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

Add warning about future computation change for ConvTranspose with auto_pad #11984

Merged
merged 8 commits into from
Jun 29, 2022
Merged

Add warning about future computation change for ConvTranspose with auto_pad #11984

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
99c47d8
Add warning about future computation change for Convtranspose with au…
jcwchen Jun 24, 2022
554e5bc
improve msg
jcwchen Jun 24, 2022
a7963cf
update TODO to make lint happy
jcwchen Jun 24, 2022
77f717a
update more contents for warning and add if
jcwchen Jun 24, 2022
9b10e52
valid was not infected
jcwchen Jun 24, 2022
cca0c01
move it into kernel registration
jcwchen Jun 29, 2022
b1d8430
parse auto_pad myself
jcwchen Jun 29, 2022
3b6fb87
try to use conv_transpose_attrs_.auto_pad directly
jcwchen Jun 29, 2022
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
10 changes: 9 additions & 1 deletion onnxruntime/core/providers/cpu/nn/conv_transpose.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,15 @@ namespace onnxruntime {
template <typename T>
class ConvTranspose : public OpKernel {
public:
ConvTranspose(const OpKernelInfo& info) : OpKernel(info), conv_transpose_attrs_(info) {}
ConvTranspose(const OpKernelInfo& info) : OpKernel(info), conv_transpose_attrs_(info) {
if (conv_transpose_attrs_.auto_pad == AutoPadType::SAME_UPPER ||
conv_transpose_attrs_.auto_pad == AutoPadType::SAME_LOWER) {
// TODO(jcwchen): #9740 ORT 1.13 will correct the logic by switching them to meet ONNX spec
LOGS_DEFAULT(WARNING) << "The existing bug in the padding distribution for auto_pad type"
<< " SAME_UPPER/SAME_LOWER will be fixed in next ORT 1.13 release and hence the"
<< " results of ConvTranspose operator using the above auto_pad type(s) will be different.";
}
}

Status PrePack(const Tensor& tensor, int input_idx, AllocatorPtr alloc,
/*out*/ bool& is_packed,
Expand Down