Jax Influence

Scalable implementation of Influence Functions in JaX.

Implementation of the algorithms in Scaling Up Influence Functions (AAAI 2022) for efficient calculation of Influence Functions.

Installation

manual installation

Download the repo and set up a Python environment:

git clone https://github.com/google-research/jax-influence ~/jax-influence


cd ~/jax-influence
conda env create -f environment.yml
conda activate jax-influence

pip installation

pip install jax-influence

The pip installation will install all necessary prerequisite packages, however you might want to install the most appropriate version of jax and jaxlib in case you use GPUs/TPUs.

Documentation

An end-to-end example of using the library can be found in examples/colab/mnist_tutorial.ipynb. We plan to add more examples in the future.

Disclaimer

This is not an official Google product.

Jax Influence is a research project, and under active development by a small team; we'd love your suggestions and feedback - drop us a line in the issues.