SPEAR: Programmatically label and build training data quickly.
Self-improving weak supervision: an LLM writes Python labeling functions, an AST sandbox validates them, a greedy F1 ratchet keeps only winners
Official implementation of our paper: "Structured Exploration and Exploitation of Label Functions for Automated Data Annotation"
Python weak-supervision demo using rule-based labeling functions, majority-vote aggregation, automated testing, Docker, and GitHub Actions.
Python weak-supervision demo using rule-based labeling functions, majority-vote aggregation, automated testing, Docker, and GitHub Actions.
Add a description, image, and links to the labeling-functions topic page so that developers can more easily learn about it.
To associate your repository with the labeling-functions topic, visit your repo's landing page and select "manage topics."