From 312318ad6548f65cc6af118c24811dfd40280b27 Mon Sep 17 00:00:00 2001 From: Dylan Date: Sun, 8 Nov 2026 17:32:09 +1300 Subject: feat: Resume training from any weights; balance training by outcome - train --from starts from any saved network; without it, training still carries on from the out dir's newest generation - Each epoch takes as many positions from sheep-win games as from tiger-win games (Config::balance_outcomes, on by default) - Training output shows the error on tiger-win and sheep-win positions separately, before and after training - Fix a broken sentence in CLAUDE.md's cache description Co-Authored-By: Claude Opus 5.5 --- plans.org | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'plans.org') diff --git a/plans.org b/plans.org index 5a06798..8f53255 100644 --- a/plans.org +++ b/plans.org @@ -67,3 +67,14 @@ Before the fix, tigers won ~97% of self-play games and the net only reached ~15- *** Next candidates - Pick exploratory moves among near-best ones (softmax over root scores) rather than uniformly - Sheep are still the weaker side: report the match score for each side separately +** DONE Balance tiger-win and sheep-win positions +The unbalanced net (gen 100) had 7x the error on positions from sheep-win games (0.045) as on +tiger-win ones (0.0065), so it leaned towards tiger wins, though not a constant prediction (that +would give errors near 2). =balance_outcomes= trains on equal numbers of each. +| vs =ClaudePlayer= (depth 3, 100 games) | gen 25 | gen 50 | gen 75 | gen 100 | +|----------------------------------------+--------+--------+--------+---------| +| Unbalanced | 25% | 50% | 54% | 62% | +| Balanced | 22% | 19% | 51% | 62% | +- The T and S errors are now similar (both ~0.02), but the strength reached is the same; balanced was slower + in the middle. Self-play is still ~95% tiger wins, from the random sheep moves. +- "vs prev" stays at ~50% because after 4 random plies both sides play deterministically; it says little. -- cgit v1.3.1