Optimize your laptop for Machine Learning
Integrating your Mac or Linux machine to VESSL allows you to use your personal machines as a Kubernetes-backed single-node cluster and optimize your laptop for ML.
- Launch training jobs in seconds on VESSL's easy-to-use web interface or CLI, without writing YAML or scrappy scripts.
- Build a baseline model on your laptop and transition seamlessly to VESSL Cloud to scale your model.
- Keep track of your hardware and runtime environments together with
VESSL Clusters' single-line CLI command automatically checks, installs, and configures all dependencies such as Kubernetes, and connects your laptop to VESSL.
There is an ongoing 🔗 issue related to Kubernetes hostname containing capital letters. Please make sure your machine's hostname is in lowercase.
You should first have Docker, Helm, and VESSL SDK installed on your machine.
brew install docker
brew install helm
You should have Docker running in the background with your account logged in while using your laptop with VESSL Clusters.
Set up a VESSL environment on your laptop and grant access. Refer to our docs on VESSL Client for more detailed setup guides.
pip install vessl
The following single-line command connects your Mac.
vessl cluster create --name '[CLUSTER_NAME_HERE]' --mode single
The most common flags used with
vessl cluster createcommands are as follows. Check out our docs on VESSL CLI for additional flags.
--name— Define your cluster name
--mode single— Specifies that are installing a single-node cluster.
--help— See additional command options.
The command will automatically check dependencies and ask you to install Kubernetes. This process will take a few minutes. Proceed by entering
If you have Kubernetes installed on your machine, the command will then ask you to install VESSL agent on the Kubernetes cluster. Enter
By this point, you have successfully completed the integration.
Confirm your integration using the
listcommand which returns all the clusters available in your Organization.
vessl cluster list
Finally, try running a training job on your laptop. Your first run may take a few minutes to get the Docker images installed on your device.
Windows support for single-node cluster integration is currently in beta. Cluster usage and status monitoring may be limited for Windows machines.