Latest Updates:
Very Hot Topic (More than 25 Replies) New version of LcO (Read 12326 times)
Jupp53
God Member
*****
Offline


be

Posts: 920
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: New version of LcO
Reply #72 - 11/19/19 at 14:21:22
Post Tools
The german newspaper ct' has in its current edition (24/2019; till this weekend) a proposal for a silent pc fitting to these ideas. I think if you don't want to play competitive correspondence chess on highest level one RTX 2070 or 2060 model are sufficient for having good support in working for yourself.

The pc is cheaper with two GPUs than my first one 36 years ago and I had less money than. So power efficiency for the climate is my main concern meanwhile.
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Back to top
 
IP Logged
 
JFugre
Junior Member
**
Offline


I Love ChessPublishing!

Posts: 81
Joined: 01/22/19
Re: New version of LcO
Reply #71 - 11/18/19 at 23:11:50
Post Tools
IsaVulpes wrote on 11/18/19 at 17:10:13:
RTX2060 - 18500 NPS - $320
GTX1660ti - 8500 NPS - $260
+117% performance, for +23% price


The GTX 1660 Ti is incredibly cost inefficient, no matter the application (Leela or games). In Europe, it seems to cost almost as much as the RTX 2060.

The non-Ti version is 40% cheaper. But he was looking for a laptop so the options are always rather different then.

The important thing is to not get a non-Turing card (i.e. GTX 10x0 series), or you'll pay more for ~25% less performance. RTX is nice if you can afford it.

The ChessBase guide for system building is hilarious. Building a system with 2 x RTX 2070 and a 2700X CPU? It will be hot, power hungry and a nightmare to keep quiet. Useful for showing off on playchess or in correspondence, not so useful if you actually study for OTB.
  
Back to top
 
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #70 - 11/18/19 at 17:10:13
Post Tools
IsaVulpes wrote on 11/09/19 at 23:15:56:
General benchmarks are of very little use for a comparison between RTX cards and others for Leela purposes

The RTX cards are boosted with tensor cores, which are an incredible performance boost for all machine learning applications.

RTX 2060 and GTX 1660i are at a similar pricepoint, but the former should bring out vastly more power for Leela (whether anyone needs that, is of course a different question).
RTX 2080 will outperfom any GTX card by miles.


https://en.chessbase.com/post/fat-fritz-what-videocard-to-buy Here is a chessbase article underlining what I said

RTX2060 - 18500 NPS - $320
GTX1660ti - 8500 NPS - $260
+117% performance, for +23% price

If you want to buy a GPU for Leela, get an NVidia RTX card.
  
Back to top
 
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #69 - 11/09/19 at 23:15:56
Post Tools
General benchmarks are of very little use for a comparison between RTX cards and others for Leela purposes

The RTX cards are boosted with tensor cores, which are an incredible performance boost for all machine learning applications.

RTX 2060 and GTX 1660i are at a similar pricepoint, but the former should bring out vastly more power for Leela (whether anyone needs that, is of course a different question).
RTX 2080 will outperfom any GTX card by miles.
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #68 - 10/21/19 at 08:35:33
Post Tools
I finally bought a new laptop:  https://www.alternate.be/ASUS/TUF-Gaming-FX505DU-AL057T-BE/html/product/1555008?...

I estimate that the Nvidea GTX 1660i is about 40 elo weaker than the recommended Nvidea RTX 2080 see https://gpu.userbenchmark.com/Compare/Nvidia-RTX-2080-vs-Nvidia-GTX-1660-Ti/4026...

However I didn't want to spend 2-3 times more money for such small gain.

I compared the new portable with my old desktop (My AMD FX(tm)-6300 6 core processor 3,5Ghz is configured with a Nvidia Geforce GTX 960.) and noted the following gains of speed:
Komodo 11 and Stockfish 10: + 70%
Lc0: + 1000% (using Cuda instead of opencl)

I organized 2 new 100 rapidmatches (Lc0 used net 42850)
Lc0 - Komodo 11: 62,5 - 37,5
Lc0 - Stockfish 10: 52,5 - 47,5

Below just one amazing game from Lc0 against Stockfish:

The return-game was easily drawn by Lc0.
  
Back to top
WWW  
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #67 - 09/20/19 at 17:37:44
Post Tools
Ok. Sorry for commenting. Good luck in the future!
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #66 - 09/20/19 at 05:56:43
Post Tools
The questions are all answered in my article.

The matches with Lc0 were all organized in the months May, June and July with the release which was at that time available (see dates of the games). I know around that period they were bringing out new engine releases which were each time decreasing the evaluation.

I also write in my article that in my last match I was using the net 42820. 42850 was not available at that time yet but I expect very little or no difference at all in strength between both.

Finally don't forget that the matches were played at a rapidrate. Many well played games were spoiled by Lc0 as it was running out of time and was just living from the 10 seconds increment for many moves which clearly hurt the performance.

In both my articles I write that I would love to test Lc0 on stronger hardware but financially I can't/ won't in the nearby future. Donations are of course always welcome.
  
Back to top
WWW  
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #65 - 09/19/19 at 22:13:26
Post Tools
Furthermore, in the third game, after 17...Bxe5, 18. dxe5 is not that highly evaluated by Lc0, although the pv. But once the move is played, the evaluation at depth 27 goes to about the same advantage as SF10 at depth 35. But, yes, I agree, here Lc0 is clearly weaker than SF10.

I guess my point is that on stronger hardware and with the latest T40 net, Lc0 would likely do even better than in your match.
  
Back to top
 
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #64 - 09/19/19 at 17:35:56
Post Tools
Interesting analysis!

