https://github.com/MartinKondor/positron

Blazingly fast deep learning library for Python.
https://github.com/MartinKondor/positron

ai algorithms artificial-intelligence deep-learning deep-neural-networks machinelearning math mathematics neural-network neural-networks

Last synced: 2 months ago
JSON representation

Blazingly fast deep learning library for Python.

• Host: GitHub
• URL: https://github.com/MartinKondor/positron
• Owner: MartinKondor
• License: bsd-3-clause
• Created: 2021-09-01T14:55:02.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2022-08-05T07:44:56.000Z (over 2 years ago)
• Last Synced: 2024-08-02T05:23:10.671Z (5 months ago)
• Topics: ai, algorithms, artificial-intelligence, deep-learning, deep-neural-networks, machinelearning, math, mathematics, neural-network, neural-networks
• Language: Python
• Homepage:
• Size: 1.31 MB
• Stars: 2
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-blazingly-fast - positron - Blazingly fast deep learning library for Python. (Python)

README

        


    



# ⚡ Positron

[![Project Status](https://img.shields.io/badge/status-active-brightgreen.svg)](https://github.com/MartinKondor/positron/)

[![version](https://img.shields.io/badge/version-v1.1-brightgreen.svg)](https://github.com/MartinKondor/positron)

[![GitHub Issues](https://img.shields.io/github/issues/MartinKondor/positron.svg)](https://github.com/MartinKondor/positron/issues)

![Contributions welcome](https://img.shields.io/badge/contributions-welcome-brightgreen.svg)

[![License](https://img.shields.io/badge/license-BSD-brightgreen.svg)](https://opensource.org/licenses/BSD)

Blazingly fast Deep Learning library for Python.

The main focuses of Positron are

* Speed

* Functionality

* Scalability

* Understandability

_There is a [Medium Article](https://martinkondor.medium.com/positron-linear-algebra-library-for-python-8a3c5c3e1c00) written about this library, make sure to check it out!_

## Features

| File      | Description |

| --------- | ----------- |

| **example.py**      | **A simple neural network adaptation** working with this library.       |

| **deep.py**      | **Deep Learning** learning related functions: feedforward, backprop etc.       |

| **activ.py**      | Common Deep Learning **Activation Functions**.       |

| **maths.py**      | **Matrix operations**: inverse, determinant, adjungate etc.       |

| **prep.py**      | **Data Preprocessing Functions**: time stamp to date, date to time stamp etc.        |

| **loss.py**     | **Loss/cost Functions** and their derivatives: residual sum of squares, mean absolute error etc.        |

## Getting Started

### Prerequisites

* Python 3.8+

* Anaconda 4.9.2+ (optional)

* Python modules from the [requirements.txt](./requirements.txt) file

### Dependencies

For the dependencies, see the contents of the [requirements.txt](./requirements.txt) file, also let's note here that:

*This library should and never will depend on an other ML/DL library (like **keras**, **tensorflow** etc.), not even on **scikit-learn**.*

## Tests

To run tests on a file, run the test module directly, for example:

```$ python -m tests.loss```

Will test the ```positron/loss.py``` file.

### Examples

You can try to run an example file to see the library in work:

```$ python -m positron.loss```

## Contributing

This project is open for any kind of contribution from anyone.

### Steps

1. Fork this repository

2. Create a new branch (optional)

3. Clone it

4. Make your changes

5. Upload them

(5+1). Make a pull request here

## Authors

* **[Martin Kondor](https://github.com/MartinKondor)**

## License 

Copyright © 2022 Martin Kondor.

See the [LICENSE](./LICENSE) file for details.