Skip to content

should I use openxla or intel extension for transformers to load Whisper Jax Model? #36

Open
@sleepingcat4

Description

@sleepingcat4

I wanted to load Whisper Jax on Intel datacentre GPU Max series. I was wondering should I use Intel OpenXLA extension or Intel extension for transformers? not sure if openxla supports quantisation by default or not

Metadata

Metadata

Assignees

No one assigned

    Labels

    questionFurther information is requested

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions