#machine-learning #candle #abstraction-layer #neural-network #machine #learning

bin+lib candlelighter

A Keras like abstraction layer on top of the Rust ML framework candle

4 releases (breaking)

0.4.0 Apr 14, 2024
0.3.0 Mar 31, 2024
0.2.0 Mar 26, 2024
0.1.0 Jan 5, 2024

#444 in Hardware support

Download history 3/week @ 2024-02-14 7/week @ 2024-02-21 1/week @ 2024-02-28 55/week @ 2024-03-20 237/week @ 2024-03-27 66/week @ 2024-04-03 145/week @ 2024-04-10 15/week @ 2024-04-17

262 downloads per month

MPL-2.0 license

280KB
2.5K SLoC

Rust Lighter

This project was started as my RUST exercise to abstract the Rust minimalist ML framework Candle (https://github.com/huggingface/candle) and introduce a more convenient way of programming neural network machine learning models.

The behaviour is inspired by Python KERAS (https://keras.io) and the initial step based on the Rust-Keras-like code (https://github.com/AhmedBoin/Rust-Keras-Like).

So let's call the project Candle Lighter 🕯, because it helps to turn on the candle light and is even easier to implement.

Examples can be found below the lib/examples/ directory.

To use it as library just call 'cargo add candlelighter'

CONTRIBUTORS ARE HIGHLY WELCOME

Note: It is by far not production ready and is only used for own training purposes. No warranty and liability is given. I am a private person and not targeting any commercial benefits.

Supported Layer types

Meta Layer Type State Example
Sequential model -
- Feature scaling 🏃 DNN and TNN
- Dense DNN
- Convolution CNN
- Pooling -
- Normalization -
- Flatten -
- Recurrent RNN 1st throw
- Regulation -
- Feature embedding S2S 1st throw
- Attention 🏃 TNN 1st throw
- Mixture of Experts 🏃 ENN 1st throw
- Feature masking and -quantization 🏃 -
Parallel model (in sense of split) - 🏃 PNN 1st throw
Parallel model Merging 🏃 PNN 1st throw
- Model fine tuning 🏃 -

License

Tripple-licensed to be compatible with the Rust project and the source roots.

Licensed under the MPL 2.0, MIT license or the Apache license, Version 2.0 at your option.

Dependencies

~14–24MB
~361K SLoC