Install the cli
To begin, you need to install the Unweave Command Line Interface (CLI) by following these steps:
brew tap unweave/unweave brew install unweave
Next, log in to your Unweave account from the terminal using:
Create a new project
In the next version of our CLI, you'll be able to create projects from your command line directly!
To create a new project, navigate to the Unweave dashboard by running the command:
Click the big blue button to create your project.
Link your project
Once you've created your project, you need to link it to your local directory. Linking a project associates a directory on your computer to the Unweave project in your account. Once linked, any commands you run in this directory, or any of its subdirectories, will be run in the context of the project.
unweave link <unweave-username>/<project-name>
Unweave ships with a super handy command to launch VS Code in the context of your project.
unweave code --gpu-type rtx_5000 --image pytorch/pytorch:2.0.0-cuda11.7-cudnn8-devel
This will launch VS Code, setup SSH access, and sync your local directory onto the VM at the
/home/unweave path. The example above uses a VM with the RTX 5000 GPU type and the
pytorch/pytorch:2.0.0-cuda11.7-cudnn8-devel Docker image as the base.
Congratulations! You're now ready to start working with your GPU-accelerated environment in VS Code.
You can check the status of any running sessions using the
unweave ls command:
This command will display a list of all your running sessions, along with relevant information such as session ID, machine type. Remember to always clean up and terminate your sessions when you are finished to avoid unnecessary charges.
Once you are done using the GPU-powered machine, remember to power it down to avoid unnecessary costs. Run the following command:
The Unweave CLI offers various powerful features. To explore more options, use the following command: