Repository navigation
fix: Support zero-size shapes in variance scaling initializers - #683
Open
georgeguimaraes wants to merge 1 commit into
Open
georgeguimaraes wants to merge 1 commit into
georgeguimaraes wants to merge 1 commit into
Conversation
compute_fans got the receptive field size by dividing the size of the shape by the fans. A dimension of 0 then divided 0 by 0, for example for the hidden state of an LSTM with a batch of 0. The receptive field size is now the product of the dimensions before the last two. A shape of rank 1 has fans of 1, which is the value that the division gave. The fans of other shapes do not change. This needs zero-size tensors in Nx (elixir-nx/nx#1885).
polvalente
approved these changes
Oct 20, 2026
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Depends on elixir-nx/nx#1885 and elixir-nx/nx#1886 (zero-size tensors in Nx). CI fails until they merge.
compute_fansdivided the size of the shape by the fans, so a dimension of 0 divided 0 by 0. For example, the hidden state of an LSTM with a batch of 0 raisedArithmeticError. The receptive field size is now the product of the leading dimensions. The fans of other shapes do not change.