lucidrains / meshgpt-pytorch

Implementation of MeshGPT, SOTA Mesh generation using Attention, in Pytorch
MIT License
732 stars 60 forks source link

Version conflict with x-transformer-1.31.6 #89

Closed hyeonjang closed 3 months ago

hyeonjang commented 3 months ago

Issue: AttributeError in RMSNorm class

Description

When running the code, an AttributeError is raised, indicating that the RMSNorm class does not have a unit_offset attribute.

Error Message

  File "C:\Users\user00\miniconda3\envs\vrmx\Lib\site-packages\x_transformers\x_transformers.py", line 634, in forward       
    return F.normalize(x, dim = -1) * self.scale * (self.g + self.unit_offset)
                                                             ^^^^^^^^^^^^^^^^
  File "C:\Users\user00\miniconda3\envs\vrmx\Lib\site-packages\torch\nn\modules\module.py", line 1709, in __getattr__        
    raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'")
AttributeError: 'RMSNorm' object has no attribute 'unit_offset'

Workaround

The issue was temporarily solved by downgrading to version 1.30.19 of the package.

lucidrains commented 3 months ago

@hyeonjang hey, i've fixed the dep to be less than 1.31 for now