Welcome to PyMAIA documentation!

PyMAIA

What is PyMAIA?

Hive is a Python package to support Deep Learning data preparation, pre-processing. training, result visualization and model deployment across different frameworks (nnUNet , nnDetection, MONAI ).

Local Environment Setup

To install the package, run the following commands:

conda install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia
or conda install cudatoolkit cuda-version=11
pip install nnunetv2/nndetection
pip install pymaia-learn

More information can be found in the documentation.

Tutorials

• nnUNet Tutorial

• nnDetection Tutorial

Docker and Singularity

PyMAIA can be run in a containerized environment using Docker or Singularity. To create the PyMAIA image, you can use HPPCM, a tool to create container images for HPC applications from given recipes.

pip install hpccm

hpccm --recipe recipe.py --format singularity > PyMAIA.def
singularity build PyMAIA.sif PyMAIA.def

hpccm --recipe recipe.py --format docker > Dockerfile
docker build -t pymaia .

Contents:

• PyMAIA package
  • Subpackages
  • Module contents
• Configs
  • nnDetection config
  • nnUNet config
• Scripts
  • PyMAIA Scripts
  • PyMAIA Scripts for nnDetection
  • PyMAIA Scripts for nnUNet
• Tutorials
  • nnDetection Tutorial
  • nnUNet Tutorial
  • nnUNet Tutorial Notebook

Indices and tables

• Index

• Module Index

• Search Page