Unsloth
Library for fine-tuning large language models significantly faster and with less memory than standard training approaches.
Overview
Unsloth optimizes the fine-tuning process for popular open-weight LLMs, claiming substantially faster training and lower memory use than standard Hugging Face training code through custom kernel optimizations, letting fine-tuning jobs that would need high-end multi-GPU setups run on more modest hardware. It supports popular open models and integrates with common fine-tuning workflows.
Unsloth is a fit for teams or individuals fine-tuning open-weight LLMs on limited hardware budgets, a specialized efficiency layer that complements DeepSpeed (also in this catalog) - both address training efficiency, with Unsloth focused specifically on making fine-tuning accessible on smaller hardware.
- Categories
- AI & Machine Learning
- Languages
- Python
- License
- Apache-2.0
Related Projects
DeepSpeed
Apache-2.0Deep learning optimization library from Microsoft for training and running very large models efficiently across multiple GPUs.
- AI & Machine Learning
CAMEL
Apache-2.0Open-source framework for research and building systems of multiple communicating, role-playing AI agents that collaborate autonomously.
- AI & Machine Learning
CrewAI
MITMulti-agent orchestration framework for building teams of collaborating, role-based AI agents that work together autonomously to complete complex tasks.
- AI & Machine Learning
DeepEval
Apache-2.0Open-source testing framework for evaluating LLM outputs, treating LLM application quality checks like conventional unit tests.
- AI & Machine Learning
Spotted an error? Suggest an edit on GitHub.