Not logged inCSS-Forum
Forum CSS-Online Help Search Login
CSS-Shop Impressum Datenschutz
Up Topic Hauptforen / Schachprogrammierung / Hybride Engines
- By Lothar Jung Date 2022-01-14 17:41
Hier eine Ansatz - mit Programmcode -  A/B-Suche und DeepLearning:

https://towardsdatascience.com/implementing-a-deep-learning-chess-engine-from-scratch-275e5c9b9e21
- By Lothar Jung Date 2022-01-14 17:45
Hier ein Projekt mit hybrider Bewertung, die Engine Chrysaora:

https://github.com/dylanagreen/chrysaora
- By Lothar Jung Date 2022-01-14 17:49 Edited 2022-01-14 17:52
Hier das Projekt der Kaiss64 Engine, minimax Suche und neurales Netzwerk:

„Hybrid (minmax + neural nets) Chess Engine for Research and Fun.

I'm using Python chess library for research and C++ parts to implement NNUE nets.“

https://github.com/gotzmann/kaissa64
- - By Lothar Jung Date 2022-03-07 11:48 Edited 2022-03-07 11:52
Die HybridEngine A/B-Helper von Hans Eckbrandt vereinigt die MCTS und die A/B-Suche in Lc0.

Hier der GitHub Link:

https://github.com/hans-ekbrand/lc0/wiki/Leela-taking-advantage-of-a-helper-engine

Hier ein Thread über die Diskussion im Forum:

https://forum.computerschach.de/cgi-bin/mwf/topic_show.pl?tid=12816
Parent - - By Jörg Oster Date 2022-03-07 12:10 Upvotes 1
Da würde ich sagen, das stimmt so nicht.
Er benutzt ja eine AB-Helper-Engine und vereinigt nicht MCTS und AB-Suche in Lc0.
Dann würde er ja keine externe Engine benötigen.
Parent - By Lothar Jung Date 2022-03-07 12:28 Edited 2022-03-07 12:33
Ich würde sagen, er modifiziert Lc0 29 dev, um die Suchergebnisse der A/B- Helper Engine in die CPUCT Suche von Lc0 einfließen zu lassen.

Siehe GitHub Einführung:

„When Vanilla Leela decides which edges in the search tree to extend, she traverses the tree from root and applies the CPUCT formula to decide which branch to go down, until a leaf is found and the edge is then extended, i.e. the node that represents the position that the move represented by the edge leads to is created, and unless that node is terminal or already in the cache the neural net is queried about that node. This version of Leela combines that way of choosing edges to extend with another way: edges representing moves recommended by another engine are also extended.

A small fraction of all nodes in Leelas search tree are sent to an auxillary helper (via UCI), and the returned principal variation(s) are used to choose edges to extend in Leelas search tree. For the system to be effective, the auxillary helper should be a CPU-based engine, and Leela should be running on a GPU.“
Up Topic Hauptforen / Schachprogrammierung / Hybride Engines

Powered by mwForum 2.29.3 © 1999-2014 Markus Wichitill