dilyabareeva/quanda
A toolkit for quantitative evaluation of data attribution methods.
This toolkit helps machine learning practitioners and researchers quantitatively assess how well different data attribution methods explain a model's predictions. You input your trained PyTorch model, its training data, and the attribution method you're interested in, and it outputs a detailed evaluation of that method's performance. It's designed for data scientists and ML researchers who need to understand and validate why their models make certain decisions.
No commits in the last 6 months. Available on PyPI.
Use this if you need to systematically compare and evaluate various training data attribution techniques for your PyTorch models to ensure they provide reliable and insightful explanations.
Not ideal if you are looking for a tool to generate data attributions themselves rather than evaluate existing methods, or if you are not working with PyTorch models.
Stars
57
Forks
—
Language
Jupyter Notebook
License
MIT
Last pushed
Jul 14, 2025
Commits (30d)
0
Dependencies
10
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/dilyabareeva/quanda"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
obss/sahi
Framework agnostic sliced/tiled inference + interactive ui + error analysis plots
tensorflow/tcav
Code for the TCAV ML interpretability project
MAIF/shapash
🔅 Shapash: User-friendly Explainability and Interpretability to Develop Reliable and Transparent...
TeamHG-Memex/eli5
A library for debugging/inspecting machine learning classifiers and explaining their predictions
csinva/imodels
Interpretable ML package 🔍 for concise, transparent, and accurate predictive modeling...