Show HN: Psychedelia fine tuned Stable Diffusion model(aiart.fm)
aiart.fm
Show HN: Psychedelia fine tuned Stable Diffusion model
https://aiart.fm/
5 comments
Thank you for sharing!
However, I think that the landing page would benefit from some improvements.
I see "Join the AI Art revolution. Unleash your creativity and discover the beauty of AI-generated art!". Well - it could be more distinguishable from hundreds (or thousands?) websites. Neither carries a hint on why it is different or better from other products nor acts as a hook. Then I click, and it asks me to sign in, while I still see no value proposition. Then I close it.
The images are indeed beautiful, but it does not suffice.
> Each image gets tagged and stored in a database so you can go back and look at what prompts, seeds, step, models people have used to make nice images. There is also a rudeimntary voting system and the homepage is sorted by highest rated.
This is interesting! If you expose it (also with an example), I would be more inclined to test your website.
However, I think that the landing page would benefit from some improvements.
I see "Join the AI Art revolution. Unleash your creativity and discover the beauty of AI-generated art!". Well - it could be more distinguishable from hundreds (or thousands?) websites. Neither carries a hint on why it is different or better from other products nor acts as a hook. Then I click, and it asks me to sign in, while I still see no value proposition. Then I close it.
The images are indeed beautiful, but it does not suffice.
> Each image gets tagged and stored in a database so you can go back and look at what prompts, seeds, step, models people have used to make nice images. There is also a rudeimntary voting system and the homepage is sorted by highest rated.
This is interesting! If you expose it (also with an example), I would be more inclined to test your website.
Hey I appreciate the feedback! Thanks!
Ya... It was built for a hackathon in a few days this weekend so there is still quite a bit left to be desired. Perhaps my excitement led me to share it too hastily.
I'm definitely shipping the aditional features like image search using CLIP, and tagged/category based filtering soon. Thanks again.
Ya... It was built for a hackathon in a few days this weekend so there is still quite a bit left to be desired. Perhaps my excitement led me to share it too hastily.
I'm definitely shipping the aditional features like image search using CLIP, and tagged/category based filtering soon. Thanks again.
I think that it is great you shared! :)
My point was not on features, but a taglines or anything that announces these features. Otherwise, it won't matter at all how many cool features you have.
My point was not on features, but a taglines or anything that announces these features. Otherwise, it won't matter at all how many cool features you have.
I did a hackathon this weekend for leapml.dev and built out a site that lets you generate images with stable diffusion in the browser:
https://aiart.fm
Each image gets tagged and stored in a database so you can go back and look at what prompts, seeds, step, models people have used to make nice images. There is also a rudeimntary voting system and the homepage is sorted by highest rated.
But the most exciting bit is I added my own custom fine tuned model to it last night that I'm calling "Psychedelia". All of the top images one the homepage are using this model.
It seems to work best with basic prompts and I often still get artifacts like double heads, but I'm quite happy with the initial results and will continue to fine tune.
The site is built with NextJS, TailwindCSS and Supabase and you can login and generate images for free.
Hope you enjoy!
Thanks, Alex