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.
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
ONNX conformance test results #21078
Labels
bug
category: dnn (onnx)
ONNX suport issues in DNN module
effort: ∞
Do not start this alone. You can discuss / improve existed proposals.
feature
Milestone
Comments
This was referenced Nov 29, 2021
Merged
This was referenced Mar 13, 2023
6 tasks
@fengyuentau I believe "Gemm B blob is supported only as const" is complete? |
No, B can be either const or variable. See opencv/modules/dnn/src/layers/gemm_layer.cpp Lines 210 to 215 in 76548e2
|
Related: #25442 |
asmorkalov
pushed a commit
that referenced
this issue
Apr 22, 2024
…reduce Comments for parser denylist #25465 Relates to #21078 This PR is designed to figure out why the test in `test_onnx_conformance_layer_parser_denylist.inl.hpp` fails. Currently, conformance tests do not pass for the following reasons: 1. BOOL, INT(8, 16) types are not supported **(MAJOR)** 2. Some layers can not be created due to various reasons **(MAJOR)** 3. Shape mismatches while creating layers **(MAJOR)** 4. Some layers are expected to support dynamic parameter initialization **(MAJOR)** 5. Some layers are expected to receive weight as inputs (no idea why that is needed) **(MAJOR)** 6. Other unknown reasons **(MAJOR)** - These are the most frequently encountered reasons for test failure. The style of comments is not consistent everywhere. Let's keep this PR without merging, just for our reference. A couple of tests are commented on since they have passed on the MacOS platform. ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [x] I agree to contribute to the project under Apache 2 License. - [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [x] The PR is proposed to the proper branch - [x] There is a reference to the original bug report and related work - [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [x] The feature is well documented and sample code can be built with the project CMake
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Labels
bug
category: dnn (onnx)
ONNX suport issues in DNN module
effort: ∞
Do not start this alone. You can discuss / improve existed proposals.
feature
System information (version)
Detailed description
I ran ONNX conformance tests from official repo, and the found the following issues:
DataLayer
Add
AveragePool
count_include_pad
not supportedauto_pad
is deprecated, but we support it, but not making distinction betweensame_upper
andsame_lower
MaxPool
ceil_mode
is set by checking the existence ofpad_mode
; fix ceil_mode for Average/MaxPooling #21159auto_pad
is deprecated, but we support it, but not making distinction betweensame_upper
andsame_lower
dilations
BatchNormalization
mean
blob is supported only as constClip
min
andmax
shouldn't be differentiable, model error?; Fix issue 22015, let Clip layer support 1-3 inputs #22100Concat
Concat
output dims are [2, 2] instead of [4]Constant
ConvTranspose
W
blob is supported only as constDiv
ELU
alpha
parameter; Add alpha parameter to ELU layer #21160, Add alpha parameter to ELU (CUDA) #21161Flatten
axis
, ifaxis
=0 outputs [X, 1] instead of [1, X]; fix Flatten layer #21283Gemm
B
blob is supported only as const, ONNX conformance test results #21078 (comment)GRU
InstanceNormalization
S
blob is supported only as constLeakyReLU
alpha
should be 0.01; fix Clip, LeakyReLU, LRN, Split defaults #21152LogSoftMax
/SoftMax
axis
should be -1 - if the opset is v11, otherwise it's 1..._expanded
models should be fused; Add Log/Softmax simplification #21162LRN
alpha
should be 0.0001; fix Clip, LeakyReLU, LRN, Split defaults #21152LSTM
Max/Min/Mul
Pow
PRelu
Slope
blob is supported only as constReduceMax/ReduceMean
keepdims
=true; Fix a few issues in MaxUnpool, ReduceMean/Max/Sum, Pad #21259ReduceSum
axes
arent handled as attributes, should be optional const input; DNN: Reduce Layer (add dynamic batch and ReduceSum support) #22199Resize
Split
axis
is set to 1 instead of 0; fix Clip, LeakyReLU, LRN, Split defaults #21152Sub
Sum
Transpose
Upsample
scales
blob is supported only as const or attributeAND
Steps to reproduce
Clone ONNX repo, clone my branch, fix paths in the code and run
opencv_test_dnn
.Issue submission checklist
forum.opencv.org, Stack Overflow, etc and have not found solution
The text was updated successfully, but these errors were encountered: