nnU-Net Custom Training: Feature Request

NVIDIA / DeepLearningExamples

State-of-the-Art Deep Learning scripts organized by models - easy to train and deploy with reproducible accuracy and performance on enterprise-grade infrastructure.

13.28k stars 3.18k forks source link

nnU-Net Custom Training: Feature Request #1115

Open sbmalik opened 2 years ago

sbmalik commented 2 years ago

Update Custom Training Notebook for nnU-Net

Examples:

PyTorch
Jupyter notebooks

Is your feature request related to a problem? Please describe. Current notebook for custom training not working.

Describe the solution you'd like An updated version of custom 2d dataset training notebook

Additional context some imports cause errors, but I resolved them. After that DataLoader for validation data is giving following error update() takes 3 positional arguments but 4 were given

sbmalik commented 2 years ago

@michal2409 or @nv-kkudrynski can you please look into this.. ?

michal2409 commented 2 years ago

Hi @sbmalik, do you run the notebook on xView2 dataset or using something different? In case you are using different dataset you might to modify some code, e.g. the metric update function.

sbmalik commented 2 years ago

@michal2409 thank you for response, I tried with the same dataset but getting the error because of validation dataloader (I think), mentioned in question. Moreover you can reproduce the error by just running the notebook.