Skip to content

why the mse_loss is performed on the scaled gt latents and the unscaled x0_pred #52

Description

@garychan22

loss_x0 = F.mse_loss(x0_T.float(), latents.float(), reduction="mean")

Hi, I have a question that why the mse_loss is performed on the scaled gt latents and the unscaled x0_pred instead of the scaled x0_pred?

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

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions