Skip to content

Conversation

varunagrawal
Copy link
Contributor

@varunagrawal varunagrawal commented Oct 16, 2018

  • Fixes and improvements to ROIAlign forward pass.
  • Working implementation and improvements of ROIAlign backwards for GPU & CPU.
  • Tests based on results from Caffe2 to verify correctness.

Support for autograd.gradcheck is still needed.

@fmassa
Copy link
Member

fmassa commented Oct 17, 2018

This is awesome, thanks!

There are some conflicts with this branch, can you rebase and add an autograd.gradcheck test, and the this is good to merge

@varunagrawal varunagrawal force-pushed the roi-align-backwards-and-tests branch from 1ab7b50 to d3dc4a1 Compare October 17, 2018 17:14
@varunagrawal varunagrawal force-pushed the roi-align-backwards-and-tests branch from 8fcd3e5 to 110d998 Compare October 17, 2018 17:18
@varunagrawal varunagrawal force-pushed the roi-align-backwards-and-tests branch from b5bd917 to ae3c453 Compare October 17, 2018 22:12
@varunagrawal
Copy link
Contributor Author

@fmassa done and done!

@fmassa
Copy link
Member

fmassa commented Oct 18, 2018

Merged #632, so this is redundant, right?

@varunagrawal
Copy link
Contributor Author

Yup. Closing this.

rajveerb pushed a commit to rajveerb/vision that referenced this pull request Nov 30, 2023
* disabled documents shuffling (pytorch#11)

Co-authored-by: Anmol Gupta

* updated the mlperf logging keywords (pytorch#12)

Co-authored-by: Anmol Gupta

* updates to mlperf logger keywords

* updates to readme and validation preprocessing script (pytorch#13)

Co-authored-by: Anmol Gupta <[email protected]>

* updated the model name

---------

Co-authored-by: anmolgupt <[email protected]>
Co-authored-by: Anmol Gupta <[email protected]>
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