Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

GPU #1

Open
hitsfm opened this issue Aug 12, 2022 · 1 comment
Open

GPU #1

hitsfm opened this issue Aug 12, 2022 · 1 comment
Labels
enhancement New feature or request

Comments

@hitsfm
Copy link

hitsfm commented Aug 12, 2022

I really like this. I really do. Was going to build my own similar idea using some other python packer and make this as a standalone app as I could not find such until now. Only issue I have with this is I have 4 computers all with good NVIDIA cards with CUDA. In some of my system I was forced to "downgrade" to install your recommended CUDA version. My RTX card on one system does not show up. The the Geforce cards do show up but all cause a app crash during dream start. I tried on like 4 systems with windows 10 and all GPU's that support CUDA with the Cuda framework installed. I have an i9 CPU so using CPU really only takes a minute or so. Not a big issue. But just thought I would let you know as the average person probably struggles with this GPU issue. Thanks for packing this up. Very useful.

@dreavjr dreavjr added the enhancement New feature or request label Aug 13, 2022
@dreavjr
Copy link
Owner

dreavjr commented Aug 13, 2022

Thank you for this report. User-friendly GPU support for Tensorflow is kind of mess right now: the framework does not really seem intended to be shipped inside consumer applications. Now that PyTorch added support for Apple Silicon Metal, I wonder if it would be worthy porting the entire thing to it.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

2 participants