SDXL-Turbo: Faster, clearer open source local artificial intelligence image generator
Haebom
1
Stability AI has come up with another great thing. After accelerating the open source ecosystem for image generation with Stable Diffusion, they released SDXL-Turbo, which has the speed of SDXL, which was released in the middle of this year. It reduces the image generation time of existing SDXL by more than 30%. (It varies depending on the computer specifications.)
You can test it right here on Clipdrop above, or you can download the model from Hugging Face and run it locally. As always, it is open source and free. This is the first time that ADD (Adversarial Diffusion Distillation) technology has been released as an actual model and product outside of a paper.
adversarial_diffusion_distillation.pdf8.073205MB
It's amazing that they've achieved both speed and quality, but it's also impressive that they've reduced the number of steps in which the existing Diffuison methods process noise step-by-step from 100 to 50 in the recent SDXL, down to 1. 'This works?' It's amazing how these guys approached it this way. It seems that there will definitely be a lot of cost reductions and new attempts in image generation. Plus, they've also released a technology to convert images into video formats, so I wonder if the posts they create will spread on Tictok, YouTube, and Meta.
📚 Welcome to Haebom's archives. --- I post articles related to IT 💻, economy 💰, and humanities 🎭. If you are curious about my thoughts, perspectives or interests, please subscribe. Would you like to be notified when new articles are posted? 🔔 Yes, that means subscribe. haebom@kakao.com
