lucidrains / se3-transformer-pytorch

Implementation of SE3-Transformers for Equivariant Self-Attention, in Pytorch. This specific repository is geared towards integration with eventual Alphafold2 replication.
MIT License
262 stars 23 forks source link

denoise.py bugfix #15

Closed javierbq closed 3 years ago

javierbq commented 3 years ago

Fixes issue related to the constructor

Traceback (most recent call last):
  File "/home/jcastellanos/projects/se3-transformer-pytorch/denoise.py", line 22, in <module>
    transformer = SE3Transformer(
  File "/home/jcastellanos/projects/se3-transformer-pytorch/se3_transformer_pytorch/se3_transformer_pytorch.py", line 1072, in __init__
    self.num_degrees = num_degrees if exists(num_degrees) else (max(hidden_fiber_dict.keys()) + 1)
AttributeError: 'NoneType' object has no attribute 'keys'