X Tutup
The Wayback Machine - https://web.archive.org/web/20201106182945/https://github.com/oneapi-src/oneDNN/pull/844
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

Added references to relevant articles #844

Open
wants to merge 2 commits into
base: master
from

Conversation

@jczaja
Copy link
Contributor

@jczaja jczaja commented Sep 24, 2020

Added references to articles covering oneDNN or its concepts

Description

This PR is extending boiler plate README with section listing articles that are relevant oneDNN library.

Rendered document: https://github.com/jczaja/mkl-dnn/tree/prv-papers (section "Relevant articles")

Added references to articles covering oneDNN or its concepts
Copy link
Contributor

@emfomenk emfomenk left a comment

Thanks Jacek!

README.md Outdated
@@ -52,6 +52,10 @@ applications enabled with oneDNN:
* [API reference](https://oneapi-src.github.io/oneDNN/modules.html) provides
a comprehensive reference of the library API.

# Relevant articles

This comment has been minimized.

@emfomenk

emfomenk Sep 24, 2020
Contributor

Suggested change
# Relevant articles
# Relevant Publications

This comment has been minimized.

@jczaja

jczaja Sep 25, 2020
Author Contributor

done

README.md Outdated
# Relevant articles
- [Anatomy Of High-Performance Deep Learning Convolutions On SIMD Architectures](https://arxiv.org/abs/1808.05567) Evangelos Georganas, Sasikanth Avancha, Kunal Banerjee, Dhiraj Kalamkar, Greg Henry, Hans Pabst, Alexander Heinecke
- [Applying the Roofline model for Deep Learning performance optimizations](https://arxiv.org/abs/2009.11224) Jacek Czaja, Michal Gallus, Joanna Wozna, Adam Grygielski, Luo Tao

This comment has been minimized.

@emfomenk

emfomenk Sep 24, 2020
Contributor

(random place)

I would suggest moving the "Relevant Publications" right before # Support section.
But this is a purely stylistic thing, so I would like to here the opinion of others.

On the other hand, I am not sure if we really want to have the publications on the main page. I checked github main page of some OSS projects (LLVM, TF, PyTorch, FBGEMM, FFTW, MxNet, TVM, ...) and almost nowhere found the publication section. So, maybe the alternative is to introduce a new section in the documentation (similar to what MLIR has: https://mlir.llvm.org/talks/).

Tagging @vpirogov

This comment has been minimized.

@jczaja

jczaja Sep 25, 2020
Author Contributor

I moved "Relevant Publications" to pointed location . On which document it should be part of, my thinking was like that:
If I were first time looking at this project I would check where it is used, github activity etc. and then seeing it has some related publications I would be additionally a bit more convinced that it is industry and academia proven open source software.

Additionally I made this PR , because there were two occasions when I was looking for article describing inner mechanism on oneDNN (Anatomy of High-Performance....) and remembered only that it was mentioned in issues. So each time it took a little bit too long to find it. So with this PR it will be very straight forward.

Looking forward to your opinions.

This comment has been minimized.

@sbogusev

sbogusev Oct 7, 2020
Contributor

I agree with @emfomenk suggestion to create a separate Relevant Publications page in documentation (I would suggest ideally before API References). I think this is the easiest way for the user to see the studies that use oneDNN. Plus, some additional content can be added to such page, including other papers, presentations, etc.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Linked issues

Successfully merging this pull request may close these issues.

None yet

3 participants
You can’t perform that action at this time.
X Tutup