Repository Issue Activity (beta)

#2730 Please return the old Flux.train!.

#2444 test Enzyme gradient for loss functions

#1385 TagBot trigger issue

#2722 Failure with distributed data parallel training (DDP) with non-CUDA/ROCM aware MPI

#2723 pad_mode argument for conv layers missing when tranferring to AMDGPU

#2410 Collecting PyTorch -> Flux migration notes

#2543 display fieldnames in show of a container

#2013 Functional equivalents missing for some layers

#2431 SamePad() for even sized filters.

#1917 Conv with circular padding

#2407 Zero-sized arrays cannot be applied to Dense layers.

#2571 infinite recursion when applying `f32` on empty layer

#2154 `BatchNorm` is not twice-differentiable

#2179 wrong gradient for BatchNorm on gpu in testmode

#91 ConvRNN Cell and ConvLSTM Cell

#1992 Batchnorm's variance becomes NaN when minibatch contains just one sample

#2122 ForwardDiff + destructure is different from Zygote, on a model with BatchNorm

#2469 DifferentiationInterface testing

#2572 Create a FluxCore package?

#2715 Audit Open Issues - August 2026

#2658 broken AMDGPU support

#2648 Zero-sized Array after indexing causes exception during gradient

#2588 `Nil` doesn't understand SpecialFunctions functions

#2616 BatchNorm Memory Blowup on Very Large CUDA Arrays

#2584 Some models saved under Flux v0.14 do not load on v0.15+