Ceechess - UCI chess engine,
Author: Hamza Inan, Rating CEDR=2376
v.1.4 - The tuning update:
+150 ELO self-play (1s / move)
Increased Futility and Reverse Futility Pruning Depths
Tweaked LMR
Cleaner, easier to understand code
Researches null window first before searching full window
Added Second Set of Killer Moves
Added King Safety in the form of King Tropism
Extra bonus to diagonals in line with the king
Extra bonus to attack if enemy king is near semi-open files
Weighted by attacker's material
Evaluation tuned using a logistic regression over the zurichess-quiet-labeled dataset, similar to the Texel Method
Black box tuning was done using Simulated annealing + local search, using a pseudo-huber loss
pseudo-huber loss was used here since there are likely outliers that would unfavorably skew the relatively simple evaluation function. This was a choice I made based on what I understood about the dataset, and made a marginal improvement to the evaluation quality as opposed to the traditional MSE loss (+10ish elo from 2000 games). If my evaluation were more complex, I might be more tempted to stay with MSE loss, as long as on-board checkmates are removed from the dataset
Some experimentation is also being done with a custom constructed dataset, if it performs better there will be a v1.4.1
Fixed some timeout bugs
Increased hash table stability
Comments
Post a Comment