endpoint from a commandΒΆ

endpoint using a raw start command that works with any programming language or framework.

  • name: name of the deployment endpoint, this will be the final part of the URL
  • image: the Docker image that will be used as the deployment environment
  • server-command: command that runs a HTTP server
  • port: (optional) where should Valohai expect to find the web server, defaults to 8000. Note the server must be bound to (all interfaces), not only localhost.
  • description: (optional) more detailed human-readable description of the endpoint
  • files: (optional) files that will be loaded into the image, for example the trained model. The files will be in the same directory as your code, modified by the path property.
- endpoint:
    name: server-endpoint
    image: python:3.6
    server-command: python run_server.py
    port: 1453
      - name: model
        description: Model output file from TensorFlow
        path: model.pb

Valohai automatically runs pip install --user -r requirements.txt in case you have additional dependencies defined in requirements.txt in your project root. Commands from installed packages end up in ~/.local/bin, following the standards for pip --user installation.