complex-loss is a Python package that provides CUDA accelerated loss functions for PyTorch.
This library was created in response to pytorch/pytorch#47052.
Complex numbers are far richer than the real numbers and as such need their own loss functions.
The project is currently in active development, with the intention of devloping
These aren't available yet, but hopefully will be migrated to PyPi.
If you're installing from source, you need to install PyTorch.
git clone https://github.com/rjkilpatrick/complex-loss
python3 ./setup.py-[ ] torch.nn.ELU
-[ ] torch.nn.Hardshrink
-[ ] torch.nn.Hardsigmoid
-[ ] torch.nn.Hardtanh
-[ ] torch.nn.Hardswish
-[ ] torch.nn.LeakyReLU
-[ ] torch.nn.LogSigmoid
-[ ] torch.nn.MultiheadAttention
-[ ] torch.nn.PReLU
-[ ] torch.nn.ReLU
-[ ] torch.nn.ReLU6
-[ ] torch.nn.RReLU
-[ ] torch.nn.SELU
-[ ] torch.nn.CELU
-[ ] torch.nn.GELU
-[ ] torch.nn.Sigmoid
-[ ] torch.nn.SiLU
-[ ] torch.nn.Softplus
-[ ] torch.nn.Softshrink
-[ ] torch.nn.Softsign
-[ ] torch.nn.Tanh
-[ ] torch.nn.Tanhshrink
-[ ] torch.nn.Threshold