Skip to content

Code for Hyperparameter Optimization project for "Big Data & Quantum Mechanics" (Andrew Medford)

Notifications You must be signed in to change notification settings

Arrrlex/bdqm-hyperparam-tuning

Repository files navigation

AmpOpt

Contents

Introduction

Code for Hyperparameter Optimization project for "Big Data & Quantum Mechanics" (Andrew Medford).

Structure

  • jobs contains .pbs files defining jobs to run on PACE
  • src/ampopt contains code for preprocessing, running hyperparameter optimization jobs, and analyzing results
  • src/ampopt_cli exposes the functionality of ampopt via a CLI
  • data contains the source .traj files, as well as the preprocessed .lmdb files (once ampopt preprocess has been run).
  • env_gpu.yml and env_cpu.yml contain dependencies for machines with and without (respectively) GPU.
  • setup-session.sh is a convenience script for PACE jobs
  • setup.py defines how ampopt Python package and command are installed.

Installation

For setup and installation instructions, please refer to SETUP.md.

Usage

For notebook tutorials, see:

For more comprehensive (but drier) instructions, please refer to USAGE.md.

Finally, for full explanation of all options, see ampopt --help or ampopt <command> --help.

About

Code for Hyperparameter Optimization project for "Big Data & Quantum Mechanics" (Andrew Medford)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published