Subject: Latest abrok Development Version December 30th Update default net to nn-b1e55edbea57.nnue (SF The gift that keeps on giving Sun Dec 31, 2023 11:39 am
Author: Linmiao Xu Date: Sat Dec 30 11:08:03 2023 +0100 Timestamp: 1703930883
Update default net to nn-b1e55edbea57.nnue
Created by retraining the master big net `nn-0000000000a0.nnue` on the same dataset with the ranger21 optimizer and more WDL skipping at training time.
More WDL skipping is meant to increase lambda accuracy and train on fewer misevaluated positions where position scores are unlikely to correlate with game outcomes. Inspired by: - repeated reports in discord #events-discuss about SF misplaying due to wrong endgame evals, possibly due to Leela's endgame weaknesses reflected in training data - an attempt to reduce the skewed dataset piece count distribution where there are much more positions with less than 16 pieces, since the target piece count distribution in the trainer is symmetric around 16
The faster convergence seen with ranger21 is meant to: - prune experiment ideas more quickly since fewer epochs are needed to reach elo maxima - research faster potential trainings by shortening each run