Cristian Garcia

Given that `from_state_dict`'s current behaviour is to check on a best-effort basis and its widely used inside the code base for various auxiliary tasks, efforts to fix have felt a...

I'll look into this. I'll document the signature for each and clarify that `bias_initializer`s don't work with the `variance_scaling` family.

Hey @dslisleedh, thanks for creating this PR! I think this looks good, however, before merging we should probably add a test using this flag.

Just noticed there aren't any tests for `pooling.py`, I'll create an issue, we can add a test for this in the future.

@dslisleedh can you create one of more tests under `PoolTest` using this new argument? You can copy an existing test, rename it, and change it to use the new `count_include_pad`...

Awesome @dslisleedh! Can you commit changes to the tests?

@dslisleedh can you add this test? None of the other tests used `padding="SAME"` which is the more interesting case. ```python @parameterized.parameters( {'count_include_pad': True}, {'count_include_pad': False}) def test_avg_pool_padding_same(self, count_include_pad): x =...

I was playing around with mixins to see how they interact with `Module`. See experiments Following case works: ```python import flax.linen as nn import jax.numpy as jnp import jax class...

There is work in progress to remove TensorFlow as a dependency in #2073 and #2444. This should be solved in the future.

I like the idea! Want to send a PR? > Note that class Module would have to be updated so that the Module fields are defined (note doing this today...