Hier die Bekanntmachung auf Discord:
Announcing CeresTrain - a platform for training deep neural networks for chess. As described on the project page, focus of the initial release has been endgames for which tablebases exist. Some empirical findings are described which seem at least mildly interesting and encouraging.
More generally the primary goals of the project are to:
* facilitate research into chess neural networks providing a modular set of flexible software building blocks which span the subtasks of training data collection, neural network architectural definition, optimization, network evaluation, introspection, and integration into MCTS chess engines such as its sister project (Ceres, 2020)
* provide a high performance platform for training of chess neural networks at scale, including a two parallel and interoperable backend implementations (based on C# via TorchSharp and one in Python) which leverage PyTorch 2.0 in an (optionally) distributed setting
As always, critical feedback and suggestions will be greatly appreciated from the community. Please be aware that the documentation of installation/configuration is likely imperfect, particularly when using the Linux/Pytorch backend.
<
https://github.com/dje-dev/CeresTrain>