conversation zone, music, game, programming, cooking, whatever else. they
Pan
Loading...
one thing thats nice about offline RL is that it means i can continue scaling how good the model is with more replay data, i doubt ill get much more at this point but i might go around and ask for ones with the rarer characters
Xrd model now uses RL, this isn't a finished training run, and no online RL has been applied yet as i'm still working on the mod setup for that, but i'm very happy with the progress! here's the model playing against itself, i think it's a massive step up from the last footage i posted
it seems like the offline stuff is really effective at adding combos and setups to the models Bag of Tricks, which is great because i think the model would probably take 1 million years to learn any of that playing against itself
once i get the training stuff set up i might wanna create some kind of livestream on twitch where you can tune into the training matches, something like that could be fun