Intel Optimized Stable Diffusion

Ankur Singh

Ankur Singh

San Jose, California

1 0
  • 0 Collaborators

Stable Diffusion has changed the complete landscape of Generative Modeling. In this project, I plan to implement Stable Diffusion from scratch. The primary focus would be to use Intel Optimized libraries and packages. ...learn more

Project status: Concept

oneAPI, Artificial Intelligence

Intel Technologies
oneAPI, Intel NUC, Intel Python, MKL, Intel CPU, Intel Opt ML/DL Framework

Code Samples [1]

Overview / Usage

Since its release, Stable Diffusion has taken the world by storm. First, its open source (unlike DALLE-2 and ImageGen). Second, its very light weight. Its so light that you can run the inference on your laptop.

You may ask, what is Stable Diffusion? It is a latent-diffusion model. It is primarily used to generate detailed images conditioned on text descriptions called prompt, though it can also be applied to other tasks such as inpainting, outpainting, and generating image2image translations guided by a text prompt.

I am currently doing the "Fastai 2022 Course: From Foundations to Stable Diffusion". By the end of this course, I will be able to build Stable Diffusion model from scratch. In this project, I want to use Intel optimization to speed up the inference on Intel hardware. This will make Stable Diffusion more accessible to the masses.

I am super excited to see what people will do with it !

Repository

https://github.com/Ankur-singh/stable-diffusion

Comments (0)