Added - Support for custom instructions and initializers in e3nn.flax.Linear and e3nn.haiku.Linear.
Changed - Fix documentation build errors.
0.20.6
Added - `e3nn.where` function - Add optional `mask` argument in `e3nn.flax.BatchNorm`
Changed - replace `jnp.ndarray` by `jax.Array`
0.20.5
Added - `e3nn.ones` and `e3nn.ones_like` functions - `e3nn.equinox` submodule
Fixed - python 3.9 compatibility
0.20.4
Fixed - Fix `pyproject.toml`, the documentation build was broken. Thanks to SauravMaheshkar!
Added - Support for [`s2fft`](https://github.com/astro-informatics/s2fft) in `e3nn.to_s2grid` and `e3nn.from_s2grid`, thanks to ameya98! - Add a special case implementation for `e3nn.scatter_mean` when `map_back and nel is not None`.
0.20.3
Added - `e3nn.flax.BatchNorm` - `e3nn.scatter_mean` - Add `e3nn.utils.vmap` also directly to `e3nn` module: `e3nn.vmap`
0.20.2
Added - `with_bias` argument to `e3nn.haiku.MultiLayerPerceptron` and `e3nn.flax.MultiLayerPerceptron`
Fixed - Improve compilation speed and stability of `s2grid` for large `lmax` (use `is_normalized=True` in `lpmn_values`)