README.md

Nanoconfinement

Link to the application: https://nanohub.org/resources/nanoconfinement

References:

• https://doi.org/10.1016/j.jocs.2020.101107

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 nanoconfinement
   cd nanoconfinement
   mkdir output

3. Get the datasets for training

   ${SBI_FAIR_DIR}/tools/scripts/load_dataset.py ${SBI_FAIR_DIR}/datasets/nanoconfinement/datasets.yaml nanoconfinement_4050

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/softmaterialslab/nanoconfinement-md for the list of all options.

Using Docker

1. Build Docker container

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

2. Run Training

   GPU_SWITCH='--runtime=nvidia --gpus all' # or '' for CPU workloads
   # Mount the directories with the dataset
   VOLUME_MOUNTS='-v ./nanoconfinement_4050/train/data_dump_density_preprocessed_train.pk:/input/train_dataset -v ./nanoconfinement_4050/test/data_dump_density_preprocessed_test.pk:/input/test_dataset -v ./output:/output -v ./options.yaml:/input/options.yaml'
   docker run ${GPU_SWITCH} ${VOLUME_MOUNTS} nanoconfinement run train

Using Apptainer

1. Build Apptainer container

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

2. Run Training

   GPU_SWITCH='--nv' # or '' for CPU workloads
   # Mount the directories with the dataset
   VOLUME_MOUNTS='--bind ./nanoconfinement_4050/train/data_dump_density_preprocessed_train.pk:/input/train_dataset --bind ./nanoconfinement_4050/test/data_dump_density_preprocessed_test.pk:/input/test_dataset --bind ./output:/output --bind ./options.yaml:/input/options.yaml'
   apptainer run --app train ${GPU_SWITCH} ${VOLUME_MOUNTS} ${SBI_FAIR_DIR}/models/nanoconfinement/nanoconfinement.sif

Results

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