Skip to content

Question about Pre-normalization #16

@NEUMUKI

Description

@NEUMUKI

Hi there,
I read you paper, notice that you mention the RMSNorm normalizing function is used in your code, but I can't find any class's forward is called which has the RMSNorm normalizing function in generation. Could you tell me where to use it? is it traning only?

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