What’s the base foundation model you’re using for generating assets?
We are currently using Stable Diffusion 2.1 and Stable Diffusion XL and we fine tune it with our customer's assets. We call this “Style Training".
Where is the server?
Netherlands, EU.
Does Layer depend on a particular LLM?
We designed the product such that we can select different foundation models depending on the need.
Do you pass any data to Stability AI?
None of the customers input whether they are prompts or artwork is shared with Stability AI.
Creating a style means training AI, right?
Yes. A style at Layer AI is considered a custom trained AI model.
Can I upload my own AI models to Layer?
Yes, you can upload your own AI models to Layer. You can learn more about it here.
Does Layer allow seed editing?
Seed editing - it's an advanced feature and in most cases it's best to leave it as "random". However, you can for example, get close variations by changing the prompt but keeping the seed the same.