Installation¶

TorchFort can be installed in multiple ways but we highly recommend building and using a Docker container.

Docker Installation¶

We provide a Dockerfile which contains all relevant dependencies. In order to build TorchFort using Docker, simply clone the repo and call:

docker build -t torchfort:latest -f docker/Dockerfile .

from the top level directory of the repo. Inside the container, TorchFort will be installed in /opt/torchfort.

CMake Installation¶

For a native installation TorchFort provides a CMakeList.txt file. Please make sure that the following required packages are installed on your system before installing TorchFort:

Requirements for core functionality and examples:
- CUDA 11.8 (CUDA 12.x is not supported yet)
- python version 3.6 or higher
- pybind11
- yaml-cpp from https://github.com/jbeder/yaml-cpp.git
- C++ compiler nvcc and Fortran compiler nvfortran. We recommend installing the NVIDIA HPC SDK Toolkit.
- HDF5
- the Python modules specified in requirements.txt
Additional requirements for building this documentation:
- Doxygen
- the Python modules specified in docs/requirements.txt

To build TorchFort, clone the repo and set the environment variable NVHPC_CMAKE_DIR to the directory where the NVIDIA HPC SDK Toolkit CMake config files are located. Those are usually located in <NVHPC root directory>/cmake. Then, call from the root directory:

mkdir build && cd build
cmake -DCMAKE_INSTALL_PREFIX=<TorchFort installation prefix> \
      -DNVHPC_CUDA_VERSION=11.8 \
      -DCMAKE_PREFIX_PATH="`python -c 'import torch;print(torch.utils.cmake_prefix_path)'`;${NHPC_CMAKE_DIR}" \
    ..
make -j install

Build Documentation¶

The documentation can be built with the corresponding Makefile in the docs directory. Make sure that the requirements are installed and call:

cd docs && make html

The docs will be located in docs/_build/html and can be viewed locally in your web browser.

Directory Structure¶

Independent of how you decide to install TorchFort, the directory structure will be as follows:

<TorchFort installation prefix>
|--- bin
     |--- examples
          |--- cpp
          |--- fortran
     |--- python
|--- include
|--- lib

The bin folder contains the examples written in C++ or Fortran located in the corresponding subdirectories. The python subfolder contains the Python wrappers for Weights and Biases Support.

The Fortran module torchfort.mod as well as the C headers can be found inside the include folder and the dynamic libraries inside the lib folder.