A command line app for downloading Huggingface repositories with Rust.
RustyFace does not require installing additional dependencies such as git
or git lfs
etc. It aims to be lightweight and portable.
In addition to that, RustyFace is friendly to users who live in Mainland China, where HuggingFace accessibility is unstable, as this CLI app adopted a mirror that can be accessed globally.
The mirror site used in this project is hf-mirror.com
First, you need to have Rust installed. For those new to Rust, please refer to the official installation guide.
You don't need to install Rust if you download the corresponding binaries to your platform from the Release section. That way, you can just type this command to download Huggingface repositories:
rustyface_windows_x86 --repository sentence-transformers/all-MiniLM-L6-v2 --tasks 4
rustyface_windows_x86
is the binary file name that you have downloaded from the Release section.--repository
is followed by therepo_id
of the repository that you want to download from HuggingFace.--tasks
is followed by the number of concurrent downloads. For example, 4 means downloading 4 files at once. It is recommended to use a lower number if your network conditions do not support higher concurrency.
If you would like to reuse the program, it is recommended to install RustyFace onto your system rather than using the binaries. Here is how you can do it.
On Linux and macOS:
curl https://sh.rustup.rs -sSf | sh
On Windows, you can download the installation executable via this link: https://win.rustup.rs/
After done installing Rust, just type this to your terminal:
cargo install rustyface
Try RustyFace out with this simple command line:
rustyface --repository sentence-transformers/all-MiniLM-L6-v2 --tasks 4
--repository
is followed by therepo_id
of the repository that you want to download from HuggingFace.--tasks
is followed by the number of concurrent downloads. For example, 4 means downloading 4 files at once. It is recommended to use a lower number if your network conditions do not support higher concurrency.
Any participation is appreciated! Feel free to submit an issue, discussion or pull request. You can find me on WeChat: baoxinyu2007
or Discord: https://discord.gg/UYfZeuPy
This project is licensed under the MIT License. See the LICENSE file for details.