Repository files navigation

This repo contains a torchsharp implementation forstable diffusion 2 model.

To run the stable diffusion 2 model on your local machine, the following prerequisites are required:

• dotnet 6
• git lfs, this is to download the model file from hugging face

To get stable-diffusion-2 model weight, run the following command to download model weight from huggingface. Be sure to have git lfs installed.

git clone https://huggingface.co/stabilityai/stable-diffusion-2

Clone this repo and replace themodelFolder folder with where you download huggingface model weight inProgram.cs

Then run the following command to start the model:

dotnet run

(a photo of an astronaut riding a horse on mars)

You can load fp16 model weight by settingdtype to ScalarType.Float16 inProgram.cs. The inference on fp16 model weight is faster and more GPU memory efficient.

• Add support for loading fp16 model weight

• Torchsharp-llama: A torchsharp implementation for llama 2 model
• Torchsharp-phi: A torchsharp implementation for phi model