What net have you been using for LC0?

With Lc0 v0.22.0 and net 42850, at depth 20, the position after 11...c5 is evaluated as +0.55, and at lower depth it was never more than +0.69. After 30. Ng3, I get more around +-2.60. Do you have the evaluation set to centipawns?
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #63 - 09/19/19 at 14:58:51
Post Tools
brabo wrote on 09/19/19 at 06:47:46:
I wrote a second article on my blog which should normally erase any doubts why we all should use neural networks (in this case Lc0): https://chess-brabo.blogspot.com/2019/09/testing-chess-engines.html

very interesting games, thanks
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #62 - 09/19/19 at 06:47:46
Post Tools
an ordinary chessplayer wrote on 09/12/19 at 13:26:05:
brabo wrote on 09/11/19 at 17:20:56:
I finally found some time to write a small review on my blog of why we should use Lc0: https://chess-brabo.blogspot.com/2019/09/leela-lc0.html

Thanks, there's a lot of ideas in that blog post, but to my mind it does not quite explain why we should use Lc0. We can construct several related propositions:
  1. We should prepare.
  2. We should prepare with an engine.
  3. We should prepare with the latest engine.
  4. We should prepare with a specific engine.

(In the above chain, for "engine" we could substitute "book" or "database"; for the professionals additionally "second" or "team".)

For example, I think in your game against Guy Baete, you could have won by preparing the same opening without an engine, just using the database, so I don't see it as an argument in favor of Lc0 per se.

I wrote a second article on my blog which should normally erase any doubts why we all should use neural networks (in this case Lc0): https://chess-brabo.blogspot.com/2019/09/testing-chess-engines.html
  
Back to top
WWW  
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #61 - 09/15/19 at 05:33:05
Post Tools
Keano wrote on 09/14/19 at 18:32:59:
MartinC wrote on 09/13/19 at 20:30:18:
LC0 simply won't be running very well if its on a laptop CPU.


why not?

Because the entire engine architecture is different, and the type of operations that neural nets (as opposed to classic engines like Stockfish) engage in are much more efficiently handled by a GPU than a CPU.

For Leela to show her "true" power, you need something upwards of an RTX2060 card - usable in general, especially if you let it run for a while, it is of course no matter your hardware (I have the LD2 net on my phone), but it will be significantly weaker - and most notably, much *more* significantly weaker than Stockfish becomes with dropping hardware levels.

Meaning: You can't really give an opinion on the engine's evaluations if your hardware is so poor that it barely runs. If I somehow manage to install Stockfish on my Nokia 3330, and it gets 1 node/second, going on record with "This engine is actually quite weak" becomes a bit silly.
  
Back to top
 
IP Logged
 
Keano
God Member
*****
Offline


Money doesn't talk, it
swears.

Posts: 2891
Location: Toulouse
Joined: 05/25/05
Gender: Male
Re: New version of LcO
Reply #60 - 09/14/19 at 18:32:59
Post Tools
MartinC wrote on 09/13/19 at 20:30:18:
LC0 simply won't be running very well if its on a laptop CPU.


why not?
  
Back to top
 
IP Logged
 
trw
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1319
Joined: 05/06/08
Gender: Male
Re: New version of LcO
Reply #59 - 09/14/19 at 18:14:55
Post Tools
It's been the standard for over 10 years to use two engines at once... though I have never seen or used a computer strong enough to run three at any real depth.

Anyhow, disagreement usually just means that the position is sufficiently complicated and warrants more investigation... Not that one is always right.
  
Back to top
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #58 - 09/14/19 at 11:29:09
Post Tools
One of the ideas I gleaned(stole) is using three engines at the same time say Stockfish, Komodo, and LcO. If any two agree that is the best choice. If all three differ use LcO's choice. Which I guess is an argument for a desktop.
  
Back to top
 
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #57 - 09/14/19 at 09:51:17
Post Tools
It is not the same, but maybe somewhat similar to a tiny Smart car with Hayabusa engine (very powerful engine from motorcycle). The engine has lots of potential, but the steering, traction etc are not. Imagine racing with such a car and trying to make sharp turns in high speed, like real racing cars can.

In this case the net would correspond to the engine and your laptop to the Smart car I guess. When you compare Lc0 to Komodo, it doesn't say much of the true potential of the Lc0.
  
Back to top
 
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #56 - 09/13/19 at 21:49:12
Post Tools
Your best bet will be the LD2 net, if you're on a slow machine without decent GPU @Leon

It's a smaller net, so it's notably worse for deep analysis type business, and retains a lot of tactical mishaps, but it's general evaluation is similar to T40, and it will be stronger on a weak machine.
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1982
Joined: 07/24/06
Re: New version of LcO
Reply #55 - 09/13/19 at 20:30:18
Post Tools
LC0 simply won't be running very well if its on a laptop CPU.
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #54 - 09/13/19 at 19:39:25
Post Tools
I usually make Komodo 10 make moves against Leela and try to see what happens. Many times the evaluation goes down when Komodo plays Leela in hypodermic (Modern/Pirc) or blocked (positional Winawer) lines. Quite often Leela can go down by half, sometimes even to only +0,20.

Or maybe Leela has a high White evaluation because it runs slower on my laptop before it can see Black's counterplay ¿ Who knows.
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #53 - 09/13/19 at 16:04:51
Post Tools
an ordinary chessplayer wrote on 09/13/19 at 15:05:38:
an ordinary chessplayer wrote on 09/12/19 at 19:40:39:
Trying to make new preparations shortly before a game just leads me to play worse.


brabo wrote on 09/13/19 at 06:56:08:
I can show you many of my won games in which I can prove that they were won largely due to new preparations made just before the games.


I didn't mean that new opening preparations make me play worse in the opening. They make me play worse "overall". I do a rather sophisticated errors-by-phase-of-game meta analysis of my play, and stuffing my brain right before a game leads to later errors. During my school-days, cramming for a test was also a net negative strategy, despite obviously leading me to answer some specific questions correctly.

Winning a lot of opening games due to preparation is fine if you don't lose any additional middle-games or endgames due to the same preparation. But for some reason, I do. My best hypothesis is that intensive opening study messes with the evaluation function somehow. This is entirely aside from the missed opportunity for relaxation before a game, which is a big contributor to endgame play.

Unless I am woozy before the game, I *never* lose the opening. And I actually like it when strong players have obviously prepared against me. I try to verify that in the post-mortem. My performance rating in these encounters is better than when they just play chess.

I find it all very weird what you experience. The endgame is for me my strongest phase so I definitely don't feel any need to decrease my hours of preparations. Also I found it quite normal at the university to study 12 hours or more per day during the exams. I obtained my engineer degree with high honors, never having failed any exam.

It is always dangerous to give advise based on own experiences but what else can we do. So I just say what works for me and think it could be useful for somebody else.
  
Back to top
WWW  
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 729
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: New version of LcO
Reply #52 - 09/13/19 at 15:05:38
Post Tools
an ordinary chessplayer wrote on 09/12/19 at 19:40:39:
Trying to make new preparations shortly before a game just leads me to play worse.


brabo wrote on 09/13/19 at 06:56:08:
I can show you many of my won games in which I can prove that they were won largely due to new preparations made just before the games.


I didn't mean that new opening preparations make me play worse in the opening. They make me play worse "overall". I do a rather sophisticated errors-by-phase-of-game meta analysis of my play, and stuffing my brain right before a game leads to later errors. During my school-days, cramming for a test was also a net negative strategy, despite obviously leading me to answer some specific questions correctly.

Winning a lot of opening games due to preparation is fine if you don't lose any additional middle-games or endgames due to the same preparation. But for some reason, I do. My best hypothesis is that intensive opening study messes with the evaluation function somehow. This is entirely aside from the missed opportunity for relaxation before a game, which is a big contributor to endgame play.

Unless I am woozy before the game, I *never* lose the opening. And I actually like it when strong players have obviously prepared against me. I try to verify that in the post-mortem. My performance rating in these encounters is better than when they just play chess.
  
Back to top
 
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #51 - 09/13/19 at 14:39:29
Post Tools
JFugre wrote on 09/13/19 at 09:21:58:
fling wrote on 09/13/19 at 08:27:10:
RTX 2060, it is not supposed to be great with the larger nets.

I think you should try one of the nets that are claimed to be better for CPU in case you want to use Lc0 on your laptop.


You have to take into account that there is a scaling issue here. A faster GPU is the same as letting the engine think longer, i.e. a longer timecontrol. Small nets are good for "blitz" because the engine needs to be able to do some minimal searching to see enough tactics. But if the larger nets are better on faster hardware, that also means they are better for longer analysis.


Well, no. It was a writing mistake. It should've been RTX 960, sorry.

Better is sometimes a bit hard to define. It might work better for longer analysis with a decent analysis rate. But at something like 1 knps, the analysis might take too much time in practice, i.e. it is not that good anyway.
  
Back to top
 
IP Logged
 
JFugre
Junior Member
**
Offline


I Love ChessPublishing!

Posts: 81
Joined: 01/22/19
Re: New version of LcO
Reply #50 - 09/13/19 at 09:21:58
Post Tools
fling wrote on 09/13/19 at 08:27:10:
RTX 2060, it is not supposed to be great with the larger nets.

I think you should try one of the nets that are claimed to be better for CPU in case you want to use Lc0 on your laptop.


You have to take into account that there is a scaling issue here. A faster GPU is the same as letting the engine think longer, i.e. a longer timecontrol. Small nets are good for "blitz" because the engine needs to be able to do some minimal searching to see enough tactics. But if the larger nets are better on faster hardware, that also means they are better for longer analysis.
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1982
Joined: 07/24/06
Re: New version of LcO
Reply #49 - 09/13/19 at 08:55:49
Post Tools
Leon_Trotsky wrote on 09/12/19 at 22:21:03:
Very strategic/locked type positions and hypermodern openings like Pirc and Modern, I find that Leela overestimates, sometimes to a big degree, Black's chances.

In my experience, in openings like the Pirc and French Winawer where Black deliberately cedes space and assumes a "turtle" approach, it gives outrageously high White evaluations, like +1,30, when Komodo would give around 0,00 to 0,10.


I presume you mean underestimate? It doesn't like the Pirc that much, no (or the Modern Benoni come to it). It does seem to be utterly brilliant at handling 'classical' advantages, building up the positions and squashing tactics so it probably backs that up!

It has a very 'unpleasant' habit of reducing seemingly fun lines to two result style pain.

The Winaver its been fairly reasonable about when I've looked.

Leon_Trotsky wrote on 09/12/19 at 22:21:03:
Except that I have no desktop. I have not had a desktop since around 1991 or 1992  Cheesy


Which is fine, its just that LC0 won't run at a worthwhile speed on a laptop unless you get one with a mobile GPU of course, but that does cost.
  
Back to top
 
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #48 - 09/13/19 at 08:55:46
Post Tools
Leon_Trotsky wrote on 09/12/19 at 22:21:03:
In my experience, in openings like the Pirc and French Winawer where Black deliberately cedes space and assumes a "turtle" approach, it gives outrageously high White evaluations, like +1,30, when Komodo would give around 0,00 to 0,10.

The question is: Who is correct?

Have you followed the given lines by Leela, figured out where she diverges from SF/Komodo/etc, and gotten the impression that she just misses all avenues of counterplay, or why she so grossly 'overestimates' the chances?

https://lichess.org/analysis/standard/2r1r3/1p3nkp/pP1pn1p1/P2N1p2/2P2P2/R3P2P/6... in this sample position out of a Dutch Leningrad, SF @depth32 spits out +0.5 with a variety of moves (Rc3, Rc1, Ra4, Bf1). White Advantage, but nothing dramatic.

https://www.chess.com/computer-chess-championship#game=44 Leelenstein did go for one of the SF suggestions (35.Rc3), but thinks White is completely winning (+2.6), and btw has been thinking so.. for pretty much the entire game!
The initial position after the opening book is evaluated by SF as +0.29, and by LS as a whopping +1.59

As late as move 54, SF believes to be in a tenable position (+0.76), while Leelenstein is long confused why there's no resignation coming (+4.26)..
After 55.Bf1, SFs eval suddenly explodes in White's favour, and never recovers.
Did Leelenstein overestimate White's chances here..?
  
Back to top
 
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #47 - 09/13/19 at 08:30:05
Post Tools
brabo wrote on 09/13/19 at 07:03:52:
Leon_Trotsky wrote on 09/12/19 at 22:21:03:
Except that I have no desktop. I have not had a desktop since around 1991 or 1992  Cheesy

I think any ambitious chessplayer needs a desktop and a laptop. The desktop is used for making good analysis at home when you have plenty of time. The laptop is needed when you go to tournaments. I see some players using their smartphone instead of the laptop. In theory via the cloud anything is accessible but I prefer a bigger screen to look at.

In the last 20 years I always had a desktop and laptop at the same time which I used for chess. In the beginning I replaced them every 3 years but now I try to wait a bit longer. Those replacements don't come cheap and the gain is not always big.


I agree with you here. A desktop is much better price/performance than a laptop. But, as you say, the gains are not always big. Diminishing returns I guess.

When playing games not at home, I usually bring my laptop, but often use my phone due to laziness, although I also prefer the much larger screen of a laptop.
  
Back to top
 
IP Logged
 
fling
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1571
Joined: 01/21/11
Gender: Male
Re: New version of LcO
Reply #46 - 09/13/19 at 08:27:10
Post Tools
MartinC wrote on 09/12/19 at 21:30:42:
It probably won't be running very well on a laptop? Much happier running on a decent GPU.

Its very classical and likes space, which are obviously somewhat arguable things. I've generally found it to 'feel' really quite sensible with its +/= sort of claims.


I agree with Martin here. Lc0 does not run well on a laptop. Even on a desktop, like my old desktop with core i5-6600K and RTX 2060, it is not supposed to be great with the larger nets. EDIT: It should have been RTX 960.

I think you should try one of the nets that are claimed to be better for CPU in case you want to use Lc0 on your laptop.

The point is that you judge the engine based on a rig that is not good for it.
« Last Edit: 09/13/19 at 14:40:30 by fling »  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #45 - 09/13/19 at 07:03:52
Post Tools
Leon_Trotsky wrote on 09/12/19 at 22:21:03:
Except that I have no desktop. I have not had a desktop since around 1991 or 1992  Cheesy

I think any ambitious chessplayer needs a desktop and a laptop. The desktop is used for making good analysis at home when you have plenty of time. The laptop is needed when you go to tournaments. I see some players using their smartphone instead of the laptop. In theory via the cloud anything is accessible but I prefer a bigger screen to look at.

In the last 20 years I always had a desktop and laptop at the same time which I used for chess. In the beginning I replaced them every 3 years but now I try to wait a bit longer. Those replacements don't come cheap and the gain is not always big.
  
Back to top
WWW  
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #44 - 09/13/19 at 06:56:08
Post Tools
an ordinary chessplayer wrote on 09/12/19 at 19:40:39:
Trying to make new preparations shortly before a game just leads me to play worse. With only a couple of hours available, a better investment of my time is to relax, go for a one-hour walk, work a few tactics puzzles, and refresh my memory as above. Not necessarily in that order.

I can show you many of my won games in which I can prove that they were won largely due to new preparations made just before the games. Besides I also make analysis after a game, have files of likely opponents, study analysis from others and repeat before a game with chesspositiontrainer old files. My private library has in the past 15 years grown to several 1000 of lines which I can use. Still I notice my library lacks very often answers on lines my opponents play.

So I am convinced there is a link between success and preparing games. I play around 2300 level but I also believe this link applies to lower levels.

It makes no sense to talk about using Lc0 in preparing games just before the games if you are not convinced of that link.

I don't know your situation but personally I do quite some sport and puzzles during the year so during a tournament I can fully focus at the preparations and of course also the games.
  
Back to top
WWW  
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #43 - 09/12/19 at 22:21:03
Post Tools
Very strategic/locked type positions and hypermodern openings like Pirc and Modern, I find that Leela overestimates, sometimes to a big degree, Black's chances.

In my experience, in openings like the Pirc and French Winawer where Black deliberately cedes space and assumes a "turtle" approach, it gives outrageously high White evaluations, like +1,30, when Komodo would give around 0,00 to 0,10.

MartinC wrote on 09/12/19 at 21:30:42:
It probably won't be running very well on a laptop? Much happier running on a decent GPU.


Except that I have no desktop. I have not had a desktop since around 1991 or 1992  Cheesy
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1982
Joined: 07/24/06
Re: New version of LcO
Reply #42 - 09/12/19 at 21:30:42
Post Tools
It probably won't be running very well on a laptop? Much happier running on a decent GPU.

Its very classical and likes space, which are obviously somewhat arguable things. I've generally found it to 'feel' really quite sensible with its +/= sort of claims.
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #41 - 09/12/19 at 21:08:01
Post Tools
I find that it tends to favour White in a lot of positions that I do not see White's advantage, and where Komodo usually gives =.

It is very heavy on the battery though. I cannot use Leela without having my laptop charged into the socket, or else it might turn off by itself.
  
Back to top
 
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 729
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: New version of LcO
Reply #40 - 09/12/19 at 19:40:39
Post Tools
brabo wrote on 09/12/19 at 17:02:42:
It is a bit hidden in the article but if you also read the comments next to the moves then you can find the real reason why I am so fond of using Lc0.

I did read those comments before. Although I can appreciate that you like Lc0, it's far from shown that it's necessary.

brabo wrote on 09/12/19 at 17:02:42:
There were probably more than 100 positions I needed to check in just a couple of hours. Books and databases are nice if you have lots of time but Leela is a real asset when you don't have time.

This is 180 degrees the opposite of the way that I prepare for an opponent. I make my preparations well ahead of time, i.e. just after a game, either one of mine against them or one I found published somewhere. I have files on likely opponents, or if I don't have a file on them then I just pick the most appropriate of one of my usual openings. Right before the game I spend a few minutes refreshing my memory from the correct file.

Trying to make new preparations shortly before a game just leads me to play worse. With only a couple of hours available, a better investment of my time is to relax, go for a one-hour walk, work a few tactics puzzles, and refresh my memory as above. Not necessarily in that order.
  
Back to top
 
IP Logged
 
IsaVulpes
Senior Member
****
Offline


No.

Posts: 326
Joined: 12/09/07
Re: New version of LcO
Reply #39 - 09/12/19 at 18:11:44
Post Tools
brabo, of interest to you might be https://github.com/dkappe/leela-chess-weights/wiki/Bad-Gyal , or perhaps even https://github.com/dkappe/leela-chess-weights/wiki/Mean-Girl:--the-most-fun-leel...

These aren't as strong as Stockfish (or even base Leela), but have their own sets of ideas, and might be more useful for 'practical' preparation / analysis than the 'godly' engines, in particular if you aren't a 2800+ player.

Bad Gyal 7 in particular is supposedly quite good at finding ideas that work nicely against humans.
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #38 - 09/12/19 at 17:02:42
Post Tools
an ordinary chessplayer wrote on 09/12/19 at 13:26:05:
brabo wrote on 09/11/19 at 17:20:56:
I finally found some time to write a small review on my blog of why we should use Lc0: https://chess-brabo.blogspot.com/2019/09/leela-lc0.html

Thanks, there's a lot of ideas in that blog post, but to my mind it does not quite explain why we should use Lc0. We can construct several related propositions:
  1. We should prepare.
  2. We should prepare with an engine.
  3. We should prepare with the latest engine.
  4. We should prepare with a specific engine.

(In the above chain, for "engine" we could substitute "book" or "database"; for the professionals additionally "second" or "team".)

For example, I think in your game against Guy Baete, you could have won by preparing the same opening without an engine, just using the database, so I don't see it as an argument in favor of Lc0 per se.

It is a bit hidden in the article but if you also read the comments next to the moves then you can find the real reason why I am so fond of using Lc0.
"I started recently to use Leela when I want a quick answer in a game-preparation for an early opening-position which I didn't analyze before."

At this moment I practically only use Leela for that purpose. Besides I also want to add that my opponent Guy Baete played something very different to the line I discuss. There were probably more than 100 positions I needed to check in just a couple of hours. Books and databases are nice if you have lots of time but Leela is a real asset when you don't have time.

My articles are full of those hidden pieces of advice. It is a way to get some interaction with the readers which are interested in the subject and avoid trolls.
  
Back to top
WWW  
IP Logged
 
an ordinary chessplayer
God Member
*****
Offline


I used to be not bad.

Posts: 729
Location: Columbus, OH (USA)
Joined: 01/02/15
Re: New version of LcO
Reply #37 - 09/12/19 at 13:26:05
Post Tools
brabo wrote on 09/11/19 at 17:20:56:
I finally found some time to write a small review on my blog of why we should use Lc0: https://chess-brabo.blogspot.com/2019/09/leela-lc0.html

Thanks, there's a lot of ideas in that blog post, but to my mind it does not quite explain why we should use Lc0. We can construct several related propositions:
  1. We should prepare.
  2. We should prepare with an engine.
  3. We should prepare with the latest engine.
  4. We should prepare with a specific engine.

(In the above chain, for "engine" we could substitute "book" or "database"; for the professionals additionally "second" or "team".)

For example, I think in your game against Guy Baete, you could have won by preparing the same opening without an engine, just using the database, so I don't see it as an argument in favor of Lc0 per se.
  
Back to top
 
IP Logged
 
MartinC
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1982
Joined: 07/24/06
Re: New version of LcO
Reply #36 - 09/11/19 at 20:32:56
Post Tools
In some ways I think that's especially true for ameteurs & openings - LC0 is very strongly classical in its tastes and isn't remotely as prone to lines purely depending on tactics as say Stockfish is.

So you get very reliable opening ideas. Useful for fallible memories!
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #35 - 09/11/19 at 17:20:56
Post Tools
I finally found some time to write a small review on my blog of why we should use Lc0: https://chess-brabo.blogspot.com/2019/09/leela-lc0.html
  
Back to top
WWW  
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #34 - 07/08/19 at 21:47:56
Post Tools
brabo wrote on 06/28/19 at 14:51:37:
tony37 wrote on 06/28/19 at 14:43:01:
@brabo: I also would assume that the more recent networks should be better than number 42000, and yet I see on the rating list that for example number 42600 has 25 elo less, while having a much higher ordo elo. Maybe I'm misunderstanding something here. And of course not all the networks are on that list.

Have you also tested some version of Leelafish?


No, I have only limited resources available. In fact I wouldn't do those tests if there were good benchmarks available. Progress of Lc0 is so fast that any tests are quickly becoming outdated.

the LCZ rating list now finally gives network number 42700 as better than 42000, so I'm definitely switching to that one
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #33 - 07/05/19 at 10:36:43
Post Tools
I do know that you have to compile it yourself to run it with Linux. If that helps, I have no experience with Linux, hope this helps. You are welcome.
  
Back to top
 
IP Logged
 
proustiskeen
God Member
*****
Offline


Hello from Omaha!

Posts: 658
Joined: 08/11/08
Re: New version of LcO
Reply #32 - 07/04/19 at 16:41:14
Post Tools
I don't cover Linux installs, only Windows -> Chessbase.
  
Back to top
WWW  
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 920
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: New version of LcO
Reply #31 - 07/04/19 at 10:17:57
Post Tools
Thank you for answering.
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #30 - 07/04/19 at 09:30:06
Post Tools
John Hartmann (sp?) has a youtube clip about installing LcO that looks very useful.
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #29 - 07/03/19 at 15:31:08
Post Tools
Jupp53 wrote on 07/03/19 at 10:20:40:
Is there any forum with help for running Lc0 under Linux? Couldn't find it by my search, in the Lc0-blog or the github-site. Which hardware should I get, as I have only at least five yo pc's without extra graphic card?

I've managed to compile and run it on Linux. But if you want to buy new hardware I suppose you want some Nvidia card and I have no experience with that. But when I google I find several guides how to install proprietary Nvidia drivers.
In my case it helped installing a recent Linux kernel (5.0 or 5.1), otherwise my mouse was hanging while running Lc0.
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #28 - 07/03/19 at 11:14:27
Post Tools
I think some people run LcO with an internal graphics card. The best place I know of is LcO.org. There is a getting started section. There may be something on youtube of course.
  
Back to top
 
IP Logged
 
Jupp53
God Member
*****
Offline


be

Posts: 920
Location: Frankfurt/Main
Joined: 01/04/09
Gender: Male
Re: New version of LcO
Reply #27 - 07/03/19 at 10:20:40
Post Tools
Is there any forum with help for running Lc0 under Linux? Couldn't find it by my search, in the Lc0-blog or the github-site. Which hardware should I get, as I have only at least five yo pc's without extra graphic card?
  

Medical textbooks say I should be dead since April 2002.
Dum spiro spero. Smiley
Back to top
 
IP Logged
 
JFugre
Junior Member
**
Offline


I Love ChessPublishing!

Posts: 81
Joined: 01/22/19
Re: New version of LcO
Reply #26 - 06/30/19 at 16:36:35
Post Tools
Leon_Trotsky wrote on 06/27/19 at 22:00:03:
In the run box I received a lot of message saying this:

warning: Linking two modules of different target triples: ' is spir64' whereas '<origin>' is 'vISA_64'

and some warning with repeated numbers on base 2 like 8:8:8:64:8:8:1024

I updated drivers because I saw how old drivers can mess up the process, so drivers are not problem.

Those warnings *are* from the OpenCL drivers though. What video card is this? I don't remember seeing this on AMD.
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #25 - 06/28/19 at 14:51:37
Post Tools
tony37 wrote on 06/28/19 at 14:43:01:
@brabo: I also would assume that the more recent networks should be better than number 42000, and yet I see on the rating list that for example number 42600 has 25 elo less, while having a much higher ordo elo. Maybe I'm misunderstanding something here. And of course not all the networks are on that list.

Have you also tested some version of Leelafish?


No, I have only limited resources available. In fact I wouldn't do those tests if there were good benchmarks available. Progress of Lc0 is so fast that any tests are quickly becoming outdated.
  
Back to top
WWW  
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #24 - 06/28/19 at 14:43:01
Post Tools
@brabo: I also would assume that the more recent networks should be better than number 42000, and yet I see on the rating list that for example number 42600 has 25 elo less, while having a much higher ordo elo. Maybe I'm misunderstanding something here. And of course not all the networks are on that list.

Have you also tested some version of Leelafish?
  
Back to top
 
IP Logged
 
brabo
God Member
*****
Offline


Welcome chessfriend

Posts: 942
Joined: 02/02/07
Re: New version of LcO
Reply #23 - 06/28/19 at 12:46:05
Post Tools
tony37 wrote on 06/27/19 at 21:17:42:
personally I'd take number 42000 (link) as this still appears to be the best one according to this table, but you can't go very much wrong if you take a recent one with a high elo.

I think that file is already a bit old.
The 42000 network has an ordo of 3140.
I tested last month the network 42372 with ordo 3200. That network achieved on my mediocre computer 44-56 against Komodo 11 with tempo 15 minutes + 10 seconds increment.
I just finished a new match over 100 games with the network 42626: 3259 ordo and that one scored 50-50 with same settings and positions.

I know 100 games isn't much but I do get the impression that progress has been made so I would consider instead of network 42000 the most recent one with the highest ordo.

I also think on a slower tempo the results would even improve for Lc0 not to mention what kind of gain you could get running it on a much stronger Nvidia card.
  
Back to top
WWW  
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #22 - 06/28/19 at 10:35:38
Post Tools
Oh ok, I did because the blog said "run the client".
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #21 - 06/28/19 at 08:59:47
Post Tools
Pawnpusher wrote on 06/28/19 at 02:18:33:
Shouldn't he run the client before he tries to install the engine? I thought that was a prerequisite.

I think the client is only needed if you want to contribute to the networks
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #20 - 06/28/19 at 02:18:33
Post Tools
Shouldn't he run the client before he tries to install the engine? I thought that was a prerequisite.
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #19 - 06/27/19 at 22:41:23
Post Tools
does the program stop after those warnings? or does it keep running and do you see stuff like this:
(1/578) KWG=32 KWI=2 MDIMA=8 MDIMC=8 MWG=16 NDIMB=8 NDIMC=8 NWG=16 SA=0 SB=0 STRM=0 STRN=0 VWM=1 VWN=1 68235.6 us (7.9 GFLOPS)
(4/578) KWG=32 KWI=2 MDIMA=8 MDIMC=8 MWG=16 NDIMB=8 NDIMC=8 NWG=32 SA=0 SB=0 STRM=0 STRN=0 VWM=1 VWN=1 53084.7 us (10.1 GFLOPS)
(5/578) KWG=32 KWI=2 MDIMA=8 MDIMC=8 MWG=32 NDIMB=8 NDIMC=8 NWG=32 SA=0 SB=0 STRM=0 STRN=0 VWM=1 VWN=1 42294.7 us (12.7 GFLOPS)
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #18 - 06/27/19 at 22:00:03
Post Tools
In the run box I received a lot of message saying this:

warning: Linking two modules of different target triples: ' is spir64' whereas '<origin>' is 'vISA_64'

and some warning with repeated numbers on base 2 like 8:8:8:64:8:8:1024

I updated drivers because I saw how old drivers can mess up the process, so drivers are not problem.
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #17 - 06/27/19 at 21:17:42
Post Tools
I very much doubt you can use the cuda file without Nvidia card, so you should take the lc0-v0.21.2-windows-opencl.zip file (link), you extract the files, and you put a networks file in the same folder as the extracted files. You can find the networks at http://lczero.org/networks, personally I'd take number 42000 (link) as this still appears to be the best one according to this table, but you can't go very much wrong if you take a recent one with a high elo. You can give the file a more sensible name if you like ('network42000' or so), but the name doesn't matter.
Then you run the lc0.exe program, you type "go nodes 100" (or another number) and if there are no errors the program will start a configuration process which can take some 5 minutes or so. In my case it says: "Started OpenCL SGEMM tuner with batch size 16.
Will try 578 valid configurations."
And after that you get Leela's evaluation of the starting position using 100 nodes, for example: "bestmove g1f3 ponder d7d5"
Now you can quit the program and start the engine on Chessbase.
  
Back to top
 
IP Logged
 
JFugre
Junior Member
**
Offline


I Love ChessPublishing!

Posts: 81
Joined: 01/22/19
Re: New version of LcO
Reply #16 - 06/27/19 at 21:17:14
Post Tools
1) try "uci" before giving the go nodes command. Probably not actually required though.
2) Is the download corrupt? You should have a ~54M network file for the latest weight series.
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #15 - 06/27/19 at 18:48:20
Post Tools
I extract the cuda download file, make a new folder and puts its contents in folder. Then I put network there too.

I open ChessBase, game window then "create engine" but at the end, analysis window is blank.

When running lc0.exe before, I tried "go ndes 100" into command bar. Gives me "error network weights file not found"
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #14 - 06/27/19 at 11:50:37
Post Tools
As far as I understood, you don't need the client (I compiled the engine on Linux, never bothered with client files).
https://github.com/LeelaChessZero/lc0/wiki/Getting-Started says: "To play or analyze games with Lc0, you don't need client.exe, but you do need the engine lc0.exe and a network file."
Maybe you need newer graphics drivers for OpenCL support. I looked up Intel HD 620 and it certainly supports OpenCL: https://www.intel.com/content/www/us/en/support/articles/000005524/graphics-driv...
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #13 - 06/27/19 at 11:23:31
Post Tools
Are you running the client before you try to install?
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #12 - 06/26/19 at 22:02:22
Post Tools
I deleted and tried installing Leela again. Now says "there is exception" and "there is failure in the engine" before crashing.

I think that installing this thing is harder than playing against it...
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #11 - 06/25/19 at 13:01:58
Post Tools
I have noticed that, it seems,  Houdini's move choices are more like LcO than either Komodo or Stockfish. Just an observation, I certainly haven't done any sort of research.
  
Back to top
 
IP Logged
 
Jonathan Tait
Senior Member
****
Offline



Posts: 489
Location: Nottingham
Joined: 07/11/06
Re: New version of LcO
Reply #10 - 06/24/19 at 19:42:41
Post Tools
Leon_Trotsky wrote on 06/24/19 at 03:32:28:
I find that in blocked or closed/fixed-pawn structure type positions, Stockfish overestimates and gives me bullshit evaluations. I feel like I have to subtract at least 0,50 from side with space advantage.


If I think Stockfish is being too optimistic, I switch to Houdini.
  

blog inspired by Bronstein's book, but using my own games: http://200opengames.blogspot.co.uk/
Back to top
WWW  
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #9 - 06/24/19 at 03:32:28
Post Tools
I find that in blocked or closed/fixed-pawn structure type positions, Stockfish overestimates and gives me bullshit evaluations. I feel like I have to subtract at least 0,50 from side with space advantage.

