From f11793dc0fddfcaad04b89827634cbb40fa98e59 Mon Sep 17 00:00:00 2001 From: tslil Date: Wed, 2 Sep 2026 12:21:20 +0100 Subject: default engines to depth 4, add a rolling window to nn_train.py I've noticed high variability in training efficacy, part of that i think is due to the fact that the dataset per iteration is not really that large. Sure, it has many rows, but it's fundamentally given by the same engine playing itself with some entropy injection. To remedy this i've expanded the input data for any given iteration to contain --replay-iters many previous datasets as input too. Alongside that decreased the epsilon decay rate for slower "cooling", and included much more of the early game in the training set (min ply 8 -> 4). --- src/ctlm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src/ctlm.c') diff --git a/src/ctlm.c b/src/ctlm.c index c9844c1..d2b59c9 100644 --- a/src/ctlm.c +++ b/src/ctlm.c @@ -487,7 +487,7 @@ int main(int argc, char **argv) { puts(license); - negamax_search_depth = 5; + negamax_search_depth = 4; new_game(5); negamax_init(5); -- cgit v1.3.1