This repo contains a few training and evaluation scripts for instance segmentation with PyTorch-RDC-Net.
The commands to train on the MoNuSeg dataset. For other datasets, the commands are similar, but the dataset name and paths need to be changed. Please check the pixi.toml file under the [task] section.
WD=runs/MoNuSeg-baseline pixi run build_rdcnet_config
WD=runs/MoNuSeg-baseline pixi run build_trainer_config
The configs will be saved in the runs/MoNuSeg-baseline directory. If the MoNuSeg-baseline directory does not exist, it will be created.
WD=runs/MoNuSeg-baseline ACCOUNT=your_cluster_account pixi run submit_train_MoNuSeg
WD=runs/MoNuSeg-baseline pixi run train_MoNuSeg
This project was generated with the faim-ipa-project copier template.