Skip to content

Latest commit

 

History

History
28 lines (21 loc) · 1.78 KB

CONTRIBUTING.md

File metadata and controls

28 lines (21 loc) · 1.78 KB
Raw

This is a community project, so please feel free to contribute (and use in your project)!

Discord Server

t2i

Contributing

We'd love to accept code contributions. Please feel free to drop by our Discord community!

📢 We're looking for code contributions for these features (or anything else you'd like to work on):

  • CodeFormer upscaling (please maintain the required copyright notices).
  • Using custom Textual Inversion embeddings.
  • Seamless tiling.
  • Outpainting.
  • Mac support.
  • AMD support.
  • Allow other samplers for img2img (instead of only DDIM).

Setting up a developer environment

If you're on Windows or Linux, please install CUDA-compatible torch and torchvision (if you haven't already) using: pip install torch torchvision --extra-index-url https://download.pytorch.org/whl/cu116

  1. Create a fork of this repository using https://github.com/easydiffusion/sdkit/fork

  2. Copy the link to your repository using the "Code" button in your repository. Important: Use YOUR fork for this. image

  3. Install an editable copy of your fork using: pip install -e [email protected]:YOUR_FORK_NAME/sdkit.git#egg=sdkit or pip install -e git+https://github.com/YOUR_FORK_NAME/sdkit.git#egg=sdkit. Replace YOUR_FORK_NAME with what your fork has.

  4. This will checkout a working copy of sdkit, along with any required dependencies.

  5. You can now edit the code and submit pull-requests for any changes you'd like to contribute.