Reference parameters for large resolution training

[snakers4]

Hi!

Thank you for your amazing work with this repository.
I am experimenting with unsupervised learning - I would like to use the G / D of the trained GAN as encoders (there are ways to use G via knowledge distillation into a simpler CNN or papers like this one. I would like to build a model that can recognize 2 images of the same room.

The problem is that standard techniques (hashing, CNN => dimensionality reduction => distance, siamese networks) will either find only very similar images / clustered images or image replicas. If the image is shot from a bit different angle - these methods fail. Ideally I need to build an index of images by such variables as couches / TVs / carpets. Looks like DCGAN / WGAN learns something like this as a latent variable. If we have 100+ latent variables that represent this - they may work for different images.

I have a dataset of ~300-500k similar images, that mostly resemble lsun living room, but are shot in Russia. So far I have tried the variations of hyperparameters similar to these (please refer to the PR to see what the new parameters mean):

python3 main.py \
	--dataset fastfolder --dataroot ../data/data_for_wgan/all_imgs \
	--imgList ../data/generator_imgs_wgan \
	--ngf 64 --ndf 64 \
	--workers 6 --batchSize 256 --imageSize 256 \
	--nz 100 --niter 50 \
	--cuda --ngpu 2 \
	--n_extra_layers 0 \
	--experiment flat_dataset_224_0extra \
	--adam \
	--tensorboard_images --tensorboard

So far I did not achieve any success yet. Can you please add some guidelines into README on training on larger resolutions / batch sizes?

[snakers4]

Also on my dataset I managed to replicate training on 64x64 resolution (both for lsun and my dataset), but 256x256 still fails to converge meaningfully.

[kopytjuk]

@snakers4 I assume in order to create images with higher resolution you have to increase model's compelexity in terms of parameter number - maybe you will have some success with extra layers?