I used an Anaconda environment with Python3.8. The dependencies are in the requirements.txt file. To replicate the environment, you can just run:
conda create --name <envname> --file requirements.txt
Running python project/setup_optimizer_config.py will create the the json file(s) that are needed for hypermapper. The benchmark programs are located in project/programs and can be imported and run in the main file using python main.py. The benchmark class for RocksDB is defined in project/benchmark.py