Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/bkonkle/burn-transformers

A WIP port of the HuggingFace Transformers library for the Burn deep learning framework
https://github.com/bkonkle/burn-transformers

Last synced: 3 months ago
JSON representation

A WIP port of the HuggingFace Transformers library for the Burn deep learning framework

Awesome Lists containing this project

README 

        
# 🔥 Burn Transformers



`burn-transformers`



[![Crates.io](https://img.shields.io/crates/v/burn-transformers.svg)](https://crates.io/crates/burn-transformers)

[![Docs.rs](https://docs.rs/burn-transformers/badge.svg)](https://docs.rs/burn-transformers)

[![CI](https://github.com/bkonkle/burn-transformers/workflows/CI/badge.svg)](https://github.com/bkonkle/burn-transformers/actions)

[![Coverage Status](https://coveralls.io/repos/github/bkonkle/burn-transformers/badge.svg?branch=main)](https://coveralls.io/github/bkonkle/burn-transformers?branch=main)



## State-of-the-art Machine Learning for the Burn deep learning library in Rust



Burn Transformers is an early-stage work-in-progress port of the [Huggingface Transformers](https://huggingface.co/docs/transformers/index) library from Python to Rust. It endeavors to provide many pretrained models to perform tasks on different modalities such as text, vision, and audio.



🔥 Transformers provides APIs to quickly download and use those pretrained models on a given input, taking advantage of the great tools Huggingface has provided as part of [Candle](https://github.com/huggingface/candle) - their minimalist ML framework for Rust - and fine-tune them on your own datasets using the excellent [Burn](https://burn.dev/) deep learning library in Rust.



⚠️ Alpha Disclaimer

NOTE: This library is in early development, and the API may shift rapidly as it evolves. Be advised that this is not yet recommended for Production use.



## Features



| Pipeline | Model | Status |

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

| Text Classification | BERT | ✅ |

| | RoBERTa | ❌ |

| | ALBERT | ❌ |

| | DistilBERT | ❌ |

| | Others... | ❌ |

| Token Classification | BERT | ✅ |

| Zero Shot Classification | ... | ❌ |

| Text Generation | ... | ❌ |

| Summarization | ... | ❌ |

| Image / Video / Audio Classification | ... | ❌ |

| Question Answering | ... | ❌ |

| Others... | ... | ❌ |



## Installation



### Cargo



Install Rust and Cargo by following [this guide](https://www.rust-lang.org/tools/install).



Add this to your Cargo.toml:



```toml

    [dependencies]

    bert-transformers = { git = "https://github.com/bkonkle/burn-transformers" }

```



Once `bert-burn` is published to crates.io, this project will be able to be published there as well.



## Development



To set up a development environment to build this project, first install some helpful tools.



### Clippy



For helpful linting rools, install [Clippy](https://github.com/rust-lang/rust-clippy)



Run it with `cargo`:



```sh

cargo clippy --fix

```



If you're using VS Code, configure the `rust-analyzer` plugin to use it (in _settings.json_):



```json

{

    "rust-analyzer.checkOnSave.command": "clippy"

}

```



### pre-commit



Install pre-commit to automatically set up Git hook scripts.



In Ubuntu, the package to install is `pre-commit`:



```sh

sudo apt install pre-commit

```



On Mac with Homebrew, the package is also `pre-commit`:



```sh

brew install pre-commit

```



### libclang



The `cargo-spellcheck` utility depends on [`libclang`](https://clang.llvm.org/doxygen/group__CINDEX.html).



In Ubuntu, the package to install is `libclang-dev`:



```sh

sudo apt install libclang-dev

```



### Cargo Make



To use build scripts from the _Makefile.toml_, install Cargo Make:



```sh

cargo install cargo-make

```



Run "setup" to install some tooling dependencies:



```sh

cargo make setup

```



### Update Dependencies



First, install the `outdated` command for `cargo`:



```sh

cargo install cargo-outdated

```



Then, update and check for any major dependency changes:



```sh

cargo update

cargo outdated

```



## License



Licensed under the MIT license ([LICENSE](LICENSE) or ).