-
Currently, Monika uses a chess engine named Stockfish for playing chess with the player. While it is a good engine, it is not suitable for Monika for these reasons:
1. Stockfish is a top level engi…
-
https://www.youtube.com/watch?v=OP_3h__MleY
24min Demis Has****s
Demis Has****s about AlphaZero at 34:36 : "we would need to play under tournament conditions to know for sure". If he says this, I …
-
hardware: Intel i5 Quad - 4 CPU´s, 4 GB RAM, AVX, POPCNT and SSE4.2 is supported, but not BMI2, 4 GB RAM, Windows 7 64 bit SP1; Cuda, Blas. OpenCL or anything else? GUIs: WinBoard Interface, Arena etc…
-
The bottom elo on [ccrl 40/4](http://www.computerchess.org.uk/ccrl/404/) is only 276 elo and sometimes loses to a random mover, but it requires java.
A good first target might be [chessputer](https:/…
-
Download client.exe here:
https://github.com/glinscott/leela-chess/releases/download/v0.2/client.exe
Then download leela go here and take out the openblas and opencl dll files:
https://github.com/g…
-
Since the use of ELF network seems to yield so good result why don't we use even other networks.
For example we could keep training the 20b network from @bjiyxo and when is better than the current, …
-
Recently leelaz is looks like to stall again, which seems too early than 5 block and 6block.
I'm in doubt if fixed playout is better than fixed visit count.
We all know that as subtree is reused…
-
It's not 100% tailored to Leela's workload, but the article is still useful.
https://www.phoronix.com/scan.php?page=article&item=nvidia-rtx2080ti-compute
ghost updated
6 years ago
-
I downloaded the client.exe v0.16.0 for running on Windows 10 (1803). I dropped it in a new folder. Upon running the file I was prompted to enter credentials and it then proceeded to download networks…
-
Leela Chess had from the beginning of its implementation set up two parameters policy_loss_weight and value_loss_weight in training.py for the purpose of adjusting the total training loss term relativ…