flytekitplugins.inference.Model

class flytekitplugins.inference.Model(name, mem='500Mi', cpu=1, modelfile=None)

Represents the configuration for a model used in a Kubernetes pod template.

Parameters:

• name (str) – The name of the model.

• mem (str) – The amount of memory allocated for the model, specified as a string. Default is “500Mi”.

• cpu (int) – The number of CPU cores allocated for the model. Default is 1.

• modelfile (str | None) – The actual model file as a JSON-serializable string. This represents the file content. Default is None if not applicable.

Methods

Attributes

cpu: int = 1

mem: str = '500Mi'

modelfile: str | None = None

name: str