Skip to content

Update Weights Operation #13

Open
Open

Description

@EscVM

Hi @eric-mitchell,

Thank you for the repo and the great paper. I was looking at your code and I have a doubt on this line of code: why is there a plus? Shouldn't be a minus?
As you also wrote in the paper, the final edited weights should be:

$$\tilde{W} = W_l − α * \tilde{\nabla} W_l$$

Thanks for your help.

Activity

  1. eric-mitchell commented on May 19, 2023

    @eric-mitchell
    Owner

    You're right about this- good catch! Turns out this is indeed a bug in the implementation. Have you been able to train a model with this corrected, and seen if there is a difference in performance?

  2. EscVM commented on May 21, 2023

    @EscVM
    Author

    Actually, I've never run your code, but I adapted it for another use case. Anyway, the code works also with the plus. The optimization process makes the network flip all gradient signs. However, I expect slightly better results with the right sign from the beginning, simply because the task should be a little simpler from the point of view of hypernetworks.

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

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions