terarachang/DataICL
Data Valuation on In-Context Examples (ACL23)
This project helps machine learning practitioners improve the reliability of large language models (LLMs) when used for new tasks like sentiment analysis or question answering. It takes a dataset of examples and identifies the most impactful ones, producing a smaller, curated subset of examples. Data scientists and ML engineers can use this to make their LLMs perform more consistently.
No commits in the last 6 months.
Use this if you are using large language models via in-context learning and experiencing inconsistent or variable performance due to the choice of training examples.
Not ideal if you are looking for a solution that modifies the LLM's architecture or fine-tunes the model itself.
Stars
24
Forks
4
Language
Python
License
MIT
Category
Last pushed
Jan 12, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/terarachang/DataICL"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.