TalkChess.com

a) look here https://lczero.org/play/configuration/flags/
b) help Channel on discord https://discord.gg/trkUZqaZ
b) change settings and see difference in taskmanager

dkappe wrote: ↑

Sun Jan 10, 2021 10:01 pm

I added an avx2 Windows build with the shared libnnueprobe.so included. Works under wine, but I have no other way of testing.

Hi Dietrich, I got it working under Windows with Arena 3.5.1.
Thanks for the .exe
best wishes
Werner

Sorry -I DO NOT UNDERSTAND THIS QUESTION !

I am in contact with Daniel, and I test his engines when he asks me. He has already results with his newest net

Hi all, our actual rating lists are online and can be found under the attached links! 40 / 20: New games: 1.740; 21 different engines Total: 1.427.353 NEW Engines 226 Wasp 4.50 x64 1CPU : 3090 - 1030 games (+69 to v. 4.00) UPDATES 255 Arasan 22.2 x64 1CPU : 3070 - 1200 games (-4, close to v. 22.1) 1...

Hi all, our actual rating lists are online and can be found under the attached links! 40 / 20: New games: 3.288; 44 different engines Total: 1.425.613 NEW Engines 245 Arasan 22.2 x64 1CPU : 3074 - 1000 games (close to v. 22.1) 620 Francesca MAD 0.30 : 2818 - 1000 games (+19 to v. 0.29) 104 Igel 2.9....

I used 2 CPU threads for LC0 as set in lc0.config. I hope Ceres is using the same configuration. The taskmanager is showing the same CPU usage here 12% from 8 threads (I have 4CPU)

For all users of a GTX Card there is no improvement in speed. I made an example with my GTX 1650 running Ceres with LS15 and LC0 0.26.3 with LS15 under Arena 3.5.1 using a position from old WM test: with Ceres,LS15, GTX 1650 nps 365 FEN: rnbr2k1/pp3p2/1qp3pp/4Q1N1/4P3/2P3P1/PP3PBP/R3R1K1 w - - 0 1 8...

Sorry ,
the cuda drivers are older only, the lc0 version is the same:
cublasLt64_11.dll 214.235.648 16.09.2020 05:17 -a-- in the CUDA package
or
cublas64_100.dll 67.318.784 26.08.2018 06:33 -a-- in the CUDNN package

clear now?

I tested the 2 versions on my GTX 1650 Super under the same conditions with go movetime 60000; NN 10x128 LC0 Cuda 0.26.3 backend cuda nps 19247 LC0 Cuda 0.26.3 backend cuda-fp16 nps 6634 LC0 CUDNN 0.26.3 backend cudnn nps 22990 LC0 CUDNN 0.26.3 backend cudnn-fp16 nps 37701 LC0 CUDNN 0.26.3 backend c...

I have released Francesca 0.30. Available from www.silentshark.co.uk Not a huge jump in strength, maybe 30 ELO or so. Changes Small changes. Minor optimisations. More aggressive null move pruning. IID replaced by IIR (thanks to Ed Schroder for idea). Tweak to futility to make more aggressive or not...