nadir 0.0.3

Nadir is a library of bleeding-edge DL optimisers built for speed and functionality in PyTorch for researchers

Nadir

Nadir (pronounced nay-di-ah) is derived from the arabic word nazir, and means "the lowest point of a space". In optimisation problems, it is equivalent to the point of minimum. If you are a machine learning enthusiast, a data scientist or an AI practitioner, you know how important it is to use the best optimization algorithms to train your models. The purpose of this library is to help optimize machine learning models and enable them to reach the point of nadir in the appropriate context.

PyTorch is a popular machine learning framework that provides a flexible and efficient way of building and training deep neural networks. This library, Nadir, is built on top of PyTorch to provide high-performing general-purpose optimisation algorithms.

Table of Contents

• Nadir
• Table of Contents
• Installation
• Simple Usage
• Supported Optimisers
• Acknowledgements
• Citation

Installation

You can either choose to install from the PyPI index, in the following manner:

$ pip install nadir

or install from source, in the following manner:

$ pip install git+https://github.com/Dawn-Of-Eve/nadir.git

Note: Installing from source might lead to a breaking package. It is recommended that you install from PyPI itself.

Simple Usage

import nadir as nd

# some model setup here...
model = ...

# set up your Nadir optimiser
config = nd.SGDConfig(lr=learning_rate)
optimizer = nd.SGD(model.parameters(), config)

# Call the optimizer step
optimizer.step()

Supported Optimisers

Optimiser	Paper
SGD	https://paperswithcode.com/method/sgd
Momentum	https://paperswithcode.com/method/sgd-with-momentum
Adagrad	https://www.jmlr.org/papers/volume12/duchi11a/duchi11a.pdf
RMSProp	https://paperswithcode.com/method/rmsprop
Adam	https://arxiv.org/abs/1412.6980v9
Adamax	https://arxiv.org/abs/1412.6980v9
AdamW	https://arxiv.org/abs/1711.05101v3
Adadelta	https://arxiv.org/abs/1212.5701v1
AMSGrad	https://arxiv.org/abs/1904.09237v1
RAdam	https://arxiv.org/abs/1908.03265v4
Lion	https://arxiv.org/abs/2302.06675

Acknowledgements

We would like to thank all the amazing contributors of this project who spent so much effort making this repositary awesome! :heart:

Citation

You can use the Cite this repository button provided by Github or use the following bibtex:

@software{MinhasNadir,
    title        = {{Nadir: A Library for Bleeding-Edge Optimizers in PyTorch}},
    author       = {Minhas, Bhavnick and Kalathukunnel, Apsal},
    year         = 2023,
    month        = 3,
    version      = {0.0.2}
}