Skip to content

Reduce image size for inference scheduler #152

@kfirtoledo

Description

@kfirtoledo

There is a difference in image size between llm-d (448 MB) and the GIE version (171 MB).
This is primarily due to the base image used:
llm-d uses registry.access.redhat.com/ubi9/ubi:latest, whereas the upstream uses gcr.io/distroless/base:nonroot.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    Status

    No status

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions