OmniOpt2 allows you to easily optimize complex hyperparameter configurations. It is based on Ax and BoTorch
You can run OmniOpt2 on any linux that has python3 and some basic necessities. If something is missing that cannot be installed, it will tell you how to install it.
If the system you run it on has Slurm installed, it will use it to parallelize as you set the settings. If you run it locally without slurm, they will simply run sequentially.
OmniOpt2 installs all of it's python-dependencies automatically in a virtual environment once at first run. This may take up to 20 minutes, but has to be done once. This is done to isolate it from other dependencies and to make you focus on your task of hyperparameter-minimization.
In short: It will try out your program with different hyperparameter settings and tries to find new ones to minimize the result using an algorithm called BoTorch-Modular.
Read the documentation of how you need to change your program and then go to the OmniOpt2-GUI and start optimizing.