entrypoint

The entrypoint key defines how your code should be run

Overview

alto.yml

...
entrypoint:
  type: script
  cmd: python <script name>.py --arg1 VALUE1
...

The entrypoint key should contain a set of key-value pairs that describe how your code should be accessed in your cloud infrastructure. Different entrypoints will have slightly different acceptable key-value pairs; however, they all must have type value.

If your entrypoint does not contain a type, then Alto will throw an error!

Key Specification

Different entrypoints have slightly different key-value specifications. We list these below:

type: script

• src: the directory containing the script to execute. This directory should be specified relative to the configuration YAML. This is an optional field.
• cmd: the command used to execute the script, e.g., python main.py.

Example:

alto.yml

...
  entrypoint:
    type: "script"
    cmd: "python main.py"
...

type: function

• src: the directory containing the script to execute. This directory should be specified relative to the configuration YAML. This is an optional field.
• cmd: the command used to execute the script, formatted as <module_name>.<function_name>
• kwargs: the function’s keyword arguments formatted as a set of key-value pairs.

Example:

alto.yml

...
  entrypoint:
    type: "function"
    cmd: "main.write_txt_file"
    kwargs:
      fname: "test.txt"
...

type: jupyter

• src: the directory containing the script to execute. This directory should be specified relative to the configuration YAML. This is an optional field.
• cmd: the notebook to execute, formatted as <notebook_name>.ipynb
• kernel: the Jupyter runtime environment. The default is python3
• params: the notebooks parameters formatted as a set of key-value pairs. Notebooks can be parametrized with Papermill! Check out the documentation here.

Example:

alto.yml

...
  entrypoint:
    type: "jupyter"
    cmd: "main.ipynb"
    params:
      alpha: 0.1
      ratio: 0.1
...