ModelTC/LightCompress
A compression toolkit for large AI models including LLMs, VLMs, and diffusion generators.
Not currently ranked — collecting fresh signals.
star history
LightCompress provides off-the-shelf compression techniques for AI generated content models. It implements state-of-the-art algorithms including quantization, pruning, token merging, and token reduction to reduce model size and improve inference efficiency while maintaining model performance. The toolkit supports popular models like LLMs, VLMs, and diffusion-based video generators, with papers published at EMNLP 2024 and AAAI 2026.
Frequently asked
- What is ModelTC/LightCompress?
- A compression toolkit for large AI models including LLMs, VLMs, and diffusion generators.
- Is LightCompress open source?
- Yes — ModelTC/LightCompress is open source, released under the Apache-2.0 license.
- What language is LightCompress written in?
- ModelTC/LightCompress is primarily written in Python.
- How popular is LightCompress?
- ModelTC/LightCompress has 735 stars on GitHub.
- Where can I find LightCompress?
- ModelTC/LightCompress is on GitHub at https://github.com/ModelTC/LightCompress.