Skip to content

Latest commit

 

History

History
43 lines (26 loc) · 1.09 KB

README.md

File metadata and controls

43 lines (26 loc) · 1.09 KB

SGHMC

Implementation of Stochastic Gradient Hamiltonian Monte Carlo.

For details, refer to original paper

This project is the final homework for duke STA663, contributed by Zining Ma ([email protected]) and Machao Deng ([email protected])

Repository contents

  • development/ : jupyter notebooks for package development

  • report/ : contents for the project reoprt

  • sghmc/ : source codes

Install

To install the package, run

$ git clone https://github.com/Senlody/SGHMC.git
$ cd SGHMC/sghmc
$ python setup.py install

You may need Administrator rights to install the package.

examples

To run examples, cd to SGHMC/sghmc/tests folder, and run one of the following

$ python mixnormal.py
$ python simpleU.py
$ python bnnMPG.py

A successful run of example script ends without throwing any error.

WARNNING: mixnormal.py and simpleU.py contains sghmc_chains that only works on linux. Running these examples on windows will cause errors.

For the details of the examples, refer to project report in SGHMC/report.