Skip to content

error instead of crashing on attempt to subclass typed tensors #20283

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
nairbv wants to merge 3 commits into from
Closed

error instead of crashing on attempt to subclass typed tensors #20283

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
34bce61
error instead of crashing on attempt to subclass typed tensors
nairbv May 8, 2019
2e1a491
fix flake8 errors
nairbv May 8, 2019
5ec0cdd
fix incorrect unit test
nairbv May 8, 2019
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
13 changes: 13 additions & 0 deletions test/test_torch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11324,6 +11324,19 @@ def test_c10_layer_norm(self):
weight), torch.tensor(bias), 1, epsilon, True)
torch.testing.assert_allclose(expected_norm, actual_norm)

def test_subclass_tensors(self):
# raise an error when trying to subclass FloatTensor
with self.assertRaisesRegex(TypeError, "type 'torch.FloatTensor' is not an acceptable base type"):
Copy link
Contributor

@gchanan gchanan May 8, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm curious if this error message is actually stable -- I guess the CI will tell us.

class Foo1(torch.FloatTensor):
pass

# but allow subclassing Tensor:
class Foo2(torch.Tensor):
def foo(self):
return 5
f = Foo2()
self.assertEqual(f.foo(), 5)
Copy link
Contributor

@gchanan gchanan May 8, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'd prefer doing something with the Tensor, e.g. you could just check it's shape is () here.


# Functions to test negative dimension wrapping
METHOD = 1
INPLACE_METHOD = 2
Expand Down
4 changes: 3 additions & 1 deletion torch/csrc/tensor/python_tensor.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -159,7 +159,9 @@ static void py_initialize_tensor_type(PyTypeObject& type, const char* name, PyOb
((PyObject*)&type)->ob_refcnt = 1;
((PyObject*)&type)->ob_type = &metaclass;
type.tp_basicsize = sizeof(PyTensorType);
type.tp_flags = Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE;
// Subclassing from torch.<ScalarType>Tensor isn't supported.
// (Py_TPFLAGS_BASETYPE omitted). Subclassing torch.Tensor still allowed.
type.tp_flags = Py_TPFLAGS_DEFAULT;
type.tp_name = name;
type.tp_new = Tensor_new;
if (PyType_Ready(&type) < 0) {
Expand Down