3 ms·

If you want to contribute to something like this, but that’s open-source, I started working on a similar app - SwiftUI + Apple’s CoreML Stable Diffusion here: h

by jitl 4y ago

If you want to contribute to something like this, but that’s open-source, I started working on a similar app - SwiftUI + Apple’s CoreML Stable Diffusion here: https://github.com/justjake/Gauss https://github.com/justjake/Gauss I’m currently working on downloadable models before I publish to the App Store & notarized builds to GitHub. A download manager adds complexity but should cut the initial app install size to ~50mb, and then distribute models via GitHub Releases.

swyx 4y ago
any idea if CoreML Stable Diffusion is appreciably faster or slower than regular-SD-on-Python? ( sorry if this is somehow well known, ive been trying to pay attention but havent seen anyone comment on what effect the implementation difference has had on the system performance)
- pavlov 4y ago
  It’s supposed to be roughly twice as fast: https://arstechnica.com/information-technology/2022/12/apple-slices-its-ai-image-synthesis-times-in-half-with-new-stable-diffusion-fix/ https://arstechnica.com/information-technology/2022/12/apple...
  - espadrine 4y ago
    They mention that in the M1 Ultra, it gets within a fraction of a second of the performance of an RTX 3060, which is very impressive! Sure, an M1 Ultra computer costs 3.6K€ higher than an RTX 3060 computer nowadays, but still, it gets close to the best performance on the market.
mark_l_watson 4y ago
Hey, thanks for doing this. I am looking around your git repo. I wrote a Swift book with a couple of CoreML examples. Apple did a fine job supporting deep learning models incorporated in macOS, iOS, and iPadOS apps.
codetrotter 4y ago
This is neat! Would you consider changing the license from GPLv3 to the ISC license? https://choosealicense.com/licenses/isc/ https://choosealicense.com/licenses/isc/
- jitl 4y ago
  Generally I use a libre license like GPLv3 for end-user apps to prevent others from quickly flipping them for profit, and open source license like MIT for libraries. If you need to release a closed-source fork I’m happy to sell you a custom license.
  - codetrotter 4y ago
    Understandable. Thank you :)

by jitl 4y ago

swyx 4y ago

any idea if CoreML Stable Diffusion is appreciably faster or slower than regular-SD-on-Python? ( sorry if this is somehow well known, ive been trying to pay attention but havent seen anyone comment on what effect the implementation difference has had on the system performance)

pavlov 4y ago
It’s supposed to be roughly twice as fast: https://arstechnica.com/information-technology/2022/12/apple-slices-its-ai-image-synthesis-times-in-half-with-new-stable-diffusion-fix/ https://arstechnica.com/information-technology/2022/12/apple...
- espadrine 4y ago
  They mention that in the M1 Ultra, it gets within a fraction of a second of the performance of an RTX 3060, which is very impressive! Sure, an M1 Ultra computer costs 3.6K€ higher than an RTX 3060 computer nowadays, but still, it gets close to the best performance on the market.

mark_l_watson 4y ago

Hey, thanks for doing this. I am looking around your git repo. I wrote a Swift book with a couple of CoreML examples. Apple did a fine job supporting deep learning models incorporated in macOS, iOS, and iPadOS apps.

codetrotter 4y ago

This is neat! Would you consider changing the license from GPLv3 to the ISC license? https://choosealicense.com/licenses/isc/ https://choosealicense.com/licenses/isc/

jitl 4y ago
Generally I use a libre license like GPLv3 for end-user apps to prevent others from quickly flipping them for profit, and open source license like MIT for libraries. If you need to release a closed-source fork I’m happy to sell you a custom license.
- codetrotter 4y ago
  Understandable. Thank you :)