Error in Binary conv layer

[AnouarITI]

I am trying to refactor the results with the CIFAR-10 dataset with same model and configuration. I am now trying to train from scratch with the same nvidia docker version stated in the repo 21.06-tf1-py3, However, now I am getting this error:

/workspace/LUTNet_experiment/binarization_utils.py in build(self, input_shape)
    197                 #random.seed(self.rand_seed)
    198 
--> 199                 if self.k_lut > 1:
    200 			self.rand_map_0 = self.add_weight(name='rand_map_0', 
    201                                 shape=(self.window_size, 1),

TypeError: '>' not supported between instances of 'TFModuleWrapper' and 'int'

What did I do wrong, and how should I fix it?
Thank you