CYB08 / MambaClinix-PyTorch

MambaClinix: Hierarchical Gated Convolution and Mamba-Structured UNet for Enhanced 3D Medical Image Segmentation
11 stars 0 forks source link

how do i test model code? #1

Open Zhongrocky opened 2 months ago

Zhongrocky commented 2 months ago

this is amazing. This is a very complete set of split codes. I have a question. If I want to test the model code with the input as [1, 1, 96, 160, 160], then how should I set the hyperparameters in the model code (including input_channels, n_stages, kernel_sizes, and so on)?

CYB08 commented 2 months ago

Hi, @Zhongrocky,

We appreciate your interest in our project. MambaClinix inherits from the nnU-Net adaptive configuration, which automatically adjusts hyperparameters based on dataset characteristics, so you don't need to manually set these parameters. Simply adjust your data to match the specifications here. I see your data structure is [batchsize, Channel, D, W, H], you should reshape it to [D, W, H] for model testing.

Zhongrocky commented 2 months ago

Thank you for being so helpful. if I download the code, do I still need to download the official nnunetv2 library?

CYB08 commented 2 months ago

Thank you for being so helpful. if I download the code, do I still need to download the official nnunetv2 library?

No need. Please just follow our installation steps. Thanks!

Zhongrocky commented 2 months ago

I appreciate your help.

Zhongrocky commented 2 months ago

This model does exactly what it was supposed to do. The results are excellent. I have a question, what should I do if I only train one module for ablation experiments? For example only Resmamba or only HGCN.

CYB08 commented 1 month ago

2