Hummingbot doesn't require much power, so some users have run successfully run multiple instances on a single Raspberry Pi. The following steps are for the Raspberry Pi but it should also work with any other device that uses the same ARM architecture.
Running Hummingbot on a Raspberry Pi or similar device has the same main benefit of running it on a cloud server: having a dedicated machine for Hummingbot. Raspberry Pi’s are relatively low cost, easy to set up, and, of course, don’t have the monthly charges associated with a cloud provider.
Download 64-bit OS¶
You can also choose between CLI (command line) and Desktop GUI versions but you'll get more performance with just using the CLI version.
Load the image file to your Raspberry Pi’s SD card¶
The Raspberry Pi has an easy to follow guide with alternatives on how to load the SD card with a Raspberry Pi OS from different operating systems.
Once the OS is installed and booted then you can follow the steps below to install Hummingbot using either Docker or Source
Install from Source¶
Update the repository and install important dependencies:
Restart the terminal:
Clone the Hummingbot repository:
If you need to switch branches (ie.
development branch) then after cloning the repository use the command
git checkout [branch_name] to switch branches. For example, to switch to the development branch use
git checkout development
Change directory into the Hummingbot folder:
If you are using Ubuntu 22.04 you'll need to go into the
./setup folder first and edit the
environment-linux-aarch64.yml file and change "cryptography==2.8" to "cryptography==3.1.1" before running the
./install command otherwise you'll get an error "could not build wheels for cryptography"
Clean your Hummingbot directory and then compile: