dstack

CUDO Compute has partnered with dstack Sky the AI container orchestration tool.

dstack on CUDO Compute

dstack is an open-source alternative to Kubernetes and Slurm, built to streamline AI development and deployment across cloud providers. dstack provides the following capabilities:

  • Dev Environments: Quickly set up remote machines with your code and IDE for interactive coding. Ideal for testing code in your IDE or notebook before task scheduling or deployment.
  • Tasks: Schedule jobs or run web apps with configurable dependencies, resources, and more. Suited for training, fine-tuning, or app testing, with support for distributed clusters.
  • Services: Deploy scalable web apps or models as private or public endpoints, complete with dependency management, resource allocation, and scaling rules.
  • Fleets: Efficiently manage clusters and instances for cloud and on-prem resources, reusable across dev environments, tasks, and services.

dstack is available as a self-hosted open-source version and a hosted marketplace version.

Install dstack

Create a config file for CUDO and save it to ~/.dstack/server/config.yml. You will need to have created a project and know its id, and you will need to create an API key in the CUDO web console.

projects:
  - name: main
    backends:
      - type: cudo
        project_id: my-cudo-project
        creds:
          type: api_key
          api_key: 7487240a466624b48de22865589

Install dstack and start the server

pip install "dstack[all]" -U
dstack server

Try out a dstack deployment

Now you can create a project directory and initialize it for use with dstack:

mkdir quickstart && cd quickstart
dstack init

Create a yaml file .dstack.yaml inside your project directory to define your dstack deployment, here we are deploying a remote IDE (vscode):

type: dev-environment
# The name is optional, if not specified, generated randomly
name: vscode

python: '3.11'
# Uncomment to use a custom Docker image
# image: dstackai/base:py3.13-0.6-cuda-12.1

ide: vscode

# Use either spot or on-demand instances
spot_policy: auto

# Uncomment to request resources
# resources:
#  gpu: 24GB

Now apply the dstack configuration:

dstack apply -f .dstack.yml

For more

Please check the dstack documentation for the most up-to-date features.