IBM/Dromedary
IBM's self-aligned language model trained with minimal human supervision using principle-driven self-alignment techniques.
Not currently ranked — collecting fresh signals.
star history
Dromedary is an open-source project implementing self-alignment for language models to make them helpful, ethical, and reliable with minimal human supervision. The approach involves a principle-driven self-alignment pipeline that trains LLMs to follow guidelines while maintaining helpfulness. Dromedary-2 extends this with an improved two-stage process using diverse user prompts and a general-specific-general response style, building on the LLaMA-2 base model.
Frequently asked
- What is IBM/Dromedary?
- IBM's self-aligned language model trained with minimal human supervision using principle-driven self-alignment techniques.
- Is Dromedary open source?
- Yes — IBM/Dromedary is open source, released under the GPL-3.0 license.
- What language is Dromedary written in?
- IBM/Dromedary is primarily written in Python.
- How popular is Dromedary?
- IBM/Dromedary has 1.1k stars on GitHub.
- Where can I find Dromedary?
- IBM/Dromedary is on GitHub at https://github.com/IBM/Dromedary.