In AI Workbench, a Location refers to the machine where your Projects are stored and run. This can be a local machine, such as the computer you’re using right now, or a remote machine, like a virtual machine (VM) or server that you connect to over a network.
The Command Line Interface (CLI) uses the term “context” instead of “location”. However, the Desktop app uses the term “location”. In the future, we plan to migrate everything to use the term “location” for consistency.
During installation, AI Workbench creates a Local Location by default. This Location allows you to easily access and work with your local files, running the desktop app or CLI and service all on the same system.
Note that you can only have one Local Location, and you cannot delete it. However, you can modify its description to better suit your needs.
With AI Workbench, you can create remote locations at any time. You can have multiple remote locations, each connected to a different system, giving you the flexibility to work on various types of systems with different compute resources.
Remote locations work by establishing an SSH tunnel between your client (either the Desktop app or CLI) and the AI Workbench service on the remote system. This secure connection allows you to access and work on your projects as if you were locally connected to the remote system.
Deleting a remote location in AI Workbench simply removes the connection information, and does not affect the remote system itself. This means that any data or installations on the remote system remain intact, and you can add the remote location again at any time.
Remote locations are particularly useful when you need to access more compute power or GPUs than what’s available on your local machine. By connecting to a remote system with more resources, you can run your workloads faster and more efficiently because AI Workbench handles the labor required to move your workaround.
To synchronize your work across different locations, we use Git Server. This allows you to commit and push your changes from the Project on one Location, and then pull the changes in the other Location.
If you’re using the desktop app, you can easily move a project you’re working on to another location with just a single click. Simply click the “Open In” button, and you’ll be able to select the desired location from a list.
If you prefer using the Command Line Interface (CLI), you can open a new terminal, activate the location, and then manually clone the repository using the nvwb clone command.
By synchronizing your work through a Git Server, you can easily switch between different locations and continue working on your projects without any hassle.