Skip to content

Hypernetwork output initialization  #16

Description

@nathanieljevans

"Classical initialization techniques do not guarantee that the weights of the target network are
initialized within the same range. However, adaptive optimizers, such as Adam [30], can mitigate this issue to some
extent. " (https://arxiv.org/pdf/2306.06955)

Should come up with a method so that early training weights match common weight initialization ranges.

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

    enhancementNew feature or request

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions