Aleksa Gordić

I'm aware that Caffe used BGR, 0..255 images mean ([123.675, 116.28, 103.53]) normalized using ImageNet's mean to train VGG models. Whereas PyTorch which I'm using uses RGB, 0..1 range images...

Do we have a way of pinging Johnson for this thing? It'd be nice to explain the meaning of that magic number and whether our hypothesis is true. @jcjohnson

https://github.com/gordicaleksa/pytorch-nst-feedforward I've implemented the paper without any tanh activation and it works like a charm! Could you maybe link this repo as a PyTorch implementation of Johnson's original paper? I've...

Did you solve this one? With gym==0.17.3 I'm getting periodic slowdowns manifested in drops in my FPS (frames/env steps per second) metric. Here is an example: 

Hey! Hm, interesting Tamas one of the coauthors of igraph develops all of these tools for Linux, actually, Windows users are usually second-class citizens having problems with igraph. I had...

The current setup should work (not I'm on Windows and weird things may happen on other platforms I haven't tested it). Check out my answer to a related issue: https://github.com/gordicaleksa/pytorch-GAT/issues/2#issuecomment-773173854...

Hey I will, probably this or next weekend. I hate those stupid errors myself. I'll add a jupyter notebook hopefully that'll help.

Hey @PJ-Finlay! I don't currently have the bandwidth to deal with NST PRs/issues. I just dealt with my [DeepDream repo](https://github.com/gordicaleksa/pytorch-deepdream) - did lots of refactoring, added full support for cmd...

Hi Nick! I have it on my backlog to try and make it work on the [0, 1] range as it feels more natural for PyTorch models as they were...

@neilbGH are you able to step through the code with your setup?