Nuke 13.2 | Faster Machine Learning with Multiple GPUs

[Music] with nuke 13.2 we have continued to improve our machine learning tools enhancing the copycat node with faster training support for multiple gpus and multi-channel training we have also removed any limitation on the number of training images and enabled the ability to train on a headless nuke with the minus x flag with this update we focused on leveraging more of your gpu power to make the most of your time now not only is single gpu training up to 30 faster but you can also take advantage of setups with multiple gpus enabling you to iterate quicker through your experiments with these new changes you can either speed up training by running copycat on all your gpus simultaneously or run different training sessions on each gpu allowing you to do more experiments at the same time furthermore we want you to make the most out of your data to do so copycat can now support more than four channels exactly as many as your gpu and nuke can handle empowering you to train networks for a variety of more advanced use cases or simply to avoid training similar networks over and over again like in this case instead of having four different networks i have trained one network to do the four different use cases with this release your copycat pipeline will be more streamlined faster and more powerful than ever before so level up your workflow and make the most out of your copycat node [Music]

Channel: Foundry

Views: 3,252

Keywords: compositing showreel, compositing after effects, compositing in nuke, compositing tutorial, compositing reel, compositing vfx, compositing animation, compositing artist, compositing cgi, vfx compositing demo reel, compositing green screen footage, deep compositing, post-production, nuke, unrealReader, unreal, foundry nuke, foundry

Id: fQl7K9BAUOo

Length: 1min 33sec (93 seconds)

Published: Thu Apr 14 2022