saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-07-30 20:18:08 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Partially revert PR #7348 (#7702)

Browse source
This commit is contained in:
Hariharan Seshadri 2021-05-16 23:56:13 -07:00 committed by GitHub
parent 2b73163690
commit ebee380911
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 13 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

24
onnxruntime/core/providers/cuda/nn/conv.cc
View file

@ -205,17 +205,19 @@ Status Conv<T>::UpdateState(OpKernelContext* context, bool bias_expected) const
std::vector<int64_t> x_dims_cudnn = x_dims;
std::vector<int64_t> y_dims_cudnn = !post_slicing_required ? y_dims : y_dims_with_adjusted_pads;
if (rank < 2) {
// cudnn only takes 4D or 5D input, so pad dimensions if needed
// If input shape is [N, C, D], we pad the shape to [N, C, 1, D], as it results in
// more efficient algorithm than padding to [N, C, D, 1]
x_dims_cudnn.insert(x_dims_cudnn.begin() + 2, 1);
y_dims_cudnn.insert(y_dims_cudnn.begin() + 2, 1);
w_dims.insert(w_dims.begin() + 2, 1);
pads.insert(pads.begin(), 0);
pads.insert(pads.begin() + 2, 0);
kernel_shape.insert(kernel_shape.begin(), 1);
strides.insert(strides.begin(), 1);
dilations.insert(dilations.begin(), 1);
// TODO: Explore padding the provided input shape [N, C, D] to [N, C, 1, D]
// especially for EXHAUSTIVE algo search which may result in a better algo selection.
// Currently, we are padding it to [N, C, D, 1] as this seems to be the sweet spot
// for all algo search options: EXHAUSTIVE, HEURISTIC, and DEFAULT.
// See PR #7348 for more context.
x_dims_cudnn.push_back(1);
y_dims_cudnn.push_back(1);
w_dims.push_back(1);
pads.insert(pads.begin() + rank, 0);
pads.insert(pads.end(), 0);
kernel_shape.push_back(1);
strides.push_back(1);
dilations.push_back(1);
}
if (w_dims_changed) {