[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gnugo-devel] strength against Many Faces
From: |
David G Doshay |
Subject: |
Re: [gnugo-devel] strength against Many Faces |
Date: |
Fri, 22 Apr 2005 12:43:07 -0700 |
On 19, Apr 2005, at 10:37 AM, David Fotland wrote:
Are the games played in a way that allows Many Faces to learn? It
keeps the
learned positions in a file in the install folder called flearn19.dat.
We see flearn9.dat.
This
file is read once when MFGO starts up, and is written after each
completed
games with the new learned positions.
MFGO will prefer to repeat openings where it wins games, and will try
to
avoid openings that lead to losses.
I am a bit confused.
We have a number of PCs running MFoG playing against parts of our
clusters. On each machine we have checked, the file has the date of the
original installation. However, the contents of the file is different
on each machine, and in many places.
When the file gets updated (I understand at the end of each game), is
there a way that we might inadvertently be avoiding the updates, and if
not, why isn't the date on the file updated to indicate the last write?
Just to be complete, we run sets of games, and terminate MFoG between
changes in SlugGo's search parameters or if there is a crash, so I
expect that the file should be updated regularly ... many times since
September when we started playing SlugGo against MFoG.
As an experiment we even removed the file from one machine. Runs are
accumulating but we have nothing significant yet.
Thanks for your time,
David