acmi-lab/RLSbench

Code and results accompanying our paper titled RLSbench: Domain Adaptation under Relaxed Label Shift

38
/ 100
Emerging

This project helps machine learning researchers evaluate their domain adaptation algorithms when there are changes in the distribution of labels between training and real-world data. It provides a standardized dataset setup and code to simulate various label shifts and test different algorithms, from ERM variants to self-training and test-time adaptation methods. Researchers can input their datasets and choose simulation parameters to see how their models perform under relaxed label shift conditions.

No commits in the last 6 months.

Use this if you are an academic researcher in machine learning focused on domain adaptation and need to rigorously test algorithms under realistic label distribution shifts, particularly with limited computational resources.

Not ideal if you are a practitioner looking for a pre-built solution to apply domain adaptation to a specific business problem outside of a research context.

machine-learning-research domain-adaptation label-shift dataset-evaluation model-robustness
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

35

Forks

6

Language

Python

License

Apache-2.0

Last pushed

Jul 19, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/acmi-lab/RLSbench"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.