This is an OpenMP offload port of Tealeaf reference for GPUs (tested on AMD and Nvidia).
In most cases one needs to load the appropriate modules (including those for the accelerator in question).
Then for AMD/Nvidia GPUs (with the Cray compiler):
make clean; make COMPILER=CRAY
or for Nvidia GPUs (nvhpc compilers),
make clean; make COMPILER=PGI