Skip to content

Make BatchNorm twice-differentiable #2255

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Make BatchNorm twice-differentiable #2255

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions src/cuda/cudnn.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,9 +13,9 @@ end

function ChainRulesCore.rrule(::typeof(batchnorm), g, b, x, running_mean, running_var, momentum; kw...)
y = batchnorm(g, b, x, running_mean, running_var, momentum; kw...)
function batchnorm_pullback(Δ)
grad = ∇batchnorm(g, b, x, unthunk(Δ), running_mean, running_var, momentum; kw...)
(NoTangent(), grad..., NoTangent(), NoTangent(), NoTangent())
function batchnorm_pullback(Δ, σ²Δ)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Where is this second parameter coming from? You can see on the CI log that this raises a MethodError. The pullback should only take one argument (Δ) and destructure it from a tuple into its component parts if need be.

grad, σ²grad = ∇batchnorm(g, b, x, unthunk(Δ), running_mean, running_var, momentum; kw...)
(NoTangent(), grad..., NoTangent(), NoTangent(), σ²grad..., NoTangent())
end
y, batchnorm_pullback
end