tomesd 0.1.3

Token Merging for Stable Diffusion

Stars: 1271, Watchers: 1271, Forks: 78, Open Issues: 25

The dbolya/tomesd repo was created 1 years ago and the last code push was 10 months ago.
The project is very popular with an impressive 1271 github stars!

How to Install tomesd

You can install tomesd using pip

pip install tomesd

or add it to a project with poetry

poetry add tomesd

Package Details

Author

Daniel Bolya

License

MIT

Homepage

https://github.com/dbolya/tomesd

PyPi:

https://pypi.org/project/tomesd/

GitHub Repo:

https://github.com/dbolya/tomesd

Related Packages

Errors

A list of common tomesd errors.

Code Examples

Here are some tomesd code examples and snippets.

GitHub Issues

The tomesd package has 25 open issues on GitHub

• https://git.mmaker.moe/mmaker/sd-webui-tome link down
• Support torch.compile
• Analytical Results
• [Feature Request] Separate merge ratio, downscaling and stride per module.
• Added onnx friendly merging
• did tomesd work with AMD Gpus "AMD RX vega 56" in Linux with Rocm?
• Idea for Improvement
• I'M A SMOOTHBRAIN, NEED HELP
• Out of Memory when used with Tiled VAE
• No Module named 'tomesd'
• Request: Add compatiblity with ONNX (OnnxStableDiffusionPipeline)
• no speed up

See more issues on GitHub