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.

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

Follow up from CMSIS-NN Pooling failure #9708

Merged
merged 1 commit into from
Dec 13, 2021

Conversation

ashutosh-arm
Copy link
Contributor

The PR addresses comments by @ekalda and @manupa-arm from #9682 and #9531. Those issues were about Pooling support in CMSIS-NN and fix for a network failure in Conv2D partitioning.

Change-Id: I68861b97d294744b3474f08b23ac890c3222c16c
@ashutosh-arm
Copy link
Contributor Author

ashutosh-arm commented Dec 13, 2021

Could you folks @ekalda @manupa-arm please take a look at it?

Copy link
Contributor

@manupak manupak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

@manupak manupak merged commit c19f193 into apache:main Dec 13, 2021
@manupak
Copy link
Contributor

manupak commented Dec 13, 2021

Thanks! @ashutosh-arm

@ashutosh-arm ashutosh-arm deleted the pooling_followup branch December 15, 2021 09:34
ylc pushed a commit to ylc/tvm that referenced this pull request Jan 7, 2022
This commit fixes few comments in TIR2Runtime pass of CMSIS-NN target.
These comments specify layout used by CMSIS-NN API 
for input and filter shapes.

Another fix was done to the filter layout calculations. 
Instead of hard coded values for dimensions, 
filter_shape.find("H") was used to locate a particular value.

Third fix was done to the padding API used by Conv2D and Pooling tests.
It was made generic for TFLu's "SAME" padding type.
yangulei pushed a commit to yangulei/tvm that referenced this pull request Jan 11, 2022
This commit fixes few comments in TIR2Runtime pass of CMSIS-NN target.
These comments specify layout used by CMSIS-NN API 
for input and filter shapes.

Another fix was done to the filter layout calculations. 
Instead of hard coded values for dimensions, 
filter_shape.find("H") was used to locate a particular value.

Third fix was done to the padding API used by Conv2D and Pooling tests.
It was made generic for TFLu's "SAME" padding type.
yangulei pushed a commit to yangulei/tvm that referenced this pull request Jan 12, 2022
This commit fixes few comments in TIR2Runtime pass of CMSIS-NN target.
These comments specify layout used by CMSIS-NN API 
for input and filter shapes.

Another fix was done to the filter layout calculations. 
Instead of hard coded values for dimensions, 
filter_shape.find("H") was used to locate a particular value.

Third fix was done to the padding API used by Conv2D and Pooling tests.
It was made generic for TFLu's "SAME" padding type.
ylc pushed a commit to ylc/tvm that referenced this pull request Jan 13, 2022
This commit fixes few comments in TIR2Runtime pass of CMSIS-NN target.
These comments specify layout used by CMSIS-NN API 
for input and filter shapes.

Another fix was done to the filter layout calculations. 
Instead of hard coded values for dimensions, 
filter_shape.find("H") was used to locate a particular value.

Third fix was done to the padding API used by Conv2D and Pooling tests.
It was made generic for TFLu's "SAME" padding type.
qsqqsqqsq-intellif pushed a commit to qsqqsqqsq-intellif/tvm that referenced this pull request Apr 29, 2022
This commit fixes few comments in TIR2Runtime pass of CMSIS-NN target.
These comments specify layout used by CMSIS-NN API 
for input and filter shapes.

Another fix was done to the filter layout calculations. 
Instead of hard coded values for dimensions, 
filter_shape.find("H") was used to locate a particular value.

Third fix was done to the padding API used by Conv2D and Pooling tests.
It was made generic for TFLu's "SAME" padding type.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants