[currymj]

this is very cool.

another interesting open-source project for imperfect-information game solving is DeepMind's OpenSpiel:

https://github.com/deepmind/open_spiel

However as far as poker goes, I believe they only provide toy versions of the game, even though theoretically you could put in your own implementation.

I've noticed that in this part of the AI world, things tend to be closed-source for whatever reason and I think that limits the ability to learn. Seems good to see that changing.

[bupticybee]

Yeah I am aware of this project, and I am current trying write an alpha-star like AI for another holdem simulator https://www.pettingzoo.ml/classic, I already made some progress and I somehow believe the code for a alpha-star like texas-holdem AI will be released in sometime of the future.

[sharky6000]

Hi, thanks:)

It does have the large poker games too. You can find generalized poker in the universal_poker game, which wraps the Annual Computer Poker Competition codebase:

https://github.com/deepmind/open_spiel/blob/master/open_spie...

Check out universal_poker_test.cc for example uses. It supports limit and common abstractions as well as the full game of no-limit. There is an example of FCPA which you can play at the keyboard in examples/

[currymj]

oh cool, good to know!