ProGamerGov

@Vigilence The models you suggest are using ResNet architecture, which can require extra work to use for things like style transfer. neural-style-pt uses VGG models by default, but you can...

@gateway The multidevice strategy just simply splits the model layers across different devices. The layer order in a model dictates which layers go on which device, based on your specified...

@IridiumMaster The `-multidevice_strategy` parameter tells the code where to slice / cut the model, and in your case 2 GPUs means you want to have the model cut into 2...

@RobertGoacher Using multiple GPUs can be a bit slower than using a single GPU. Also, there's a small increase in memory that results from using multiple GPUs as well I...

@petered Are you using `-init image -init_image image.jpg`? Because `-init_image` only works when the initalization is set to `image`?

The FileNotFound errors look like errors caused by another error that occurs earlier. The first step in the starry_stanford.sh script takes your input images, and produces an output image. The...

@Subash-Chandra The PyTorch site shows that the PyTorch Conda install only supports CUDA 9.2, CUDA 10.1, and CUDA 10.2: https://pytorch.org/get-started/locally/ Also, unless you are installing from source, I think cuDNN...

> I'm not sure why it's able to compute the first 15 or so images, and only fail after that. Can you elaborate on that?

@Subash-Chandra I can't seem to figure if the error is because of a lack of memory or something else. I originally created starry_stanford.sh with a GPU with 12 GB of...

First you should check if PyTorch sees your devices correctly and that CUDA works. Try running this in the Python interpreter and seeing what it shows: ``` import torch torch.__version__...