ctak/src/nn_train.py, branch hybrid_rootstrapAn implementation of Tak and a computer opponent in Chttps://git.l-3.space/ctak/atom/src/nn_train.py?h=hybrid_rootstrap2026-09-25T15:18:21Zfactor out nn_train.py into a separate python project to make it more manageable, add plotting and offline strength fit2026-09-25T15:18:21Ztsliltslil@posteo.de2026-09-25T13:46:31Zurn:sha1:61e5b8949b2f5e8973c58f0dcc7c20c4bdade029try combining loops2026-09-20T07:29:22Ztsliltslil@posteo.de2026-09-20T07:28:09Zurn:sha1:6a2197328065fe38ac62d9e76e3e943fd297e73ctransposing the weights gives ~2.5x speedup2026-09-19T19:32:46Ztsliltslil@posteo.de2026-09-19T19:24:08Zurn:sha1:9f5f953a95c8617cd2fdd24f7402af632c1b07c2big refactor2026-09-19T18:23:34Ztsliltslil@posteo.de2026-09-19T08:59:29Zurn:sha1:fe7268110d93e2abe4d4b51ac64e83521a8d6382
* use elo
* optionally train on only diagonal
* train-inference nonlinearity alignment
grid search2026-09-18T20:19:33Ztsliltslil@posteo.de2026-09-18T14:12:13Zurn:sha1:3f918be9b31026ae546f66c565b9494dab6d82a2working towards hybrid rootstrap2026-09-18T14:12:19Ztsliltslil@posteo.de2026-09-17T06:01:29Zurn:sha1:0dba02072e9a89e35885d9078c0c8b58956448cemove data generation into the epoch loop2026-09-13T19:15:31Ztsliltslil@posteo.de2026-09-12T18:37:56Zurn:sha1:7f1566fb21dfe8c99bc770885cc2cc4e478e16b4
Previously we would spend all of the training epochs on making the network a better predictor of the *initial* TD-leaf values that we computed before the start of the first epoch. There's definitely something stale about this, given that the TD-leaf training targets involved the net evaluation at PVs. I'm not overly confident that this makes sense, but as a kind of compromise between fully on-line TD-leaf and the fully stale version we had before, we now recompute the TD-leaf values in each epoch with the fine-tuned net instead of the epoch 0 net.
Perhaps we were just lucky, but this seems to train a strong network under the default setting.
More experimentation is required.
still observing non-monotonic improvements between training generations2026-09-12T14:54:20Ztsliltslil@posteo.de2026-09-12T14:51:16Zurn:sha1:f3ed463e8a215b0ca3219aebbb14c4a7073a0f9fflood fill for roads, fix dragon, memoise leaf eval, enable diagonal in nn_train.py2026-09-12T14:47:15Ztsliltslil@posteo.de2026-09-11T17:36:10Zurn:sha1:a361728f39812e25599323c52510c0eb3131bd90move from single candidate td-leaf with gating to tournament of a population2026-09-11T16:00:18Ztsliltslil@posteo.de2026-09-11T11:56:12Zurn:sha1:6e90f1145b2c5c0f2dcc82d21ec45432274b8cad
The goal in this change was to overcome the rock-paper-scissors phenomenon that was happening in capability space by having on a single candidate learn to beat the most recent version of itself. In some sense this lead to rotating around in capability space instead of chasing monotonic improvements.
I hope that flipping to a system in which many (N=5) candidates play each other and learn from their own set of diverse games in this pool, combined with --replay-iters, will lead to a given network having to have a much more comprehensive understanding of the space of tactics. I haven't yet had time to train this thorougly, will run those experiments on my desktop computer.