Optimize the calculation in the Deformable Convolution

[younessdkhissi]

Optimizing the function that calculate the bound used to calculate the number of paralle images in the defomable convolution.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/9232

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[meta-cla]

Hi @younessdkhissi!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at [email protected]. Thanks!

[meta-cla]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[AntoineSimoulin]

Hey @younessdkhissi, thanks for the PR. My understanding of your changes is that your are modifying the get_greatest_divisor_below_bound and get_greatest_divisor_below_bound functions to look for a divisor only in the [1, min(n/2, bound)] range rather than the [1, bound] range. The idea being largest possible divisor is n/2, so we don't need to search in the ]n/2, bound] range?

[younessdkhissi]

Hello @AntoineSimoulin, your understanding is right. My idea is to reduce the useless calculations made inside this function. I have noticed a significant increase of time computing when the batch size I use, during inference, for my deformable convolution models is a prime number (it is particular use case). That's why, I proposed this PR.