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

Need some advice on how to use the repo to train a big dataset. #35

Open
zydjohnHotmail opened this issue Mar 17, 2022 · 7 comments
Open

Comments

@zydjohnHotmail
Copy link

Hello:
I am new to Pytorch, and I want to learn how to use StyleGan2, and I found a big dataset from here:
https://github.com/NVlabs/metfaces-dataset

I download the datasets and found the data is huge, more than 17GB on disk.
If I want to train the dataset and export the trained model in an ONNX format.
How I can start?
I have Python 3.9 on Windows 10, and Pytorch version 1.11.0.
My PC has 16-core CPU with 128GB RAM with a NVIDIA graphics card, so I think hardware is good enough to start a big project.
Thanks,

@bes-dev
Copy link
Owner

bes-dev commented Mar 18, 2022

@zydjohnHotmail hi, thanks for interesting to our project. So, our framework is only for compression of the pretrained StyleGAN2 models, not for training from scratch. To use our framework for your custom dataset, firstly you need to train StyleGAN2 model using external tool https://github.com/rosinality/stylegan2-pytorch, after you can follow README.md of the our repo to compress and convert to ONNX.

@zydjohnHotmail
Copy link
Author

Hello:
Can I use the repo to convert pre-trained StyleGAN2 models to ONNX?

@bes-dev
Copy link
Owner

bes-dev commented Mar 24, 2022

@zydjohnHotmail no, we support ONNX/CoreML export only for MobileStyleGAN

@zydjohnHotmail
Copy link
Author

Hello:
Could you please provide some links where I can download pre-trained MobileStyleGAN models, so I can do some testing.
Thanks,

@bes-dev
Copy link
Owner

bes-dev commented Mar 24, 2022

@zydjohnHotmail
Copy link
Author

Hello:
F:\Python\MobileStyleGAN.pytorch-develop>pip install -r requirements.txt
Collecting git+git://github.com/fbcotter/pytorch_wavelets.git (from -r requirements.txt (line 9))
Cloning git://github.com/fbcotter/pytorch_wavelets.git to c:\users\john\appdata\local\temp\pip-req-build-fvhd01ni
Running command git clone --filter=blob:none --quiet git://github.com/fbcotter/pytorch_wavelets.git 'C:\Users\John\AppData\Local\Temp\pip-req-build-fvhd01ni'
fatal: remote error:
The unauthenticated git protocol on port 9418 is no longer supported.
Please see https://github.blog/2021-09-01-improving-git-protocol-security-github/ for more information.
error: subprocess-exited-with-error

× git clone --filter=blob:none --quiet git://github.com/fbcotter/pytorch_wavelets.git 'C:\Users\John\AppData\Local\Temp\pip-req-build-fvhd01ni' did not run successfully.
│ exit code: 128
╰─> See above for output.

note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

× git clone --filter=blob:none --quiet git://github.com/fbcotter/pytorch_wavelets.git 'C:\Users\John\AppData\Local\Temp\pip-req-build-fvhd01ni' did not run successfully.
│ exit code: 128
╰─> See above for output.

note: This error originates from a subprocess, and is likely not a problem with pip.

F:\Python\MobileStyleGAN.pytorch-develop>
It seems some requirements can't be meet. Did you see this error before?

@bes-dev
Copy link
Owner

bes-dev commented Mar 24, 2022

nope, you have some problems with your git settings

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

No branches or pull requests

2 participants