Pipeline environment

When setting up Pipelines, you have the option to use either a Studio environment or a container as the source for the pipeline's execution environment.

Using a Studio

In the example below, the pipeline utilizes the Studio environment to manage the execution of a job.

Initialize the PipelineWe start by importing necessary classes from the lightning_sdk and create a Studio named 'studio_env'.Define PipelineNext, we create a pipeline instance named 'pipeline_from_studio' that will inherit from the Studio environment. we define the steps that the pipeline should execute.Define stepsIn this example, we are running a single Studio Job. The Job executes the command "python main.py" within the Studio environment.

1/3

2/3

3/3

from lightning_sdk.pipeline import Pipeline, JobStep, Studio

studio = Studio(name="studio_env")

pipeline = Pipeline(name='pipeline_from_studio', studio=studio)
pipeline.run(
    steps=[
        JobStep(command="python main.py"),
    ],
)

Using a container

Initialize the PipelineWe start by importing necessary classes from the lightning_sdkDefine PipelineNext, we create a pipeline instance named 'container_pipeline'Define stepsWe are running a single container Job. The Job executes the command "echo 'Hello, World!'" within the container.

1/3

2/3

3/3

from lightning_sdk.pipeline import Pipeline, JobStep

pipeline = Pipeline(name='container_pipeline')

pipeline.run(
  steps=[
      JobStep(
        image="ubuntu:latest",
        command="echo 'Hello, World!'"
      )
  ]
)