Skip to content

Jax implementation for the paper "Sampling-based inference for large linear models, with application to linearised Laplace"

License

Notifications You must be signed in to change notification settings

cambridge-mlg/sampled-laplace

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

sampled-laplace

This repository includes Jax code and experiments for the paper Sampling-based inference for large linear models, with application to linearised Laplace.

Run experiments

As an example, to run stochastic EM for a linearised Laplace model using the LeNetSmall architecture on the MNIST dataset on a Google Cloud TPU VM, run the following command:

python src/em_trainer.py --config experiments/mnist_gloud_em.py

Cloning the Repository

Since the repository uses submodules, it is recommended to clone the repository with the following command:

git clone --recursive sampled-laplace
git submodule update --init --recursive

Installation Instructions

sudo apt-get install python3.9-venv
python3.9 -m venv ~/.virtualenvs/sampled
source ~/.virtualenvs/sampled/bin/activate
pip install --upgrade pip
pip install -r requirements.txt
pip install -e .
python3.9 -m ipykernel install --user --name=sampled

About

Jax implementation for the paper "Sampling-based inference for large linear models, with application to linearised Laplace"

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages