Setup and Installation
If you just want to try out Rune without installing anything, head over to the tutorial for links to our online playground.
To use Rune, you will need the following:
- the
rune
CLI tool - for building and running Runes from the command-line - the Rust toolchain - for doing the actual compiling to WebAssembly
- The
tinyverseml/rune-serve
Docker image - lets you use your Rune from the Runic Mobile app without needing to upload to a registry
Rune tools work with the following systems:
- Linux
- MacOS
- Note: it is not advised to use Rune on M1 devices. We have received several reports of M1 bugs in Docker and TensorFlow.
- Windows (using WSL or the Docker images)
GitHub Releases​
Pre-compiled bundles containing the rune
CLI tool and examples are available
on our GitHub Releases page as the rune.$target.zip
asset,
where $target
is the target triple corresponding to your OS
(e.g. x86_64-apple-darwin
for x86 MacOS).
For those who like to live on the bleeding edge, our nightly release
contains the same assets but automatically generated from master
every 24
hours.
Once you have downloaded the pre-compiled bundles you will need to put the
rune
executable somewhere on your $PATH
. Typically, this will be ~/bin
or ~/.local/bin
.
$ curl -L -O https://github.com/hotg-ai/rune/releases/latest/download/rune.x86_64-unknown-linux-gnu.zip
$ unzip rune.x86_64-unknown-linux-gnu.zip
$ mkdir -p ~/.local/bin
$ cp rune ~/.local/bin/rune
As a security measure, when you download a program from the internet some
operating systems will remove their executable flag. You may need to chmod
the
rune
binary to make it executable.
$ chmod +x ~/.local/bin/rune
$ rune --version
rune 0.10.0
Docker​
We have a docker image that comes with the rune
CLI and Rust toolchain
already installed.
$ docker pull tinyverseml/rune-cli
$ docker pull tinyverseml/rune-serve
Compile From Source​
The rune
CLI tool can be installed using cargo install
just like a normal
Rust tool.
As well as requiring the Rust toolchain, the Rune project uses several native libraries so you will need to install the following build dependencies:
bindgen
(cargo install bindgen
)- Clang and LLVM
- CMake
- Docker (Linux only)
- Bazel (Windows and MacOS)
You can then install the Rune CLI from crates.io:
$ cargo install hotg-rune-cli
Alternatively, if you want to live on the bleeding edge you can install it
directly from [the GitHub repository][rune-repo] (either master
or a tagged
version):
$ cargo install --git https://github.com/hotg-ai/rune hotg-rune-cli
$ cargo install --git https://github.com/hotg-ai/rune hotg-rune-cli --rev v0.4.0
Compiling TensorFlow Lite from source can take quite a while, especially on less powerful machines, so do not panic if the build seems "stuck" at this stage:
building [=======================> ] 400/406: hotg-runecoral(build)