sts_lightspeed VS Serious-Engine

My plan would be to use CommunicationMod and sts_lightspeed to run approximately one bajillion runs and train the network based on that (there's even a pull request out for combining 'em).

If you haven't seen it, then essentially there is a command line version of sts here. I was going to see if it was at all possible for me to create a deep learning project on it for my senior year at uni, but I have been completely unable to compile the code. Has anyone tried it and had any luck?

and AI* a finished AI (that uses information that the player doesn't normally have).

There's a ultra-light-weight implementation of Slay the Spire by gamerpuppy. Some kind of ML is the ultimate goal. It's outside my wheelhouse, but I'm really interested in the concept!

https://github.com/gamerpuppy/sts_lightspeed I wrote a complete simulator with implementations for Ironclad in c++. It has python bindings for a few things right now and mcts similar to pasha’s. You might like the data analysis channel on the discord.

Croteam released the source code for the classic Serious Sam games for their 10th anniversary. Annoyingly, none of the games use exactly the version from the source, but they're generally similar enough. There are two discoveries which have been made through it, both for coop.