multiblock mask collator now samples masks including last row and column

[jackwilkie]

Previously multiblock mask position sampler used the line torch.randint(0, self.height - h, (1,))... , however since randint is exclusive of the upper bound self.height - h, and similarly self.width - w is never sampled meaning the last row and column of image patches is never used for maksing. This is fixed by adding one to the upper bound.

[facebook-github-bot]

Hi @jackwilkie!

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 cla@meta.com. Thanks!

[LostGeorge]

Almost a year old by now, but related to this, I believe the following needs to be changed in order to properly sample context size = (h, w) [ratio 1.0] as well.

ijepa/src/masks/multiblock.py

Lines 67 to 70 in 52c1ae9

	while h >= self.height:
	h -= 1
	while w >= self.width:
	w -= 1

        while h > self.height:
            h -= 1
        while w > self.width:
            w -= 1