Hi,
Have you looked at Pachi's usage of virtual wins and virtual losses?
I had read your paper. That was useful. Thanks.
http://pasky.or.cz/go/pachi-tr.pdf
I use VirtualLoss N=10 on single machine.
I have not tried VirtualWin. I'll try.
Winrate vs VirtualLoss N=1
Winrate N
0.457 0
0.532 2
0.494 3
0.544 4
0.538 5
0.536 6
0.550 8
0.559 10
0.549 11
0.556 12
0.510 20
0.536 30
0.562 40
0.553 100
(1sec/move, 4 threads, 9x9, 1000 games selfplay.)
Regards,
Hiroshi Yamashita
----- Original Message -----
From: "Petr Baudis" <[email protected]>
To: <[email protected]>
Sent: Wednesday, November 21, 2012 10:17 PM
Subject: Re: [Computer-go] TAAI details?
Hi!
On Wed, Nov 21, 2012 at 08:03:38PM +0900, Hiroshi Yamashita wrote:
Aya was running on 10 machines, each has 12 cores @3.3GHz(10x12 =120 cores).
I borrowed it from The University of Electro-Communications.
Cluster is Root Parallelization with summing up root results each 0.5 sec.
Aya uses GTP(Go Text Protocol) to communicate another machines.
I tested this cluster as AyaMC6 on KGS. But it got only 1d or 2d.
AyaMC5 (4 cores) has 2d. I think my method may be something wrong.
Have you looked at Pachi's usage of virtual wins and virtual losses?
Does using that make a difference for you?
Best,
Petr "Pasky" Baudis
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
