Models
WIP instructions ported from the old readme
To generate images using this plugin, you will first need a model. A few model presets are provided with the plugin and you can create your own by creating a StableDiffusionModelAsset though it is recommended that you duplicate one of the provided models as a starting point.
To download a model from huggingface.co, you will need an account and a https://huggingface.co/settings/tokens with read permissions.
You can use any diffusers based model from the diffusers category on huggingface.co. To choose a model, enter it in the format Username/Modelname. To use any of the default models RunwayML Stable Diffusion 1.5, Runway inpaiting, or CompViz Stable Diffusion 1.4, you will need to vist the respective model card page for each of these models and accept their usage agreement before downloading their weights otherwise you will receive an HTTP401 error.
The revision property will allow you to pick a specific branch of the model to download. I recommend the fp16 branch (if the model provides it) in case you have a limited amount of VRAM (under 6GB). You can also choose the level of precision you want to model to use where 16-bit floating point is the default.