Hey there, implementing the core architecture of style gan was fun, but i want to go more than just implementing the architecture and generating on pretrained model, I want to finetune that pretrained model with my own custom datasets. The guides available on the internet seems to be outdated and there is so source where I can get help.
please provide exhaustive guide which I can follow to create a masterpiece.
Thank you/
To create a masterpiece you need to be a master :).
In the tensorflow advanced techniques specialization they show how to fine-tune a pretrained model. I think the Deep Learning Specialization goes through that and the Generative AI for Large Language Models.
Hi, Luffy.
If you mean fine-tuning as a general matter, Gent has given you several options for courses here which cover that. If you specifically mean how to fine-tune StyleGAN on your particular dataset, then my guess would be that there is no-one listening here who has the deep expertise with StyleGAN to provide you with the exhaustive instructions that you are looking for. The best bet would be to do some searching and perhaps ask your favorite chatbot where to find that type of information. Read whatever papers there are out there about StyleGAN and see how far that gets you.
But maybe I’m wrong and you’ll get lucky and someone else listening here can actually respond. Stay tuned and hope for the best!
If you do end up going down this path under your own power, it would be a really nice public contribution if you could share what you learn about this. Then the next person who has your question could use that as a resource.
Regards,
Paul