README.md

CaloGAN

Link to the application: https://github.com/hep-lbdl/CaloGAN

References:

• https://doi.org/10.1103/PhysRevLett.120.042003
• https://doi.org/10.48550/arXiv.1705.02355

How to use (training)

Setup and download dataset

1. Get the SBI-FAIR repository

   git clone --depth 1 https://github.com/DSC-SPIDAL/sbi-fair
   SBI_FAIR_DIR=${PWD}/sbi-fair

2. Create a directory for downloading datasets and store results

   mkdir calogan
   cd calogan
   mkdir output

3. Get the datasets for training

   ${SBI_FAIR_DIR}/tools/scripts/load_dataset.py ${SBI_FAIR_DIR}/datasets/calogan/datasets.yaml calogan_eplus

   You can use any of the following datasets:

   • calogan_eplus
   • calogan_gamma
   • calogan_piplus

4. Create a file with parameters

   # Few epochs for testing
   echo 'epochs: 2' > options.yaml 

   We will update the list of available options here, in the meantime please refer to the original repository https://github.com/hep-lbdl/CaloGAN for the list of all options.

Using Docker

1. Build Docker container

   cd ${SBI_FAIR_DIR}/models/calogan
   ./build_docker.sh
   cd - # Go back to results directory 

2. Run Training

   At the moment the provided containers are likely to fail to using GPU with recent systems due to CUDA incompatibilities. We are working to fix that.

   GPU_SWITCH='' # or '--runtime=nvidia --gpus all' for GPU workloads
   # Mount the directories with the dataset
   VOLUME_MOUNTS='-v ./calogan_eplus/default/eplus.hdf5:/input/train_dataset -v ./output:/output -v ./options.yaml:/input/options.yaml'
   docker run ${GPU_SWITCH} ${VOLUME_MOUNTS} calogan run train

Using Apptainer

1. Build Apptainer container

   cd ${SBI_FAIR_DIR}/models/calogan
   ./build_apptainer.sh
   cd - # Go back to results directory 

2. Run Training

   At the moment the provided containers are likely to fail to using GPU with recent systems due to CUDA incompatibilities. We are working to fix that.

   GPU_SWITCH=''  # or '--nv' for GPU workloads
   # Mount the directories with the dataset
   VOLUME_MOUNTS='--bind ./calogan_eplus/default/eplus.hdf5:/input/train_dataset --bind ./output:/output --bind ./options.yaml:/input/options.yaml'
   apptainer run --app train ${GPU_SWITCH} ${VOLUME_MOUNTS} ${SBI_FAIR_DIR}/models/calogan/calogan.sif

Results

The outputs of the run will be available in ./output.