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

Qnn fully connected #3910

Merged
merged 7 commits into from
Sep 22, 2019
Merged

Qnn fully connected #3910

merged 7 commits into from
Sep 22, 2019

Conversation

shoubhik
Copy link
Contributor

@shoubhik shoubhik commented Sep 6, 2019

This is continuation of the Qnn dialect. After the convolution #3580 this the implementaion of Dense layer with quantized inputs. In future revisions we will add schdules for more optimal performance.

@shoubhik
Copy link
Contributor Author

shoubhik commented Sep 6, 2019

@anijain2305 @zhiics @vinx13 can you take a look at this PR.

Copy link
Contributor

@anijain2305 anijain2305 left a comment

Choose a reason for hiding this comment

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

Overall LGTM. Some minor comments.

src/relay/pass/pattern_util.h Outdated Show resolved Hide resolved
python/tvm/relay/qnn/op/qnn.py Show resolved Hide resolved
src/relay/qnn/op/dense.cc Show resolved Hide resolved
src/relay/qnn/op/dense.cc Outdated Show resolved Hide resolved
src/relay/qnn/op/dense.cc Outdated Show resolved Hide resolved
src/relay/qnn/op/dense.cc Outdated Show resolved Hide resolved
src/relay/qnn/op/dense.cc Outdated Show resolved Hide resolved
src/relay/qnn/op/dense.cc Show resolved Hide resolved
tests/python/relay/test_qnn_dense.py Outdated Show resolved Hide resolved
src/relay/qnn/op/convolution.cc Show resolved Hide resolved
@shoubhik
Copy link
Contributor Author

shoubhik commented Sep 9, 2019

@zhiics @vinx13 @jackwish can you take a look at this PR.

@anijain2305
Copy link
Contributor

@shoubhik Can you please fix the lint?

@zhiics @vinx13 Can you please review?

@shoubhik
Copy link
Contributor Author

Fixed the test related issue.

Copy link
Member

@zhiics zhiics left a comment

Choose a reason for hiding this comment

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

LGTM, just left a few nits.
@jackwish Could you take a look as well when you have cycles?

Also, plz fix the ci error.

python/tvm/relay/qnn/op/qnn.py Show resolved Hide resolved
python/tvm/relay/qnn/op/qnn.py Show resolved Hide resolved
@shoubhik
Copy link
Contributor Author

@jackwish Could you take a look

Copy link
Contributor

@zhenhuaw-me zhenhuaw-me left a comment

Choose a reason for hiding this comment

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

LGTM. Sorry that I have missed this :)

@shoubhik
Copy link
Contributor Author

@zhiics can you merge this.

@zhiics zhiics merged commit 43f54a5 into apache:master Sep 22, 2019
@zhiics
Copy link
Member

zhiics commented Sep 22, 2019

Thanks everyone. This is now merged.

wweic pushed a commit to wweic/tvm that referenced this pull request Sep 30, 2019
* Qnn Dense layer.

* Reformatting code.

* Reformatting code and making the test case more readable.

* Fixing lint issues.

* Fixing test method names to pass the nose related configurations.

* Aligning the code for code style.
wweic pushed a commit to wweic/tvm that referenced this pull request Sep 30, 2019
* Qnn Dense layer.

* Reformatting code.

* Reformatting code and making the test case more readable.

* Fixing lint issues.

* Fixing test method names to pass the nose related configurations.

* Aligning the code for code style.
wweic pushed a commit to neo-ai/tvm that referenced this pull request Oct 1, 2019
* Qnn Dense layer.

* Reformatting code.

* Reformatting code and making the test case more readable.

* Fixing lint issues.

* Fixing test method names to pass the nose related configurations.

* Aligning the code for code style.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants