5 ms·
Show HN: How to Build a Stable Diffusion text-to-image generator (OSS)
- PTlightning 4y agoIs it open source?
- mikejulietbravo 4y agoYes it is. Editing title to reflect that.
- rasbt 4y agolove how fast it is
- mikejulietbravo 4y agoThank you! There will be more coming in exactly how we did that!
- osurits 4y agoHow long would it take to build?
- mikejulietbravo 4y agoIt 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
- PTlightning 4y agoFrom 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.
- mikejulietbravo 4y agoWe 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.
- theaniketmaurya 4y agothis scales with the amount of requests and can be used as a blueprint for deploying any AI model. Really cool!!