Issues
- 0
- 1
- 1
- 0
imagenet example fails with AttributeError: 'Config' object has no attribute 'define_bool_state'
#3950 opened by sycamoreoak - 2
- 6
Problem in usage of multi-gpu (Four RTX 4090) for FLAX
#3931 opened by psj1866 - 2
nn.Dense doesn't respect the provided dtype
#3945 opened by dfdx - 3
- 5
No way to call nnx.State.from_flat_path
#3879 opened by NeilGirdhar - 1
Documentation links 404
#3917 opened by namiller - 2
More memory consume compared with Pytorch
#3766 opened by Sun-Xiaohui - 2
Tutorial request
#3890 opened by Wasim04 - 1
- 3
Deprecation Warnings with orbax 0.5.3
#3708 opened by lucidfrontier45 - 1
- 0
Add ability to easily change documentation version
#3858 opened by emergenz - 1
with_partitioning has surprising behavior with MultiHeadAttention and DenseGeneral
#3892 opened by jameslyon - 1
nn.remat_scan doesn't work with nn.with_partitioning
#3870 opened by jameslyon - 2
Improve SEO for docs pages
#3854 opened by carlosgmartin - 4
- 1
nnx static fields not part of static tree structure
#3863 opened by NeilGirdhar - 2
Orbax checkpoint for LogicallyPartitioned params
#3787 opened by mmorinag127 - 3
- 1
Memory issue when randomly initializing large parameters, sharding cannot help
#3701 opened by hr0nix - 0
- 1
- 0
- 0
GroupedConv distributed training failure
#3831 opened by sali1997s - 0
Add `BatchRenorm` layer to `linen.normalization`
#3822 opened by danielpalen - 0
Remove `tree_map` deprecation filter after Flax upgrades minimum Python version to 3.10
#3842 opened by chiamp - 2
Documentation/notebook errors
#3837 opened by st-- - 4
*Module Parameters* section of docs is outdated.
#3761 opened by PaulScemama - 5
Error when calling module tabulate involving WeightNorm
#3735 opened by DBraun - 0
Error when calling `Module.tabulate` on normalization wrappers like `WeightNorm` and `SpectralNorm`
#3779 opened by chiamp - 0
- 0
Feature request: Add ConvGRUCell
#3713 opened by carlosgmartin - 4
Add `reset_gate` flag to `MGUCell`
#3718 opened by carlosgmartin - 1
- 1
- 1
modifying params of flax.linen. Module model
#3730 opened by lopa23 - 0
- 0
flax.linen.module.init still fails under dynamic type checking for nested modules
#3756 opened by evangelos-ch - 2
Compatibility with Torch LSTM
#3746 opened by mttga - 1
Unify behavior of strides arg of Conv and ConvTranspose
#3729 opened by younik - 4
Add vanilla / Elman / simple RNN cell
#3681 opened by carlosgmartin - 1
- 4
asyncio error while loading weights
#3679 opened by mobley-trent - 1
- 3
VAE example outdated
#3666 opened by tisabe - 0
Make `self.make_rng()` callable by using a default RNG stream and have `.init()` and `.apply()` use the default RNG stream if no explicit RNG mapping is passed.
#3668 opened by chiamp