Skip to content

In the paper, how is the compression of the switch-transformer c2048 to 25.4x calculated? #6

Description

@SHUSHENGQIGUI

hi,
i am trying to understand qmoe, but it is too hard to figure it out that how you compute the theoretical compression limit of 25.4x of c2048 switch transformer. could you provide detailed computation process?

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

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