Just a wild guess here. One way to get catastrophically slow performance is to
have superfluous nested loops. For instance, one could iterate over each board
space, calling a routine to check for legality. If the legality routine also
iterates over the whole board (perhaps by updating all liber
On Fri, Nov 20, 2009 at 05:19:01AM -0800, Peter Drake wrote:
> On Nov 20, 2009, at 12:08 AM, Petr Baudis wrote:
>
> >On Thu, Nov 19, 2009 at 07:03:56PM -0800, Seth Pellegrino wrote:
> >>* I've already got a stable base that I'm working from (specifically,
> >>Peter Drake's Orego[1]).
> >
> > (You
On Nov 20, 2009, at 12:08 AM, Petr Baudis wrote:
On Thu, Nov 19, 2009 at 07:03:56PM -0800, Seth Pellegrino wrote:
* I've already got a stable base that I'm working from (specifically,
Peter Drake's Orego[1]).
(You could also choose a better base to work from,
e.g. libego, Pachi or Fuego, bu
On Fri, Nov 20, 2009 at 10:31:21AM +0100, Petr Baudis wrote:
> For the liberties, of course that applies since they are used very often
> and take little memory (if you are sensible about it), but also always
> get in mind to avoid universal truths - not all features pay off to be
> maintained incr
On Fri, Nov 20, 2009 at 10:21:25AM +0100, Thomas Lavergne wrote:
> On Thu, Nov 19, 2009 at 08:25:53PM -0800, Seth Pellegrino wrote:
> > Apologies all for replying to myself so many times, but Darren Cook
> > has been kind enough to help me clear the matter up. Essentially, I
> > had misread his pap
On Thu, Nov 19, 2009 at 08:25:53PM -0800, Seth Pellegrino wrote:
> Apologies all for replying to myself so many times, but Darren Cook
> has been kind enough to help me clear the matter up. Essentially, I
> had misread his paper. I was trying to compute the expensive features
> Rémi uses for progre
Hi!
On Thu, Nov 19, 2009 at 07:03:56PM -0800, Seth Pellegrino wrote:
> Thank you both for your speedy replies. I do apologize, though, as I
> seem to have left some important information out:
>
> * I know that 70pps is useless, hence my dismay when that's the speed
> I was getting.
> * I am ru
Apologies all for replying to myself so many times, but Darren Cook
has been kind enough to help me clear the matter up. Essentially, I
had misread his paper. I was trying to compute the expensive features
Rémi uses for progressive widening as a move selector for playouts
rather than the stripped d
It may be helpful to incrementally compute information on a move-by-move basis,
rather than recompute for all points on the board.
For example; if you choose to store the liberties of groups, and a single stone
is placed on the board, it affects only those strings which have thereby lost a
lib
Ah, curse the accidental send. A few clarifications below:
On Thu, Nov 19, 2009 at 7:03 PM, Seth Pellegrino wrote:
> Hello Tom and Petr!
>
> Thank you both for your speedy replies. I do apologize, though, as I
> seem to have left some important information out:
>
> * I know that 70pps is useless,
Hello Tom and Petr!
Thank you both for your speedy replies. I do apologize, though, as I
seem to have left some important information out:
* I know that 70pps is useless, hence my dismay when that's the speed
I was getting.
* I am running these playouts on a 19x19 board, which does slow things
do
Le 19/11/2009 à 09:35, Seth Pellegrino a écrit :
>
> Hello all,
>
> I'm working on an implementation of Rémi Coulom's Elo-based move
> prediction algorithm[1], and I seem to be running into efficiency
> issues.
Very simple question to start:
What is the programmation language ?
Do you use 1d rep
On Thu, Nov 19, 2009 at 09:43:41AM +0100, Petr Baudis wrote:
> Hi!
>
> On Thu, Nov 19, 2009 at 12:35:09AM -0800, Seth Pellegrino wrote:
> > I'm working on an implementation of Rémi Coulom's Elo-based move
> > prediction algorithm[1], and I seem to be running into efficiency
> > issues. With my c
Hi!
On Thu, Nov 19, 2009 at 12:35:09AM -0800, Seth Pellegrino wrote:
> I'm working on an implementation of Rémi Coulom's Elo-based move
> prediction algorithm[1], and I seem to be running into efficiency
> issues. With my current implementation, I'm running at around 8 or 10
> playouts per secon
Hello all,
I'm working on an implementation of Rémi Coulom's Elo-based move
prediction algorithm[1], and I seem to be running into efficiency
issues. With my current implementation, I'm running at around 8 or 10
playouts per second. I can't imagine that the playouts would be so
accurate that I wou
15 matches
Mail list logo
