This blog post will be updated based on progress.
While playing around with the Text-To-Image Stable Diffusion checkpoint back in 2023, I start thinking how I could integrate it into my work, primarily for creating design references.
So the prefect solution for me was a LoRA, a small-sized model with my own dataset for interaction with checkpoint. Thats how project hython was created.
The Project Hython is currently under active development, with a release date around Q4 2024.
Goal
The primary objective was to develop a user-friendly tool that allows me to create design inspirations easily. It needed to seamlessly integrate with Stable Diffusion, given my focus on web development and design, particularly in product and UI/UX. Therefore, it was essential for the tool to adhere to the philosophy of Tailwind UI, which I frequently use.
The main idea for custom UI is to provide to the user a simply and future-rich interface to easily change current prompt key words, such a page, components on it and colors.
Roadmap
The current roadmap have 3 big releases for this project
- Release a LoRA for Stable Diffusion 1.5
- Integrate LoRA to use with ControlNet to crate some more specific components
- Create my own UI to use with a model