Search found 83 matches

by Cheney
Sat Sep 02, 2017 4:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Cutechess-cli Time Control: Me or Bug
Replies: 24
Views: 4292

Re: Cutechess-cli Time Control: Me or Bug

Lower the increment? I have done that, it does decrease the time a bit. 0.8 increment increases the game (with 10 moves) to 8 additional seconds. Maybe go less than that? Like 0.3? Hmmm, that I'll try :), for some reason, I always wanted to avoid going that low, maybe I did not want to expose a bug...
by Cheney
Sat Sep 02, 2017 2:27 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Cutechess-cli Time Control: Me or Bug
Replies: 24
Views: 4292

Re: Cutechess-cli Time Control: Me or Bug

Sorry to touch on this old thread, but I am wondering something as I start to perform engine vs engine matches again. I want to have engines play fast games. I am not totally versed on the definitions of blitz versus rapid versus bullet, but the idea I thought was to play a game in 3 seconds with a ...
by Cheney
Sat Sep 02, 2017 2:03 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

Thanks to all with your guidance and patience on this subject (Jon, Robert, Alvaro, etc. :) ) I think I finally got this working. I first had some odd bugs which threw me off for weeks. All I wanted to test was piece values. Even though the testing function to search for better parameters appeared t...
by Cheney
Mon Aug 14, 2017 11:47 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

Here's the latest after another week of research, studying, and testing :). From calculating the derivative, I am not sure if I should be calculating the dt of (result - Sigmoid(qs))^2 or 1/N*Sum(n=1,(result - Sigmoid(qs))^2). There are other sites that calculate derivatives and I do not seem to alw...
by Cheney
Wed Aug 09, 2017 11:12 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

Thanks for the replies :). I have been trying to study up a bit on calculus and read some docs/examples on derivatives and gradient descent. Compute the derivative of the loss function (the thing you are trying to minimize) with respect to each of the parameters. Here's where I am lost. First, just ...
by Cheney
Thu Jul 27, 2017 3:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

Thanks Jon :) I do follow that concept but what I am stumped on is the idea of testing multiple parameters simultaneously. A reasonable optimization algorithm can tune several hundred simultaneously If I take four parameters and increase them all by 1, then test, and I find a lower E, maybe two of t...
by Cheney
Thu Jul 27, 2017 11:43 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

If you cannot get a positive result, that could mean your parameter values are already optimal or there are bugs in the eval I went as far as starting over with my original eval - tapered eval with PSQ. I added the pawn code for pawn structure. In base, I set bonus/penalties to 0 and in base-1 I se...
by Cheney
Thu Jul 27, 2017 11:08 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

There is no need to tune parameters one at a time I did look at the pseduo code on CPW for the local search and have a similar model. Are you saying I could, for example: 1. Set chain pawn +1 2. Set Backward pawn +1 3. Set rook on 7th -1 4. ... set a few other parameters to +/-1 5. Now run qsearch ...
by Cheney
Wed Jul 26, 2017 11:15 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Texel tuning method question

I will certainly try this out as well to see if this helps but for some reason, I am just not getting positive results with this. I am thinking I am missing something. I reread all the posts to this thread. Would someone mind shedding some light on my scenario? I am sorry if some of this is repetiti...
by Cheney
Sat Jul 22, 2017 4:28 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Texel tuning method question
Replies: 63
Views: 13539

Re: Code bug fix

Thank you :). I have not coded much with python but maybe this will help me start. I have coded my tuning within the engine source. Based on your output sample, it looks like I am walking the parameters properly... test param1 + 1, look for better E, if not then test param - 1. I do have a question ...