked_select #36859

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
aad1d0a
reenable bfloat16 masked_select
Apr 18, 2020
d3bed8a
add IS_WINDOWS
Apr 18, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 0 additions & 2 deletions aten/src/ATen/native/cuda/IndexKernel.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,8 +97,6 @@ static void index_put_kernel(TensorIterator& iter, IntArrayRef index_size, IntAr
static Tensor & masked_select_out_cuda_impl(Tensor & result, const Tensor & self, const Tensor & mask) {
NoNamesGuard guard;

TORCH_CHECK(self.scalar_type() != ScalarType::BFloat16,
"masked_select: bfloat16 not supported for CUDA implementation");
TORCH_CHECK(mask.scalar_type() == ScalarType::Byte || mask.scalar_type() == ScalarType::Bool,
"masked_select: expected BoolTensor or ByteTensor for mask");
TORCH_CHECK(self.scalar_type() == result.scalar_type(),
Expand Down
12 changes: 9 additions & 3 deletions test/test_torch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11256,7 +11256,6 @@ def test_masked_scatter_bool_tensor(self, device):
dst = dst.masked_scatter(mask, src)
self.assertEqual(dst, torch.tensor([True, True, True], device=device))

@dtypesIfCUDA(*[dtype for dtype in torch.testing.get_all_dtypes() if dtype not in [torch.bfloat16]])
@dtypes(*torch.testing.get_all_dtypes())
def test_masked_select(self, device, dtype):
if device == 'cpu':
Expand Down Expand Up @@ -11292,7 +11291,11 @@ def test_masked_select(self, device, dtype):
return

# Ensure that masks are expanded to match tensor properly
a = torch.rand(100, 100, device=device).mul(100).to(dtype)
if IS_WINDOWS and dtype == torch.bfloat16 and torch.device(device).type == 'cuda':
# TODO .to() for bfloat16 does not work on windows
a = torch.ones(100, 100, device=device, dtype=dtype)
else:
a = torch.rand(100, 100, device=device).mul(100).to(dtype)
mask_first_el_each_row = torch.zeros(100, device=device).bool()
mask_first_el_each_row[0] = True
a_masked = a.masked_select(mask_first_el_each_row)
Expand All @@ -11304,7 +11307,10 @@ def test_masked_select(self, device, dtype):
self.assertEqual(a_masked, a[0, :])

# Ensure that tensor is expanded to match mask properly
a = torch.rand(100, device=device).mul(100).to(maskType)
if IS_WINDOWS and dtype == torch.bfloat16 and torch.device(device).type == 'cuda':
a = torch.ones(100, device=device, dtype=dtype)
else:
a = torch.rand(100, device=device).mul(100).to(maskType)
mask_copy_3_times = torch.tensor([[True], [True], [False], [True]], device=device)
a_masked = a.masked_select(mask_copy_3_times)
self.assertEqual(a_masked, a.unsqueeze(0).expand(3, 100).flatten())
Expand Down