I know not how Leela is in positional evaluation, but I feel like there has to be way to get more realastic view of Komodo's over-emphasis on positional and Stockfish's over-emphasis on tactics and space.
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #8 - 06/23/19 at 21:45:39
Post Tools
There are definite differences in LcO analysis and Stockfish/Komod because the neural network is a different arrangement than SF and Komodo use. Of course they are going to find the same move at times, but LcO is stronger and different.
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #7 - 06/23/19 at 21:11:37
Post Tools
I have Intel HD 620 graphics card.

In any case, Leela never loads.

To find tactical stuff, I run Stockfish 10 for hours. I cannot waste any more time with this Leela installation.

Komodo 10, however seems more positional than tactical. If really needed I use both Stockfish 10 and Komodo 10 to make sure no blunder. Stockfish seems to give me overestimations in positional/blocked position.
  
Back to top
 
IP Logged
 
tony37
God Member
*****
Offline


I Love ChessPublishing!

Posts: 607
Joined: 10/16/10
Re: New version of LcO
Reply #6 - 06/23/19 at 16:34:02
Post Tools
trw wrote on 06/23/19 at 14:52:36:
If you just have a cpu and no GPU then don't bother, it will be an utter waste of your time and not give you any analysis worth a crap. You need a strong GPU.

This isn't even necessarily true.
When analysing some French lines, Leela more than once happened to give better analysis than Stockfish (the reverse was also true). A very concrete example is 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3+ 6.bxc3 Ne7 7.Qg4 O-O 8.Bd3 Nbc6 9.Nf3 f5 10.Qg3 Qb6 and here Leela finds (after 4 minutes on very inadequate hardware, i.e. some 30 nodes per second) the crushing move 11.Bh6, the most entertaining continuation being 11...Ng6 12.h4 c4 13.h5 Qb2 14.hxg6 Qxc3+ 15.Kf1 Qxa1+ 16.Ke2 Qxh1 17.gxh7+ with mate in 6 to come.
But Stockfish and Komodo 10 seem to be blind to this (after 400 million nodes, with PV each time at 3), even when you suggest 11.Bh6 to them, they don't see the point. Maybe Houdini is different but I don't have a recent version of it.
So even on suboptimal hardware, running Leela in a few very specific/unclear positions can be really interesting. But I'd agree that using it as your main engine is not workable without good GPU.

edit: after 600 million nodes (11 minutes when I only use one CPU core) Stockfish finds 11.Bh6 at last; Komodo is a real dumbass here, still doesn't have a clue after 3.5 billion nodes, I'm pretty sure Houdini would do better
The difference is clearest when you start analyzing from 10...Qb6 11.Bh6 Ng6. Leela sees 12.h4 almost immediately, while Stockfish easily takes a minute. Even when I use only CPU for Leela (using the openblas-backend) it is still faster than Stockfish using 2 cores. At a ridiculous 5 nodes per second!
« Last Edit: 06/23/19 at 19:35:26 by tony37 »  
Back to top
 
IP Logged
 
trw
God Member
*****
Offline


I Love ChessPublishing!

Posts: 1319
Joined: 05/06/08
Gender: Male
Re: New version of LcO
Reply #5 - 06/23/19 at 14:52:36
Post Tools
Leon_Trotsky wrote on 06/22/19 at 22:53:50:
This is still pain in the arse. Why cannot they make it without having to download these network files ¿

I tried it again and it still just gives me blank window in the evaluation section of ChessBase board.

I only have i7 Windows 10. Not sure if that is slow or fast, but the only way that I can get this Leela installed is if I bring my computer to some tech person who knows what they are doing, so that they can install it for me.  Shocked



If you just have a cpu and no GPU then don't bother, it will be an utter waste of your time and not give you any analysis worth a crap. You need a strong GPU.
  
Back to top
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #4 - 06/22/19 at 23:38:23
Post Tools
You have to have the network. Did you download the files?
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #3 - 06/22/19 at 22:53:50
Post Tools
This is still pain in the arse. Why cannot they make it without having to download these network files ¿

I tried it again and it still just gives me blank window in the evaluation section of ChessBase board.

I only have i7 Windows 10. Not sure if that is slow or fast, but the only way that I can get this Leela installed is if I bring my computer to some tech person who knows what they are doing, so that they can install it for me.  Shocked
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
Re: New version of LcO
Reply #2 - 06/22/19 at 22:29:42
Post Tools
I looked on the LcO blog, there is a post about the newest version. You do the download, extract the files to a folder. I called my folder LcO. You need to download a network, extract it to the same folder. You have to make sure the network file is named weights. Run the client. Then you install the LcO engine to CB or whatever gui you are using. The default settings work fine for me.
If you have a really good comp, you want a 40 network, medium comp you should use a 20, not so new you want a 10 network.
  
Back to top
 
IP Logged
 
Leon_Trotsky
Senior Member
****
Offline


Кто был никем — тот станет
всем!

Posts: 499
Location: Barcelona, CAT
Joined: 08/11/17
Gender: Male
Re: New version of LcO
Reply #1 - 06/22/19 at 21:52:11
Post Tools
How easy ¿ With Stockfish or Komodo all required is to download, extract files, then add to ChessBase. Is new version of Leela as easy as these two ¿
  
Back to top
 
IP Logged
 
Pawnpusher
Full Member
***
Offline


I Love ChessPublishing!

Posts: 158
Joined: 01/04/18
Gender: Male
New version of LcO
06/22/19 at 13:30:32
Post Tools
The newest version, changes to evaluation. Easy to set up and free, why not?
  
Back to top
 
IP Logged
 
Bookmarks: del.icio.us Digg Facebook Google Google+ Linked in reddit StumbleUpon Twitter Yahoo