Zaigham Abbas

Confusing output with Demo

3

Hey, so I basically followed the given demo to do CUT with the following commands in order: -> bash ./datasets/download_cut_dataset.sh grumpifycat -> python train.py --dataroot ./datasets/grumpifycat --name grumpycat_CUT --CUT_mode CUT...

When I try to run the demo with multiple GPUS, I get the following errors: **Command:** python train.py --dataroot ./datasets/grumpifycat --name grumpycat_CUT --CUT_mode CUT --gpu_ids 0,1,2,3,4,5,6,7 --batch_size=2048  It works...

Good morning, I was wondering if there is a file or something to reproduce the provided results on a given dataset. Regards, Zaigham

 Hey, so I the instruction but run into these two errors. I am using Ubuntu 18.04 Bionic bearver. Can you help?

I was just wondering if you could provide details on this.  The source_skeleton, target_skeleton are both of size (20, 256, 176). What do these 20 planes represent? i saw...

Code for using the ImageNet pretrained model

4

I thought would be helpful for other people. I had issues with getting the resnet used in this repo running properly, but the given weights work well with Pytorch's default...

Which section of the code here addresses the moco trick?

What mean, std and other datapreprocessing transforms did you use for ImageNet?

1

It would really help me to know this information.

yaml.constructor.ConstructorError

1

I am facing this weird error, dont really know what to do about it. Was running it in a CONDA env. as instructed. Below can be seen a screenshot.