E3nn

Fixed
- `ReducedTensorProducts`: `normalization` and `filter_ir_mid` where not properly propagated through the recusive calls, this bug has no effects if the default values where used
- Use `torch.linalg.eigh` instead of the deprecated `torch.symeig`

Added
- (dev only) Pre-commit hooks that run pylint and flake8. These catch some common mistakes/style issues.
- classes to do `SO(3)` Grid transform (not fast) and Activation function using it
- Add `f_in` and `f_out` to `o3.Linear`
- `PBC` guide in the doc

Changed
- `FullyConnectedNet` is now a `torch.nn.Sequential`

Fixed
- `BatchNorm` was not equivariant for pseudo-scalars

Added
- `biases` argument to `o3.Linear`
- `nn.models.v2106`: `MessagePassing` takes a sequence of irreps
- `nn.models.v2106`: `Convolution` inpired from [Batch Normalization Biases Residual Blocks Towards the Identity Function in Deep Networks](`https://arxiv.org/pdf/2002.10444.pdf`)

Added
- [`opt_einsum_fx`](https://github.com/Linux-cpp-lisp/opt_einsum_fx) as a dependency
- `p=-1` option for `Irreps.spherical_harmonics(lmax, p)`

Removed
- Removed `group/_linalg` (`has_rep_in_rep` and `intertwiners`) (should use `equivariant-MLP` instead)

Added
- `preprocess` function in `e3nn.nn.models.v2103.gate_points_networks.SimpleNetwork`
- Specialized code for `mode="uuw"`
- `instance` argument to `nn.BatchNorm`

Added
- `pool_nodes` argument (default `True`) to networks in `e3nn.nn.models.v2104.gate_points_networks`
- Instruction support for `o3.Linear`
- `o3.Linear.weight_views` and `o3.Linear.weight_view_for_instruction`
- `nn.Dropout`

Changed
- `o3.Linear` and `o3.FullyConnectedTensorProduct` no longer automatically simplifies its `irreps_in` or `irreps_out`. If you want this behaviour, simplify your irreps explicitly!

Fixed
- `TensorProduct` can now gracefully handle multiplicities of zero
- `weight_views`/`weight_view_for_instruction` methods now support `shared_weights=False`

Added
- Normalization testing with `assert_normalized`
- Optional logging for equivariance and normalization tests
- Public `e3nn.util.test.format_equivariance_error` method for printing equivariance test results
- Module `o3.SphericalHarmonicsAlphaBeta`

Changed
- Generated code (modules like `TensorProduct`, `Linear`, `Extract`) now pickled using TorchScript IR, rather than Python source code.
- e3nn now only requires PyTorch >= 1.8.0 rather than 1.8.1
- Changed `o3.legendre` into a module `o3.Legendre`

Removed
- Removed `e3nn.util.codegen.eval_code` in favor of `torch.fx`