same for network pretrain code

[QIU023]

I want to know how can you get the 3DSeg-8 dataset well-trained and has such dice value, such as the hyper-parameters to train the network from scratch. How can I train such a simple decoder architecture with 3d-resnet152 to obtain such dice value on the 3DSeg-8 dataset? Or do you use another network architecture for the decoder?

Happy every day!