Skip to content

Can I use half precision training and inference data? #496

@julcst

Description

@julcst

Is there any difference between using half precision or full precision inputs and outputs when the network internally uses half precision for weights or can I safely reduce memory consumption without consequences?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions