BrainLearn

[peter]

There is no error in the code and the experience file is perfectly created.
You have to have administrator privileges to launch the gui.
If you remove the "!", there is no Q-learning.
The "!" is a negation/not operator in cpp: be careful in what you do if you're not a programmer.

Nevermind, I wouldn't try things like that on my ownl, trying to compile and then be disappointed that it doesn't work at all.

I'm always logged in as admin, so launching and quitting the GUI can't be a problem as for that. Yet there isn't any experience.bin created by BrainLearn with Q-Learning checked, when I close the GUI, neither with Fritz nor with Shredder.

Could it be a matter of whole games played with full move history and result stored? I always tried some single moves to be played out so far.

You see, I could try out some of these things on my own of course, yet I don't like to as long as I'm not sure, if there isn't any bug maybe anyhow yet, making my trials worthless then.

I like BrainLearn and ShashChess very much, "Standard" learning works well, so I'm not in any hurry about Q-learning.

[amchess]

There is no error in the code and the experience file is perfectly created.
You have to have administrator privileges to launch the gui.
If you remove the "!", there is no Q-learning.
The "!" is a negation/not operator in cpp: be careful in what you do if you're not a programmer.

Nevermind, I wouldn't try things like that on my ownl, trying to compile and then be disappointed that it doesn't work at all.

I'm always logged in as admin, so launching and quitting the GUI can't be a problem as for that. Yet there isn't any experience.bin created by BrainLearn with Q-Learning checked, when I close the GUI, neither with Fritz nor with Shredder.

Could it be a matter of whole games played with full move history and result stored? I always tried some single moves to be played out so far.

You see, I could try out some of these things on my own of course, yet I don't like to as long as I'm not sure, if there isn't any bug maybe anyhow yet, making my trials worthless then.

I like BrainLearn and ShashChess very much, "Standard" learning works well, so I'm not in any hurry about Q-learning.

You have to complete the match or tournament. When it finished and quit the gui, it will create the bin file. You can contact me in private if you want.

[peter]

start an analysis and force some moves to close the GUI does not create anything .. but when I subsequently set a match play the BIN file is created

As I understood Andrea, it should work with selfplay, so I tried with "Shootout". Should or shouldn't that work too?

Does the game has to be ended with draw or full point for one side or may I stop before too? May the game be started from a midgame- position too or has starting position with full move history to be stored in game too?

[Zerbinati]

For information however:
in the infinity gui neither of the 2 engines works
an empty file is created in the installation folder of the gui and not the engine
The uci commands to activate the type of learning of Shashchess are not displayed in the options

[peter]

You have to complete the match or tournament. When it finished and quit the gui, it will create the bin file. You can contact me in private if you want.

Didn't see your answer before asking Marco too. Is he right, that it has to be a match between different engines or is real selfplay like "Shootout" allowed too?
May games be started from fens too or does full move history with starting position have be stored too? Is a distinct result like draw or full point necessary or are lines played out without win or draw possible too?

[Zerbinati]

@amchess I rewrite it in more detail because maybe you haven't seen it

For information however:
in the infinity gui neither of the 2 engines works
an empty file is created in the installation folder of the gui and not the engine
The uci commands to activate the type of learning of Shashchess are not displayed in the options
Eman successfully creates the BIN file
both on Playchess and on Infinity
You are both programmers
I hope you will solve it

[amchess]

@amchess I rewrite it in more detail because maybe you haven't seen it

For information however:
in the infinity gui neither of the 2 engines works
an empty file is created in the installation folder of the gui and not the engine
The uci commands to activate the type of learning of Shashchess are not displayed in the options
Eman successfully creates the BIN file
both on Playchess and on Infinity
You are both programmers
I hope you will solve it

Eman does'n't have Q-learning.
Even in linux, for example, Arena doesn't display an uci option ("Petrosian").
Conversely, in windows, it works perfectly.
So, this is a gui bug, not the engine and this is never the first time.
For example, chessbase updates continually their guis also to solve this kind of problems.
Anyway, we can show you, after a match play in Fritz gui, experience files are created.
Sorry, but we can't do more than this.
Finally, as we wrote in the readme file, Q-learning is for selfplay: why to use it in playchess or infinitychess?

[matejst]

@amchess I rewrite it in more detail because maybe you haven't seen it

For information however:
[...]

Hi, Marco. Glad you post here. I would like to know what's happening with your own computerchess projects? I like the SugaR NN engine very much (especially the 260120 version), and I think you did a lot to improve the overall end-users' experience and comfort. I hope you will continue in a similar vein, bringing projects to maturity and increasing their potential.

[Zerbinati]

@amchess I rewrite it in more detail because maybe you haven't seen it

For information however:
in the infinity gui neither of the 2 engines works
an empty file is created in the installation folder of the gui and not the engine
The uci commands to activate the type of learning of Shashchess are not displayed in the options
Eman successfully creates the BIN file
both on Playchess and on Infinity
You are both programmers
I hope you will solve it

Eman does'n't have Q-learning.
Even in linux, for example, Arena doesn't display an uci option ("Petrosian").
Conversely, in windows, it works perfectly.
So, this is a gui bug, not the engine and this is never the first time.
For example, chessbase updates continually their guis also to solve this kind of problems.
Anyway, we can show you, after a match play in Fritz gui, experience files are created.
Sorry, but we can't do more than this.
Finally, as we wrote in the readme file, Q-learning is for selfplay: why to use it in playchess or infinitychess?

Andrea I refer to the standard mode
this problem was also present in version 11
I changed the code and deleted the new version drop down menu
have the possibility to compile a version with the standard default learning activated
in fritz gui it works well
while in the infinity GUI it creates an empty BIN file
also by installing the engine in the GUI installation folder

[Zerbinati]

@amchess I rewrite it in more detail because maybe you haven't seen it

For information however:
[...]

Hi, Marco. Glad you post here. I would like to know what's happening with your own computerchess projects? I like the SugaR NN engine very much (especially the 260120 version), and I think you did a lot to improve the overall end-users' experience and comfort. I hope you will continue in a similar vein, bringing projects to maturity and increasing their potential.

Thanks Boban for the nice words ..
But I have many family problems and little time available .. although I would like to do many things.