Automatic build of preference datasets for LLM alignment
Project description
⚗️ distilabel
AI feedback framework
What's distilabel
distilabel is a framework for building datasets with AI for LLM fine-tuning, evaluation, and alignment.
It gives an extensible and scalable framework for AI Feedback (AIF) and model distillation, building on state of the art approaches such as UltraFeedback or JudgeLM.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
distilabel-0.1.0rc0.tar.gz
(1.2 MB
view hashes)
Built Distribution
Close
Hashes for distilabel-0.1.0rc0-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 84d2303a704a139a51d6050edd08fcb16824c1c29a1fd557f0409f015d8e1dcc |
|
MD5 | 7b87bcd7e5f85c6160f932954030bbe5 |
|
BLAKE2b-256 | bfb57cadd6a510e748705d88c966e471d6ae938f7727fbe52456a4771e2e9b1f |