saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-05-18 21:21:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

[WebNN] Remove validation for coordinate_transformation_mode (#22811)

Browse source 
The performance cost of falling back to the CPU EP is high for several
resampling nodes and causes multiple partitions in SD Turbo and VAE
decoder. Since the asymmetric mode with nearest to floor and integer
scales is identical to half_pixel anyway, stick with the WebNN EP.
This commit is contained in:
shiyi 2024-11-14 03:12:00 +08:00 committed by GitHub
parent ff57ac4f3d
commit 3adcf4d714
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 3 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
onnxruntime/core/providers/webnn/builders/impl/resize_op_builder.cc
View file

@ -267,15 +267,9 @@ bool ResizeOpBuilder::IsOpSupportedImpl(const InitializedTensorSet& initializers
return false;
}
// coordinate_transformation_mode
// Spec issue for supporting more coordinate transformation modes:
// https://github.com/webmachinelearning/webnn/issues/270
const std::string coordinate_transformation_mode = helper.Get("coordinate_transformation_mode", "half_pixel");
if (coordinate_transformation_mode != "half_pixel") {
LOGS(logger, VERBOSE) << "Resize does not support coordinate_transformation_mode: "
<< coordinate_transformation_mode;
return false;
}
// Ignore coordinate_transformation_mode because WebNN only supports half_pixel mode.
// TODO: Validate coordinate_transformation_mode. Related spec issue for supporting attribute coordinate
// transformation modes: https://github.com/webmachinelearning/webnn/issues/270
// exclude_outside
const auto exclude_outside = helper.Get("exclude_outside", 0);