stanfordnlp/pyreft
A Python library implementing state-of-the-art Representation Fine-Tuning (ReFT) methods for pretrained language models.
Not currently ranked — collecting fresh signals.
star history
Pyreft is a Stanford NLP library implementing state-of-the-art Representation Fine-Tuning methods for pretrained language models. It enables training ReFT with any pretrained LMs on HuggingFace, supports configurable hyperparameters, and facilitates sharing results to HuggingFace. The library provides interventions on model representations as an alternative to LoRA and other PEFT approaches.
Frequently asked
- What is stanfordnlp/pyreft?
- A Python library implementing state-of-the-art Representation Fine-Tuning (ReFT) methods for pretrained language models.
- Is pyreft open source?
- Yes — stanfordnlp/pyreft is open source, released under the Apache-2.0 license.
- What language is pyreft written in?
- stanfordnlp/pyreft is primarily written in Python.
- How popular is pyreft?
- stanfordnlp/pyreft has 1.6k stars on GitHub.
- Where can I find pyreft?
- stanfordnlp/pyreft is on GitHub at https://github.com/stanfordnlp/pyreft.