TalkChess.com

Found this info: Regarding our submission to SuFi: • Given that it's unlikely, that Lc0 in the current configuration would win SuFi (don't let DivP fool you), it's quite likely that we'll send pr1483 binary there. • PR 1483 is a port of one of Ceres features (parallel node gathering), implemented b...

Does it play stronger? The reason I ask is that Ceres did not do well in Pohl's testing. This is not Ceres, just Leela. https://github.com/LeelaChessZero/lc0/pull/1483 I dont know which built Pohl used. And there is no official release of Ceres yet. Crem's fixed node test shows around 5 elo loss.

1rb3k1/rp1p1ppp/pPpP4/P1P5/8/8/8/K5BB w - - 0 1 Hopefully the following position is good enough. sting also blunder by bxa7 try to play with your favourite engine with both sides. I think that it can be a good starting position for engine-engine games to see the strong engines that know to avoid bx...

This New Leela version (partial implementaion of Ceres idea ) will be playing in this Sufi! S20 DivP game 188 original and replay: moveno: DivP -> PR1483 8: 97976 -> 168736 +72% 9: 90965 -> 194676 +114% 10: 89835 -> 196105 +118% 11: 84191 -> 94452 +12% 12: 64423 -> 91580 +42% 13: 71462 -> 109162 +53...

I am quite happy with Ceres speed on my 1060GTX card.
17knps on 703810.pb.gz network on Leela vs 25knps on Ceres. Nearly 50% faster speed!

Collingwood wrote: ↑

Sun Jan 03, 2021 7:03 am

Nay Lin Tun wrote: ↑

Sun Jan 03, 2021 6:36 am

1.It is not full details of codes

What does this mean? Does the github dump not compile and run?

It doesn't mention about all details of A0 training. You can get a running version of weak A0.

1.It is not full details of codes 2. It is a very good documentary record on revolution of early neural networks 3. Current Lc0 running on TCEC is 5x times larger NN weights than A0 ( 30x386 vs 20x256) with much more advanced features than A0 ( e.g TB probe, TB scoring of value head, MLH etc) and li...

Prabably 21. Bf3? was the blunder as suggested my droidfish running 10-15 seconds per position. :P Maybe there are other alternatives but I think 21.Bf3 move doens't deserve a "?" tag because it is also a strong move. 29/45 +0.21 Qe1 Bf5 Bf3 Rg8 Re5 Bh3 Nf1 Qg7 Ng3 fxg3 hxg3 Bxe5 Qxe5 Rae8 Qxg7+ Rx...

Now SF classic with 232 cores has much lower nps than Komodo 128 cores. Something is wrong now :!: :?: . It looks like you dont understand actual physical cores vs hyperthreads.( 2x Threads=1 core). 232÷2= 116 cores Stockfish is not running on 232 cores, they dont have 232 cores. Komodo is running ...

Announcement from Leela discord. Long time community member and developer @dje has created a new chess engine that uses Leela Chess Zero networks. It appears much stronger than lc0 (see the performance test below). The build process is currently being worked on so please be patient while the commun...