Tanishq Abraham

@kevinbird15 is this fixed yet? I don't think so since the test is the same and not updated... What needs to be fixed here?

The difference may be from the constants `C1=(0.01)`**2 and `C2=(0.03)**2` which assume the range of pixel values is 0 to 1. If instead it is `C1=(0.01*255)**2` and `C2=(0.03*255)**2`, along which...

@neomatrix369 Thank you for the suggestion. I am aware of W&B, and in fact fastai has great support for W&B thanks to the work of Boris Dayma. I plan to...

I have added FID (https://github.com/tmabraham/UPIT/commit/432f784a4ebcc88f7a5c76d4c4a629d8e484b74f) and tested it. Horse2Zebra FID reached ~91.7 with 10 epochs of training ([here](https://github.com/tmabraham/UPIT/blob/master/nbs/05_metrics.ipynb)), which is close to a full training CycleGAN benchmark of 89.7 reported...

It seems like `SaveModelCallback` works fine, it's just that the default of monitoring the `valid_loss` obviously wouldn't work if there isn't any validation loss. Apart from FID, I plan to...

I will probably wait until fastai+Accelerate support is in a good state (see fastai Discord for current status) and will see if I need to add anything for support then...

Thank you for your interest! Let me clarify one thing, are you trying to classify how well CycleGAN-generated images fools an existing model that classifies between real and fake images?...

Okay. This may be a little hard to do with the current version of the library. I plan to add metrics soon (#3) and maybe it's possible to treat this...

If you want, you can assign me this, I think I'll work on adding this during this week...

@rwightman If you want, I am willing to help out with any efforts regarding adapting code to PyTorch XLA, having used PyTorch XLA many times in the past...