Skip to content
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

Jump to bottom

[ManifestAlloc] Handle TupleType inputs in CheckReshapeOnly #6776

Merged
merged 2 commits into from
Oct 28, 2020
Merged

[ManifestAlloc] Handle TupleType inputs in CheckReshapeOnly #6776

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
8a23866
Changes in CheckReshapeOnly to support TupleTypes as input
rohanmukh Oct 27, 2020
6a5613b
[ManifestAlloc] Handle TupleType inputs in CheckReshapeOnly
rohanmukh Oct 28, 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
5 changes: 5 additions & 0 deletions python/tvm/relay/transform/memory_alloc.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,6 +84,11 @@ def visit_call(self, call):
for arg in call.args:
self.visit(arg)

def visit_var(self, var):
var_type = var.checked_type
if not isinstance(var_type, ty.TensorType):
self.reshape_only = False


def is_reshape_only(func):
"""Check if the primitive function contains only reshape ops."""
Expand Down
16 changes: 16 additions & 0 deletions tests/python/relay/test_vm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -754,5 +754,21 @@ def test_vm_reshape_tensor():
check_result([x_np, y_np], x_np.reshape([8, 2, 8]), mod)


def test_vm_reshape_tuple(x_shape=(1, 4, 2), y_shape=(1, 2, 10)):
tup = relay.var(
"tup",
type_annotation=relay.TupleType([relay.TensorType(x_shape), relay.TensorType(y_shape)]),
)
out = relay.reshape(relay.TupleGetItem(tup, 0), (1, -1))
f = relay.Function([tup], out)

x_data = np.random.uniform(size=x_shape).astype("float32")
y_data = np.random.uniform(size=y_shape).astype("float32")

for tgt, ctx in tvm.testing.enabled_targets():
res = veval(f, (x_data, y_data), ctx=ctx, target=tgt)
tvm.testing.assert_allclose(res.asnumpy(), np.reshape(x_data, (1, -1)))


if __name__ == "__main__":
pytest.main([__file__])