![]() See the "Automatic avoid patterns" section in cpp/configs/gtp_example.cfg. These are changes relevant to users running bots online or katago internal test matches.Īdded support for automatically biasing KataGo to avoid moves it played recently in earlier games, giving more move variety for online bots. To override this, you can set the config parameter minPlayoutsPerThread. the 1-1 point).įor GTP and analysis, KataGo will automatically cap the number of threads at about 1/8th the number of playouts being performed, to prevent the worst cases of accidentally misconfiguring KataGo to use many threads being destructive to search quality when testing KataGo with low settings. Softplus error scaling - supports new squared softplus activations for value and score error predictions, as well as adjusted scaling of gradients and post-activation for those predictions, which should fix some rare outliers in overconfidence in these predictions as well as large prediction magnitudes that might result in less-stable training.įixed a bug with determining the baseline top move at low playouts for policy target pruning, which could cause KataGo at low playouts on small boards to sometimes play extremely bad moves (e.g. A one-off neural net using this policy head is attached below, KataGo's main nets at will begin including the new head soon as well. Optimistic policy - improved policy head that is biased to look more for unexpectedly good moves. If you have this issue, you may be able to work around it by compiling from source, which is usually not so hard on Linux, see the "TLDR" instructions for Linux here. Some users have encountered issues with libzip or other library compatibility issues in the past. ![]() The Linux executables were compiled on a 20.04 Ubuntu machine. They may also be slower and will use much more memory, even when only playing on 19x19, so you should use them only when you really want to try large boards. These are just for fun, and don't support distributed training but DO support board sizes up to 29x29. This release also features a somewhat outdated but stronger net using a new "optimistic policy" head in v1.13.0, attached below, the latest nets at will also start including this improvement soon.Īttached here are "bs29" versions of KataGo. ![]() KataGo is continuing to improve at and if you'd like to donate your spare GPU cycles and support it, it could use your help there!įor the TensorRT version, download it from v1.13.1 which is a quick bugfix release specific to the TensorRT version and which should only matter for users doing custom builds, but for clarity has been minted as a new release. If you're a new user, this section has tips for getting started and basic usage! If you don't know which version to choose (OpenCL, CUDA, TensorRT, Eigen, Eigen AVX2), see here. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |