Skip to content

colmont/NASLIB

Repository files navigation

Using the NASLib benchmark

Introduction

This is a fork from the original NASLib repository, with minimal changes to incorporate the proposed heat kernel, as well as run experiments more smoothly.

Installation

To install the required dependencies, please refer to the original README.md file. Our fork only needs a few additional dependencies that can be installed as follows: pip install simplejson hydra-core wandb termcolor.

Usage

To run the GP regression, execute the following command:

python naslib/runners/predictors/runner.py

which will use the values from the associated config file. All experiments will be automatically saved using Weights & Biases, and large parallel experiments can be run using the associated bash script, which contains all the experiments from the paper.

To run the BO loop, execute the following command:

python naslib/runners/nas_predictors/runner.py

which will use the values from the associated config file. All experiments will be automatically saved using Weights & Biases, and large parallel experiments can be run using the associated bash script, which contains all the experiments from the paper.

About

No description, website, or topics provided.

Resources

License

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 21