CCRL Benchmark Net

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
dkappe
Posts: 260
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

CCRL Benchmark Net

Post by dkappe » Mon Aug 26, 2019 5:06 am

I can never lay my hands on a CCRL data trained net when I want one. I've made a 128x10 classical value head and convolutional policy head net available. https://github.com/dkappe/leela-chess-w ... i/CCRL-Net

It looks to be about 200 elo weaker than Bad Gyal 7, which is surprising to me.

Max
Posts: 242
Joined: Tue Apr 13, 2010 8:41 am

Re: CCRL Benchmark Net

Post by Max » Mon Aug 26, 2019 10:20 am

Hi Dietrich,
any explanation why your small girls/gyals are so much faster than the distilled networks at the same size?

On CPU: ./lc0 benchmark --minibatch-size=8 --max-prefetch=0

Code: Select all

 nps	size     net
--------------------------
7850  	32x4 meangirl-7	
4700  	48x5 bad-gyal-1
      	48x5 bad-gyal-2
      	48x5 evilgyal-6
4000	16x2 11258-16x2-se
3500	24x3 11258-24x3-se
3020	32x4 11258-32x4-se
2800	64x6 bad-gyal-3
2475	48x5 11258-48x5-se
1780	64x6 11258-64x6-se
Hope we're not just the biological boot loader for digital super intelligence. Unfortunately, that is increasingly probable - Elon Musk

dkappe
Posts: 260
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

Re: CCRL Benchmark Net

Post by dkappe » Mon Aug 26, 2019 3:01 pm

The architecture is different, but other than that, no idea.

dkappe
Posts: 260
Joined: Tue Aug 21, 2018 5:52 pm
Full name: Dietrich Kappe

Re: CCRL Benchmark Net

Post by dkappe » Mon Aug 26, 2019 4:41 pm

dkappe wrote:
Mon Aug 26, 2019 5:06 am

It looks to be about 200 elo weaker than Bad Gyal 7, which is surprising to me.
Make that 250 elo. I think that this is a good piece of evidence that the SL tools surrounding lc0 yield suboptimal results.

Post Reply