API reference¶

Sphinx domains turn API-style content into linkable, searchable documentation. The declarations below are documentation examples and do not require a Python package.

class Pipeline(name, replicas=1)¶: Represents a deployable AI workload.

Pipeline.name: str¶: The display name for the pipeline.

async Pipeline.scale(replicas)¶

Scale the pipeline to replicas workers.

Parameters:: replicas (int) – The desired worker count.
Returns:: The updated pipeline.
Return type:: Pipeline

create_pipeline(name: str, replicas: int = 1) → Pipeline¶

Create a new Pipeline.

Parameters:

name – A unique pipeline name.
replicas – The initial worker count.

Returns:

A configured pipeline.

Usage¶

pipeline = create_pipeline("inference", replicas=3)

Return to the Configuration example in the authoring guide for the corresponding data model.