Search found 116 matches

by Pio
Sun Apr 29, 2018 9:19 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Where is bug?
Replies: 18
Views: 2202

Re: Where is bug?

Hi, the error might be that you did not update the position correctly when he took your pawn on f7. If you did not update the enemy king move correctly you would have the enemy king on e6 and taking his pawn would be illegal. If that would be the case your king move would be logical.
by Pio
Wed Apr 18, 2018 6:23 pm
Forum: Computer Chess Club: General Topics
Topic: LC0 on 43 cores had a ~2700 CCRL ELO performance.
Replies: 58
Views: 8643

Re: LC0 on 43 cores had a ~2700 CCRL ELO performance.

It would have been interesting if They had swapped hardware. I wonder how good stockfish would have been on TPU:s ;)
by Pio
Sat Apr 07, 2018 6:58 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Looking for advices
Replies: 27
Views: 6959

Re: Looking for advices

Hi Vivien!

What I do in my engine is to have one move array shared for all plies. I then have an index array telling where each ply starts. In that way you can use all the space and do not have to account for the extreme case nr of moves.

/Pio
by Pio
Sun Feb 04, 2018 4:19 pm
Forum: Computer Chess Club: General Topics
Topic: Is this position already lost for white?
Replies: 28
Views: 5264

Re: Is this position already lost for white?

This is of course a white win. However if black had responded with knight to A6 it would be a draw :wink:
by Pio
Sun Dec 24, 2017 11:15 pm
Forum: Computer Chess Club: General Topics
Topic: I can't believe that so many people don't get it!
Replies: 149
Views: 30509

Re: I can't believe that so many people don't get it!

Hi Michael! I do not think hgm does not think what you have done is great. The problem is that it will be very hard to train a network to primarily work as an opening database. If it could learn how to work primarily as an opening database it would have to learn not to trust the leaf probabilities, ...
by Pio
Sun Dec 24, 2017 7:14 pm
Forum: Computer Chess Club: General Topics
Topic: Stockfish could have drawn all ten (10) games ...
Replies: 8
Views: 2026

Re: Stockfish could have drawn all ten (10) games ...

I think you are wrong. You could probably have won at least a couple of them :)
by Pio
Sat Dec 23, 2017 7:01 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: An AlphaZero inspired project
Replies: 16
Views: 6451

Re: An AlphaZero inspired project

Hi! I agree with H.G.M. One thing that should also improve the learning is to score the solutions with the smallest proof trees higher. If you also make the network to understand symmetries it will also speed it up. For example when there are no pawns you could fix the side to move king to one of th...
by Pio
Mon Dec 18, 2017 9:34 pm
Forum: Computer Chess Club: General Topics
Topic: In chess,AlphaZero outperformed Stockfish after just 4 hours
Replies: 64
Views: 12015

Re: In chess,AlphaZero outperformed Stockfish after just 4 h

Hi Ed! The thing is that the engine will not learn openings first. It will learn the endings first just like a human. I do not find anything strange about their achievement. Of course you could do it even better by giving the ANN some very efficient patterns (as suggested by H.G.M and change the top...
by Pio
Sun Dec 17, 2017 9:24 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Parallel search/LazySMP question
Replies: 11
Views: 3169

Re: Parallel search/LazySMP question

The thing is that the depths will be different because the move order might change and that LMR is a function of move order. This in some way prevents the threads to search the same tree
by Pio
Sun Dec 17, 2017 9:02 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Parallel search/LazySMP question
Replies: 11
Views: 3169

Re: Parallel search/LazySMP question

They will search different moves due to things like the history table and other structures shared by the threads that make difference in the move ordering and LMR

another thing that will prevent duplicate searches is to launch the threads with different depths