https://github.com/amanpriyanshu/dp-hyperparamtuning
DP-HyperparamTuning offers an array of tools for fast and easy hypertuning of various hyperparameters for the DP-SGD algorithm.
https://github.com/amanpriyanshu/dp-hyperparamtuning
automl hyperparameter-optimization machine-learning pytorch reinforcement-learning
Last synced: 7 months ago
JSON representation
DP-HyperparamTuning offers an array of tools for fast and easy hypertuning of various hyperparameters for the DP-SGD algorithm.
- Host: GitHub
- URL: https://github.com/amanpriyanshu/dp-hyperparamtuning
- Owner: AmanPriyanshu
- License: mit
- Created: 2021-07-30T19:21:04.000Z (about 4 years ago)
- Default Branch: main
- Last Pushed: 2021-09-27T09:20:22.000Z (about 4 years ago)
- Last Synced: 2025-03-18T14:23:00.112Z (7 months ago)
- Topics: automl, hyperparameter-optimization, machine-learning, pytorch, reinforcement-learning
- Language: Jupyter Notebook
- Homepage:
- Size: 88.9 KB
- Stars: 24
- Watchers: 1
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
Awesome Lists containing this project
README
# DP-HyperparamTuning
The offical repository for all algorithms and code for the [Efficient Hyperparameter Optimization for Differentially Private Deep Learning](https://arxiv.org/abs/2108.03888) accepted at [PPML Workshop @ ACM-CCS'2021](https://ppml-workshop.github.io/).
A streamlined and basic implementation for all modules presented is available at:
* [GitHub Repo](https://github.com/AmanPriyanshu/DP-HyperparamTuning)
* [GitHub Notebook](https://github.com/AmanPriyanshu/DP-HyperparamTuning/blob/main/RL_DP_Demo.ipynb)
`Note:` [Colab Demo for the same](https://colab.research.google.com/github/AmanPriyanshu/DP-HyperparamTuning/blob/main/RL_DP_Demo.ipynb)
# Implementation:
## Imports:
```python
from DP_HyperparamTuning.experiment.train_single_model import Experiment
from DP_HyperparamTuning.algorithms.bayesian_optimization import Bayesian
from DP_HyperparamTuning.algorithms.grid_search_algorithm import GridSearch
from DP_HyperparamTuning.algorithms.evolutionary_optimization import EvolutionaryOptimization
from DP_HyperparamTuning.algorithms.reinforcement_learning_optimization import RLOptimization
```
## Running Given Modules:
```python
e = Experiment(get_model, criterion, train_dataset, test_dataset)
b = Bayesian(e.run_experiment, calculate_reward, num_limit, search_space_nm=search_space_nm, search_space_lr=search_space_nm)
```
Where, `get_model`, `calculate_reward` are functions, and `criterion` and `train_dataset, test_dataset` which are `` and `torch.utils.data.Dataset` respectively.
# Contributing
When contributing to this repository, please first discuss the change you wish to make via issue,
email, or any other method with the owners of this repository before making a change. We also make
available a [CONTRIBUTING.md](https://github.com/AmanPriyanshu/DP-HyperparamTuning/blob/main/CONTRIBUTING.md) and [CODE_OF_CONDUCT.md](https://github.com/AmanPriyanshu/DP-HyperparamTuning/blob/main/CODE_OF_CONDUCT.md) for easy communication and quick issue resolution.
# Paper Citation:
```bib
@misc{priyanshu2021efficient,
title={Efficient Hyperparameter Optimization for Differentially Private Deep Learning},
author={Aman Priyanshu and Rakshit Naidu and Fatemehsadat Mireshghallah and Mohammad Malekzadeh},
year={2021},
eprint={2108.03888},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
```