🎨 Stable Diffusion and generative models with David Marx
The MLOps Podcast
00:00
Is Stable Diffusion Still Python?
Stable diffusion is a model that doesn't take up a lot of resources to run. And so that actually has consequences on what efficient GPUs are. A100s are great for training, but they're kind of like chunky for the inference that we do because they just have so much memory on them. But you know, the cost effective GPU for training aren't necessarily the cost effectiveGPU for inference. They're not necessarily the message there. The DAGs are back in this and go. So I have a soft spot for go as well.
Transcript
Play full episode
