Search found 3154 matches

by Daniel Shawul
Tue Jan 15, 2019 6:48 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Egbb dll neural network support
Replies: 7
Views: 1373

Re: Egbb dll neural network support

Egbbdll now supports leela networks, which have policy head, as well besides value head. To use the lc0 networks, they should be converted to formats readable by egbbdll (i.e. either Protobuf (pb) format for Tensorflow backend, and UFF format for TensorRT support). The cpu version does not work yet ...
by Daniel Shawul
Thu Dec 27, 2018 9:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: neural network architecture
Replies: 7
Views: 719

Re: neural network architecture

One thing that helps me a lot in reducing the number of training games is to 'shortcut' the deep ResNet by adding a layer of inputs ( e.g. piece counts) to the dense network of the policy head, and also using attack tables (that should in theory reduce the number of convolution layers) required. How...
by Daniel Shawul
Fri Dec 14, 2018 7:17 pm
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

@Mathew Lai

I have some questions:

Does A0 convolution's have bias, or just convolution+normalization ?

How many games roughly does it take to capture piece values say on the 20b net using self play?
It is not clear to me what kind of filter can be used for counting precisely the number of pieces.
by Daniel Shawul
Mon Dec 10, 2018 7:37 pm
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

I do not apply my human imagination in order to explain the training process, I see it only tries to fit the output of a 800 node search by 1 node, and do so in many iterations. If you ask how that correlates to chess or any other games, then there is a easy way out, they are domain knowledge. Now ...
by Daniel Shawul
Mon Dec 10, 2018 5:34 am
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

I don't get some of the negative talk about A0 and the constant bickering about test conditions ... have you guys played through some of those games ??? Obviously there is something "game changing" happening. If you go through the games of some of the human chess geniuses of our time, on a really g...
by Daniel Shawul
Mon Dec 10, 2018 4:41 am
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

@Mathew Lai The following pseudo code for Dirichlet noise seems to be wrong # At the start of each search, we add dirichlet noise to the prior of the root # to encourage the search to explore new actions. def add_exploration_noise(config: AlphaZeroConfig, node: Node): actions = node.children.keys() ...
by Daniel Shawul
Mon Dec 10, 2018 2:54 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Linux program to convert computer algebraic in text file to PGN in 2nd text file
Replies: 9
Views: 596

Re: Linux program to convert computer algebraic in text file to PGN in 2nd text file

Anyone have the converter? It could just receive this kind of thing e2e4 e7e5 g1f3 g8f6 and output in PGN. ...and so forth, all communication being done with the input physical file of CA and output file of PGN. I had a need for that kind of tool and didn't want to write code for implementing SAN s...
by Daniel Shawul
Fri Dec 07, 2018 10:33 am
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

What do people think would happen if A0 used 4 TPUs and just 1 CPU core? How would that change its strength? I see no "fairness" reason to give it 44 cores as well as the 4 TPUs. I am pretty sure that AlphaZero hardly used any of those cores. There just wouldn't be anything to do for them while it ...
by Daniel Shawul
Thu Dec 06, 2018 11:35 pm
Forum: Computer Chess Club: General Topics
Topic: Alphazero news
Replies: 381
Views: 26652

Re: Alphazero news

https://www.chess.com/news/view/updated-alphazero-crushes-stockfish-in-new-1-000-game-match Stockfish 8 ??? They really fear real Stockfish ... Unfortunately we weren't able to get our time machine up and running before we finished writing the paper in order to test against engines that hadn't been...
by Daniel Shawul
Wed Nov 28, 2018 11:41 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Are draws hard to predict?
Replies: 18
Views: 1703

Re: Are draws hard to predict?

oh, I hadn’t worked out how to explicitly tell it WDL, I figured it was treating value as continuous via MSE and then deciding for itself it was binary or whatever by reading the y-values. Because, when I adapted to continuous value, the accuracy figure dropped to zero (I assumed that was its way o...