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
add dilation case for ConvTranspose op #1797
add dilation case for ConvTranspose op #1797
Conversation
please also push the updated doc, and generated test data; see https://github.com/onnx/onnx/blob/master/docs/AddNewOp.md#step-5-update-the-doc-and-generate-the-test-data for help |
new test model uploaded. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, Thanks!
@houseroad Hi lu just want to confirm the addition of testcase wont break things on ur side |
|
||
node = onnx.helper.make_node("ConvTranspose", ["X", "W"], ["Y"], dilations=[2, 2]) | ||
|
||
y = np.array([[[[21., 56., 13., 16., 2.], # [1, 1, 5, 5] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I checked, the results I got was:
tensor([[[[21.1887, 56.1887, 13.1887, 16.1887, 2.1887],
[63.1887, 35.1887, 67.1887, 10.1887, 14.1887],
[24.1887, 22.1887, 76.1887, 76.1887, 21.1887],
[ 9.1887, 5.1887, 88.1887, 45.1887, 63.1887],
[ 3.1887, 2.1887, 33.1887, 18.1887, 54.1887]]]]
Did I miss something here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Mind to share your script here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Never mind, I forgot to set bias. Apparently, that noise is from bias.
The result look good to me now.
|
||
node = onnx.helper.make_node("ConvTranspose", ["X", "W"], ["Y"], dilations=[2, 2]) | ||
|
||
y = np.array([[[[21., 56., 13., 16., 2.], # [1, 1, 5, 5] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Never mind, I forgot to set bias. Apparently, that noise is from bias.
The result look good to me now.
* add dilation case for ConvTranspose op * add comment * upload test data * upload changed docs * fix check err * add comment * add space * revert unintended change * revert unintended change
* add dilation case for ConvTranspose op * add comment * upload test data * upload changed docs * fix check err * add comment * add space * revert unintended change * revert unintended change
Add case to test ConvTranspose op with dilation.