Skip to content

Add proper timing for model inference. #13

@DennisSmolek

Description

@DennisSmolek

Right now timing execution actually just times submission to the GPU. Not the actual work.

There are various strategies for this we can employ.

We should also time input and output handling better so we can see where in the flow speed improvements would be possible

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions