这是indexloc提供的服务,不要输入任何密码
Skip to content

feat(blueprint): Parameterize machine_type for inference gateway blueprint #4717

@SinaChavoshi

Description

@SinaChavoshi

Description
As a follow-up to the initial blueprint implementation under PR #4699 , this feature request is to refactor the blueprint to use variables for key configuration parameters, specifically machine_type.

This suggestion, originally made by @samskillman, will allow us to use the same blueprint for different accelerator types (e.g., A3, G4) by simply changing the input variables.

Motivation
The primary motivation is to reduce the number of blueprints we need to create and maintain. Currently, supporting a new machine type would require duplicating the entire blueprint with minor changes. Parameterizing these core values will lead to a more generic, reusable, and maintainable blueprint.

Proposed Changes

  • Convert the hardcoded machine_type attribute to an input variable.
  • Update documentation (e.g., README.md) to reflect the new variables and provide examples.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions