Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: How to Build a Stable Diffusion text-to-image generator (OSS) (lightning.ai)
8 points by mikejulietbravo on Oct 12, 2022 | hide | past | favorite | 9 comments



From the article it seems like there are a lot of steps involved in building a good app. It also doesn't necessarily talk on the model-building or tuning aspects of the process. There should probably be a bigger article to go with this on stable diffusion or building apps in general as right now it could be confusing to try and get started immediately without knowing the fundamentals of how these models work.


We can absolutely create a companion on stable diffusion. The model itself isn’t the focus of this though. We just wanted to show it was easy to build a fully functioning application that deploys a model, by using our framework. We didn’t touch the training piece intentionally.


this scales with the amount of requests and can be used as a blueprint for deploying any AI model. Really cool!!


Is it open source?


Yes it is. Editing title to reflect that.


How long would it take to build?


It depends how complicated you want it to be and how much you’ve already thought of. You could have a nice clean UI and working app in around a week if you had a clear vision for what you wanted


love how fast it is


Thank you! There will be more coming in exactly how we did that!




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: