Page 2 of 2
Re: LC0 cuda speed question
Posted: Sat Aug 11, 2018 8:09 pm
by jmartus
Well how do I use those weights insead?
Re: LC0 cuda speed question
Posted: Sun Aug 19, 2018 7:55 pm
by consen
Save them (Weight-file) in same place where lczero.010.exe is.
Or a newer lczero.xxx.exe.
You not need unzipping or unpakking.
And take away other txt-files in same directory (map)
Then Lczero use the file ending in txt.gz.
(No need for unpaking gz-file).¨
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 7:07 pm
by AdminX
2x GTX 1060 Net ID 10970
--threads=4
--minibatch-size=256
--backend=multiplexing
--backend-opts=(backend=cudnn,gpu=0,max_batch=1024),(backend=cudnn,gpu=1,max_batch=1024)
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 9:02 pm
by brianr
With my GTX-770 and default options with NN 588 I get about 2,750 nps after "hashfull" which is after 130 seconds. Eleven seconds is too short for it too "spin up" and at that point I get about 2,100 nps. So, your results look like mine.
Brian
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 10:05 pm
by Milos
brianr wrote: ↑Thu Aug 23, 2018 9:02 pm
With my GTX-770 and default options with NN 588 I get about 2,750 nps after "hashfull" which is after 130 seconds. Eleven seconds is too short for it too "spin up" and at that point I get about 2,100 nps. So, your results look like mine.
Brian
net 588 is "small" i.e. 15x192 net.
You would get at least 2x smaller nps with "large" i.e. 20x256 net.
Btw, GTX770 is 2.5x slower than 1060.
With GTX770 and 20x256 net I was getting around 1400nps after 300k nodes.
With GTX1060 and 20x256 net I am getting close to 4000nps after 300k nodes. However, my GTX1060 is strongly OC'ed, otherwise it would be the usual 3600nps.
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 10:27 pm
by cc2150dx
I'm running LC0 cuda (default settings) with my Geforce GTX 760 and only getting 875 nps. Something doesn't seem right?
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 11:55 pm
by AdminX
cc2150dx wrote: ↑Thu Aug 23, 2018 10:27 pm
I'm running LC0 cuda (default settings) with my Geforce GTX 760 and only getting 875 nps. Something doesn't seem right?
That seems about right if you are using 20x256 net. On my Dell Laptop with an Nvidia MX150 GPU I am getting between 865 - 900 nps.
Re: LC0 cuda speed question
Posted: Thu Aug 23, 2018 11:56 pm
by brianr
I was responding to original post by jmartus, which I probably should have quoted, as I just happened to have a GTX 770. In any case, thanks for your info.
Currently, I'm running some test matches between 6x64, 15x192, and 20x256 NNs on my GTX 1070 system against various other engines. Although slower in nps, the larger NNs are clearly stronger. I would like to get a better idea of how much stronger.
Milos wrote: ↑Thu Aug 23, 2018 10:05 pm
brianr wrote: ↑Thu Aug 23, 2018 9:02 pm
With my GTX-770 and default options with NN 588 I get about 2,750 nps after "hashfull" which is after 130 seconds. Eleven seconds is too short for it too "spin up" and at that point I get about 2,100 nps. So, your results look like mine.
Brian
net 588 is "small" i.e. 15x192 net.
You would get at least 2x smaller nps with "large" i.e. 20x256 net.
Btw, GTX770 is 2.5x slower than 1060.
With GTX770 and 20x256 net I was getting around 1400nps after 300k nodes.
With GTX1060 and 20x256 net I am getting close to 4000nps after 300k nodes. However, my GTX1060 is strongly OC'ed, otherwise it would be the usual 3600nps.
Re: LC0 cuda speed question
Posted: Fri Aug 24, 2018 12:09 am
by cc2150dx
AdminX wrote: ↑Thu Aug 23, 2018 11:55 pm
cc2150dx wrote: ↑Thu Aug 23, 2018 10:27 pm
I'm running LC0 cuda (default settings) with my Geforce GTX 760 and only getting 875 nps. Something doesn't seem right?
That seems about right if you are using 20x256 net. On my Dell Laptop with an Nvidia MX150 GPU I am getting between 865 - 900 nps.
I'm using a 20x256 net. Well, thanks for letting me know
