Jacob Marshall

I'm not sure how you'd reconcile/merge search tree states across a single game, as the next MCTS iteration depends on the state reached from the previous one. If you know...

Looks interesting, thanks for sharing! When I have some time I may explore adding some of these ideas, not sure how well it will work with the existing batching paradigm...

It would be very very neat to be able to batch across many environments as well as across MCTS iterations!

I agree! This project is mostly focused on training at scale, but nevertheless it could be interesting to allow for a mix of batching across many environments as well as...

I was able to reproduce, then correct the behavior you are describing. First, in [ 16d2f3f](https://github.com/bubble-07/turbozero/commit/16d2f3f30fc1c80429ed67fb548f8fc87da85e96), you move the legal actions assignment to before `env.step` in `mcts.py`, meaning that legal...

87fd4d8c allows for negative rewards/evaluations. I'll keep this open until I address label smoothing as well, and perhaps debug asserts for detecting invalid actions in MCTS. Let me know if...

Apologies for the incredibly late reply... Yeah the main difference here is this repo implements subtree saving and also includes all the necessary utilities for an AZ training run in...

Sorry, missed your response here. In zed logs I see a lot of: 2025-03-24T01:32:22.899698-07:00 [ERROR] language not found When I first open Zed, I don't see ruff listed in the...

sorry for the delay here, will address this one this weekend