RomiChess P3n released

[Graham Banks]

Download link - http://kirill-kryukov.com/chess/discuss ... p?id=40457

A big thanks to Mike Sherwin (RomiChess author).

[gerold]

Download link - http://kirill-kryukov.com/chess/discuss ... p?id=40457

A big thanks to Mike Sherwin (RomiChess author).

Thanks Mike.

Thanks for posting link Graham.

Best,
Gerold.

[Daniel Anulliero]

Thanks for the new Romichess Mike

[Michael Sherwin]

Thanks guys for the kind replies! It was my pleasure.

[Werner]

This entry is moved from Topic AlphaGo Zero And AlphaZero, RomiChess done better :

Hi Mike,
so if I include all CEGT games into the Folder, merge them and repeat a match played now I will get much better result?

- original games are played without learning
- now using learning on and repeat a match

I will try : but I do not think the engines uses this lear file.
Settings here
learn_on
book_off
quit

I do not know how to use the command in the help.txt inside Romi:
douselearn ?

Now I did some more reading and I think, this is more a booklearning. So as we do not test books and use opening positions, learning will not help here. I stopped a test with opening books against DanaSah 70 after 0-20 result.

best wishes

[Michael Sherwin]

This entry is moved from Topic AlphaGo Zero And AlphaZero, RomiChess done better :

Hi Mike,
so if I include all CEGT games into the Folder, merge them and repeat a match played now I will get much better result?

- original games are played without learning
- now using learning on and repeat a match

I will try : but I do not think the engines uses this lear file.
Settings here
learn_on
book_off
quit

I do not know how to use the command in the help.txt inside Romi:
douselearn ?

Now I did some more reading and I think, this is more a booklearning. So as we do not test books and use opening positions, learning will not help here. I stopped a test with opening books against DanaSah 70 after 0-20 result.

best wishes

If you start from fen positions then there is no path to follow and no way for Romi to load the subtree before the search. If you are starting at a position at the end of a pgn file then everything would work fine.

I could create a new version that has the commands:

book_on
book_off
learning_on
learning_off
uselearn_on
uselearn_off

So with book_off reinforcement learning could still be used. And reinforcement has nothing to do with a book. The reason that everyone thinks that Romi's learning is all book learning is because a tree structure of all games played or loaded is used. It just so happens that the tree structure can be used for both a book and reinforcement learning.

Romi's reinforcement learning is not much different than a persistent hash. It is actually superior to persistent hash because it saves the entire tree of Romi's experience and loads the pertinent part of the tree into the hash. And over time those hash entries accumulate rewards and penalties which is the reinforcement learning.

Let me know if you want a special version. It will not take long.

Anyway, on a philosophical examination, if Romi starts from an empty learn file and no games are loaded in and simply chooses moves based on her own games and results can any of that be called "book learning" in the traditional sense? If Romi learns that she does best with 1.e4 can it be considered as picking a move from a book? I don't think so. Romi is only playing what she has learned is best for her from her own games. She can never choose randomly between acceptable book moves. So in effect there really is no book in RomiChess.

[Werner]

Sorry Mike,
I still do not see how to use these commands correctly. See here the answer of the program inside command prompt:

White<1>learning_on

Error (unknown command): learning_on

r n b q k b n r
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
R N B Q K B N R
hashSig = cb57540b31517d13
hashKey = bd5162a4
pawnKey = 9e8764f7
White<1>uselearn_on

Error (unknown command): uselearn_on

Do you remember the correct commands ?

And how can I uses these commands together inside Arena:

as commandline or as startparameters, and how to separate with ;?

PS: I am using a pgn file for the openings (about 8-10 moves).

best wishes

[Damir]

Thanks Mickey.

[Michael Sherwin]

Sorry Mike,
I still do not see how to use these commands correctly. See here the answer of the program inside command prompt:

White<1>learning_on

Error (unknown command): learning_on

r n b q k b n r
p p p p p p p p
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
P P P P P P P P
R N B Q K B N R
hashSig = cb57540b31517d13
hashKey = bd5162a4
pawnKey = 9e8764f7
White<1>uselearn_on

Error (unknown command): uselearn_on

Do you remember the correct commands ?

And how can I uses these commands together inside Arena:

as commandline or as startparameters, and how to separate with ;?

PS: I am using a pgn file for the openings (about 8-10 moves).

best wishes

Hi Werner, Sorry for the confusion. The commands currently in RomiChess are:

learn_on
learn_off
book_on
book_off

They cannot be entered in Arena. So to learn and use the learning enter:

learn_on
book_on
quit

Or just delete the old learn file and Romi will create a new learn file with learning enabled. The values that turn learning on and off are stored in the learn file. Arena and Winboard work fine with Romi's learning but I am not sure about any other.

I would like to thank you for your interest in Romi's learning whether you decide to test with it on or not.

[Werner]

I would like to thank you for your interest in Romi's learning whether you decide to test with it on or not.

Hi Michael,
thanks for the answer. I will give it a try for private testing under Arena, not for CEGT. For this test I have matched all CEGT games into the learn.dat.

best wishes
Werner