Duplicating hyperparameters when training a FactorVAE

[SantiagoJN]

Hi!

I've been playing a little bit with the code(congratulations for the work by the way 😄 ), and I've seen that when training a FactorVAE model, both the batch size and number of epochs are duplicated:

disentangling-vae/main.py

Lines 191 to 194 in f045219

	if args.loss == "factor":
	logger.info("FactorVae needs 2 batches per iteration. To replicate this behavior while being consistent, we double the batch size and the the number of epochs.")
	args.batch_size *= 2
	args.epochs *= 2

Does anybody know the reason behind this operation? I've reviewed the original paper but I couldn't find anything related to this.
Thanks for the help!