UCSC-REAL/DS2

[ICLR 2025] Official implementation of paper "Improving Data Efficiency via Curating LLM-Driven Rating Systems"

36
/ 100
Emerging

DS2 helps researchers and engineers improve the quality of datasets used to train large language models (LLMs). It takes raw, LLM-generated quality scores for data samples, corrects common errors in these scores, and then curates a final dataset that is both high-quality and diverse. This results in more efficient and effective LLM instruction tuning.

101 stars. No commits in the last 6 months.

Use this if you are an AI researcher or machine learning engineer struggling with noisy or inconsistent quality ratings when preparing training data for your LLMs.

Not ideal if you need a tool for general data cleaning or a system for human-in-the-loop data labeling.

LLM training data curation machine learning engineering dataset optimization AI research
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 11 / 25

How are scores calculated?

Stars

101

Forks

9

Language

Python

License

Apache-2.0

Last pushed

Mar 24, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/UCSC-REAL/DS2"

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