555 points by deeplearning 6 months ago flag hide 9 comments
johnsmith 6 months ago next
This is such an interesting topic! Can't wait to see how GANs can be used in image synthesis.
machinelearner 6 months ago next
@johnsmith I've had a lot of success using GANs with GPU acceleration on a local machine. The training time is significantly reduced when compared to using just a CPU.
gpuguru 6 months ago next
@machinelearner Absolutely. I use a GTX 1080 Ti for my GAN projects and it's been a game-changer in terms of training time.
anonymouscoward 6 months ago prev next
Anyone have any experience with running GANs on a local machine? How's the performance?
datawhiz 6 months ago next
@anonymouscoward I've found that training GANs locally can be resource-intensive. You might want to consider using a cloud platform to distribute the workload.
cloudchamp 6 months ago next
@datawhiz I've had good results using a cloud-based platform for GAN training. You can spin up and down instances as needed and the cost is relatively low.
codecrusher 6 months ago prev next
I'm new to GANs and image synthesis. Is there a recommended resource for learning about this topic?
bookworm 6 months ago next
@codecrusher I'd recommend starting with the original paper on GANs by Goodfellow et al. It's a bit dense, but it's a great place to start. You can find it on arXiv.
tutorialtips 6 months ago prev next
@codecrusher There are also many great tutorials online that walk you through the process of implementing a GAN for image synthesis. Check out some of the resources on Medium and Kaggle.