LCZero Network question

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
User avatar
Eduard
Posts: 224
Joined: Fri Oct 26, 2018 10:58 pm
Location: Germany
Full name: Eduard Nemeth
Contact:

LCZero Network question

Post by Eduard » Tue Feb 19, 2019 8:57 am

I downloaded the network 36089. It only has 11 MB. That's not much more than destilled networks, running well on CPU only! Question: Why is the network so small? (Normally these networks have around 40 MB)

Btw:
CCRL Rating list: Lc0 w36089 1 CPU 3022 Elo!

Not bad.
Last edited by Eduard on Tue Feb 19, 2019 9:26 am, edited 1 time in total.

User avatar
Guenther
Posts: 3109
Joined: Wed Oct 01, 2008 4:33 am
Location: Regensburg, Germany
Full name: Guenther Simon
Contact:

Re: LCZero Network question

Post by Guenther » Tue Feb 19, 2019 10:26 am

Eduard wrote:
Tue Feb 19, 2019 8:57 am
I downloaded the network 36089. It only has 11 MB. That's not much more than destilled networks, running well on CPU only! Question: Why is the network so small? (Normally these networks have around 40 MB)

Btw:
CCRL Rating list: Lc0 w36089 1 CPU 3022 Elo!

Not bad.
The short-lived NN35 series were just 128*10 blocks and filters.
(*18-12-10 †19-01-07 - last one 36092)

http://lczero.org/networks/
Current foe list count : [101]
http://rwbc-chess.de/chronology.htm

User avatar
Eduard
Posts: 224
Joined: Fri Oct 26, 2018 10:58 pm
Location: Germany
Full name: Eduard Nemeth
Contact:

Re: LCZero Network question

Post by Eduard » Tue Feb 19, 2019 11:30 am

Thanks, that's interesting. On only CPU running show my first tests in analysis mode that these networks (36088 and 36089) could be better than the distilled networks 11258.

User avatar
Eduard
Posts: 224
Joined: Fri Oct 26, 2018 10:58 pm
Location: Germany
Full name: Eduard Nemeth
Contact:

Re: LCZero Network question

Post by Eduard » Tue Feb 19, 2019 12:02 pm

Here for example (but I can post more). After 5 minutes thinking:



Analysis by Lc0 v0.20.2-rc1 11258 (Distilled) 112x9:

18...c4 19.b3 Sxd7 20.Lxd7 Lf6 21.Tac1 Sb4 22.bxc4 dxc4 23.Lxe6 fxe6 24.Ld6 Sd3 25.Tc2 Lxc3 26.Txc3 Txa2 27.Txc4 Sxf2 28.Tf1 Se4
+/- (0.77) Tiefe: 10/24 00:04:46 47kN
18...c4 19.b3 Sxd7 20.Lxd7 Lf6 21.Tac1 Sb4 22.bxc4 dxc4 23.Lxe6 fxe6 24.Ld6 Sd3 25.Tc2 Lxc3 26.Txc3 Txa2 27.Txc4 Sxf2 28.Tf1 Se4
+/- (0.77) Tiefe: 10/24 00:04:51 48kN
18...c4 19.b3 Sxd7 20.Lxd7 Lf6 21.Tac1 Sb4 22.bxc4 dxc4 23.Lxe6 fxe6 24.Ld6 Sd3 25.Tc2 Lxc3 26.Txc3 Txa2 27.Txc4 Sxf2 28.Tf1 Se4
+/- (0.77) Tiefe: 10/24 00:04:56 49kN

18...c4? is bad, it loses after 19. e4!.

Analysis by Lc0 v0.21.0-rc1 w36088:

18...Sh5 19.Sxd5 Lxd5 20.Txd5 Sxg3 21.Lxa6 Txa6 22.hxg3 f6 23.a3 Ta7 24.Tad1 Kf7 25.Kf1 Tc6 26.Ke2 c4 27.Sb8 Tc8 28.Sd7
+/= (0.59) Tiefe: 13/32 00:04:53 30kN
18...Sh5 19.Sxd5 Lxd5 20.Txd5 Sxg3 21.Lxa6 Txa6 22.hxg3 f6 23.a3 Ta7 24.Tad1 Kf7 25.Kf1 Tc6 26.Ke2 c4 27.Sb8 Tc8 28.Sd7
+/= (0.57) Tiefe: 13/32 00:04:58 31kN
18...Sh5 19.Sxd5 Lxd5 20.Txd5 Sxg3 21.Lxa6 Txa6 22.hxg3 f6 23.a3 Ta7 24.Tad1 Kf7 25.Kf1 Tc6 26.Ke2 c4 27.Sb8 Tc8 28.Sd7
+/= (0.56) Tiefe: 13/32 00:05:03 31kN

18...Nh5 is here one of the best moves.

But I am surprised about the depth. Distilled network 11258 after 5 minutes only in depth 10/24, w36088 in depth 13/32! How can that be? :roll:

User avatar
Eduard
Posts: 224
Joined: Fri Oct 26, 2018 10:58 pm
Location: Germany
Full name: Eduard Nemeth
Contact:

Re: LCZero Network question

Post by Eduard » Tue Feb 19, 2019 9:16 pm

Can someone explain me what the difference is between the 50xxx networks with 128x10 (eg NN 50052) and the 30xxx networks 128x10 (eg NN 36089)?

yanquis1972
Posts: 1762
Joined: Tue Jun 02, 2009 10:14 pm

Re: LCZero Network question

Post by yanquis1972 » Tue Feb 19, 2019 9:50 pm

i think 35xx+ were precursor training for t40. it was fully trained (all 4 LR drops) & in some way used to initialize T40 training. it should be much better on CPU than 256x20 nets since it's both mature & fast.

t50 i don't know anything about; it's got some new parameters & uses 10000 visits instead of the prescribed 800. i assume it's experimental in preparation for t60 (but T30 was an experimental run as well). it's very young & several hundred elo below T40 atm, even at 1'+1s TC.

brianr
Posts: 358
Joined: Thu Mar 09, 2006 2:01 pm

Re: LCZero Network question

Post by brianr » Tue Feb 19, 2019 10:07 pm

Eduard wrote:
Tue Feb 19, 2019 12:02 pm
But I am surprised about the depth. Distilled network 11258 after 5 minutes only in depth 10/24, w36088 in depth 13/32! How can that be? :roll:
Depth for Leela is an artificial estimate and not really like a/b engines.

User avatar
Eduard
Posts: 224
Joined: Fri Oct 26, 2018 10:58 pm
Location: Germany
Full name: Eduard Nemeth
Contact:

Re: LCZero Network question

Post by Eduard » Tue Feb 19, 2019 11:28 pm

I understand, thank you!

Post Reply