uni-medical / STU-Net

The largest pre-trained medical image segmentation model (1.4B parameters) based on the largest public dataset (>100k annotations), up until April 2023.

uni-medical/STU-Net Issues

pickle error occurs when running the Direct Inference command
Updated 2 months ago
How to use STUNet under nnUnet V2 version for completely starting from scratch training
Closed 3 months ago
How to train from scratch(re-pretraining)?
Closed 4 months ago1
Hello, I am truly sorry for interrupting your work
Updated 4 months ago5
Issues about pkl files
Updated 4 months ago1
DP
Updated 5 months ago
Some questions about the setting of initial_ir
Updated 6 months ago
crop size is patch size?
Updated 6 months ago
New label
Updated 6 months ago
Surpport 2D segmentation
Updated 7 months ago
Are pretrained models compatible with nnUNet v2?
Updated 8 months ago
Direct evaluation
Closed 7 months ago4
huge model training supermemory problem
Updated 8 months ago3
Request for clarification on the architecture
Closed 8 months ago10
The unpacking dataset stopped for several hours
Closed 8 months ago
Can you provide the metrics evaluation scripts for the WORDS and FLARE datasets?
Updated 8 months ago4
about output channel and label
Closed 10 months ago2
How to make the model run in DDP
Updated 8 months ago1
About the commands to pretraining the model
Updated 8 months ago4
Label overlap
Updated a year ago2
Hyper-parameter of fine-tune
Closed a year ago2
Could not find trainer class
Closed a year ago4
model infer error
Closed a year ago
Dataset conversion instructions
Closed a year ago9
What is the “3d_fullres” mentioned in README
Closed a year ago
About the data preprocessing parameters for inference
Closed a year ago3
Would it be possible to share weights for other models from the paper?
Updated a year ago4
Missing <model_name>.model.pkl for inference
Updated a year ago1