Raise friendly errors in case of num_epochs is less than num_warmup_epochs in CIFAR10 and Transformers examples

[KickItLikeShika]

I was trying to benchmark torch.inference_mode vs torch.no_grad on CIFAR10 example, so I just wanted to run for 3 epochs and average the benchmarks, the default warmup epochs is 4, so the scheduler will raise this error, which is not very clear to new users

f"Milestones should be increasing integers, but given {pair[0]} is smaller "
ValueError: Milestones should be increasing integers, but given 291 is smaller than the previous milestone 388

So I suggest to check num_epochs vs num_warmup_epochs inside run method and raise more clear error message for the users.

[himanshu007-creator]

Hey , i would like to work on this issue. Can you guide me further?! on what needs to be done🙂

[KickItLikeShika]

Hey @himanshu007-creator, thanks a lot for showing interest, we have these 2 examples:

• https://github.com/pytorch/ignite/tree/master/examples/contrib/cifar10
• https://github.com/pytorch/ignite/tree/master/examples/contrib/transformers

in both, if the num_epochs is less the num_warmup_epochs, the error in the description will be raised,
we want to check inside the run method in both examples, if num_epochs is less than num_warmupds_epochs we should raise a friendly error message explaining what's the error and why, and what the user should do (either increase the num_epochs or decrease num_warmup_epochs)

[FarehaNousheen]

Please brief me about the issue a little more. I'll work on this issue.

[KickItLikeShika]

Hi @FarehaNousheen, please check the comment above #2207 (comment), it explains what we need to do in depth. If you still can't get what we need to do, please ping me on discord.

[DebarshiChanda]

Hey @KickItLikeShika, I think adding an assertion can solve this issue. Can you assign this to me?

[theory-in-progress]

Hey @KickItLikeShika, I think I've made the necessary changes in three of the files to raise an user friendly message when num_epochs is greater than num_warmup_epochs. Am I free to create a Pull Request, since this hasn't been assigned to me...

[vfdev-5]

@DebarshiChanda @theory-in-progress to be fair if someone is assigned to the issue his/her PR has higher prio (conditioned also by the time it takes to work on the issue). Thanks a lot to you both on working on this tiny issue ! In this particular case, @theory-in-progress 's PR looks better and we'll proceed with it.
@DebarshiChanda please pick another issue to work on, thanks a lot !

[theory-in-progress]

Of course! sorry for barging in... Thanks for the opportunity @vfdev-5! I'm happy to work on this issue