CITATION.cff

cff-version: 1.2.0
title: '"DySweep": Enhanced Weights and Biases Sweeps for Systematic Experimentation'
message: >-
  If you use this software, please cite it using the
  metadata from this file.
type: software
authors:
  - given-names: Hamid, Hamidreza
    family-names: Kamkari
    email: hamidrezakamkari@gmail.com
repository-code: "https://github.com/HamidrezaKmK/dysweep"
abstract: >-
  Dysweep is a powerful Python library designed to enhance the functionality of Weights and Biases sweeps. With Dysweep, conducting systematic and efficient deep learning experiments becomes a breeze. Its features include checkpointing for the Sweep Server, allowing for the resumption of specific runs, and the ability to run sweeps over hierarchies, eliminating the need for hard-coded selection between different classes. Inspired by DyPy, Dysweep provides a versatile configuration set, enabling the definition of experiments at any level of abstraction. Whether it's large-scale hyperparameter tuning or parallel execution of experiments, Dysweep empowers researchers with a systematic and streamlined approach to deep learning experimentation.
keywords:
  - Lazy Configurations
  - Weights and Biases
  - Sweep
  - Hierarchical Configuration
  - Deep Learning
  - Experiments
license: MIT
date-released: "2023-05-21"