Fix iteration count

[jimmyyhwu]

This commit changes the iteration count to reflect the true number of iterations a model has been trained for. Previously, a model trained for n + 1 iterations would be saved as model_n.pth.

This also fixes the issue where resuming from a checkpoint would result in the model training for one iteration and then immediately saving to the same checkpoint file it resumed from.

[facebook-github-bot]

Thank you for your pull request and welcome to our community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. In order for us to review and merge your code, please sign up 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 the corporate CLA signed.

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

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Facebook open source project. Thanks!

[fmassa]

Hi,

Thanks a lot for the PR!

I think that a potentially better fix would be to modify the iteration_based_batch_sampler.py so that it takes into account this issue?

Or instead of modifying the Sampler, why can't we do

iteration = iteration + 1

right after the for iteration, ... loop?
This way we avoid propagating the +1 everywhere in the code.

Let me know what you think.

[jimmyyhwu]

Yes, that should work. I modified the commit to reflect the changes.

[fmassa]

Thanks!