AttributeError: module 'torch.nn' has no attribute 'Conv64d' -- Can you share your full list of dependencies?

[blutjens]

Do you know how to fix the error AttributeError: module 'torch.nn' has no attribute 'Conv64d' ? Would you be able to print all your dependencies?

Specifically, I assume there's a mismatch in one of the following modules; My setup is:

• pytorch 0.4.1post2
• torchvision 0.2.1
• cudatoolkit 10.1
• Python 3.6.10
• unet 0.2.0

Thank you very much!

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-1-9dd2f6c8fecf> in <module>
     26                             shuffle=True)
     27 
---> 28 model = UNet(3,3,64).cuda()
     29 """
     30 

/opt/conda/envs/eie-floods-seg-tst2/lib/python3.6/site-packages/unet/unet.py in __init__(self, in_channels, out_classes, dimensions, num_encoding_blocks, out_channels_first_layer, normalization, pooling_type, upsampling_type, preactivation, residual, padding, padding_mode, activation, initial_dilation, dropout)
     48             activation=activation,
     49             initial_dilation=initial_dilation,
---> 50             dropout=dropout,
     51         )
     52 

/opt/conda/envs/eie-floods-seg-tst2/lib/python3.6/site-packages/unet/encoding.py in __init__(self, in_channels, out_channels_first, dimensions, pooling_type, num_encoding_blocks, normalization, preactivation, residual, padding, padding_mode, activation, initial_dilation, dropout)
     40                 activation=activation,
     41                 dilation=self.dilation,
---> 42                 dropout=dropout,
     43             )
     44             is_first_block = False

/opt/conda/envs/eie-floods-seg-tst2/lib/python3.6/site-packages/unet/encoding.py in __init__(self, in_channels, out_channels_first, dimensions, normalization, pooling_type, preactivation, is_first_block, residual, padding, padding_mode, activation, dilation, dropout)
    106             activation=activation,
    107             dilation=dilation,
--> 108             dropout=dropout,
    109         )
    110 

/opt/conda/envs/eie-floods-seg-tst2/lib/python3.6/site-packages/unet/conv.py in __init__(self, dimensions, in_channels, out_channels, normalization, kernel_size, activation, preactivation, padding, padding_mode, dilation, dropout)
     28             padding = total_padding // 2
     29 
---> 30         conv_class = getattr(nn, f'Conv{dimensions}d')
     31         conv_layer = conv_class(
     32             in_channels,

AttributeError: module 'torch.nn' has no attribute 'Conv64d'