Error with Load Weight and Configuration Modification Impacting Accuracy: Base Model Training Inquiry

[NieMinxin]

Thank you so much for doing such a good job, I did get 86.1 accuracy with your base_best weights, but I used 3090 graphics (unlike your V100) and the environment was the same. I directly used your commands to train the model, but there was an error of load weight. I modified your configuration file and trained Base from the beginning, and the accuracy was only 84.7. I am not sure if there is any error in your configuration file, or could you please tell me how you trained base 85.7, thank you very much

[peijie-chiu]

@NieMinxin Thank you for being interested in our work and reporting issues. It makes sense to achieve a lower performance training from scratch (without having a pretrained weight) on the multi-organ segmentation. I checked the config file, it should be correct. Could you be more specific about your error of loading the weight (e.g., the error log). Is it loading the pretrained weight from DAT++ ? If that's the case, I'll look into that (there might be some errors after I reorganized the code). It seems that I do not face any bugs when loading the pretrained weight from DAT++ from my side. I'll appreciate if you can provide more details on your errors; it would be very informative for me to verify things and save my time.

[Linz-Z]

Thank you so much for doing such a good job，I encountered an issue while installing the package with "RuntimeError: Couldn't load custom C++ops. Recompile C++extension with: Python setup.py build_ext -- in-place". I understand that the version of torch vision needs to be greater than 0.3 to use ops, but the torch vision version that is compatible with CUDA 12.1+PyTorch 2.2 is 0.17. Can you please provide the complete package and version? thanks

[thekingofbigdate]

Hello, I have also encountered [zl1216]（ https://github.com/zl1216 ）The question I raised is that I used pytorch=2.20 torch vision=0.17.0 torch studio=2.20 pytorch studio=12.1. May I know your version

[allsundayy]

Thank you so much for doing such a good job, I used CUDA 12.1, PyTorch 2.2.1, and NATTEN 0.15.0 to train the Base model from scratch, but the accuracy is only 0.845676.I am not sure if there is any error in your configuration file, or could you please tell me how you trained base 85.7, thank you very much.