Dann Corbit wrote: ↑Mon Jan 18, 2021 3:19 pm
...
"The way I see it, an intelligent person who disagrees with me is
probably the most important person I'll interact with on any given
day."
--Billy Chambless
Wise maxim
Dann Corbit wrote: ↑Mon Jan 18, 2021 2:03 pm
I suggest you examine copyright law.
I am right
You are wrong'
Simple as that.
...
Angle wrote: ↑Tue Jan 19, 2021 3:40 am
I only want to say that "NNUE Era" and "Copy-Paste SF NNUE Era" are two basicaly different rhings. The first is beautiful while the second is unworthy. In many discussions here, these two concepts are not separated, which is a great mistake.
The Night Nurse nets (and the Toga ones, for that matter) play quite differently from the SF ones, as the various similarity testers will tell you. Are you saying that all NNUE nets play the same?
Fat Titz by Stockfish, the engine with the bodaciously big net. Remember: size matters. If you want to learn more about this engine just google for "Fat Titz".
Angle wrote: ↑Tue Jan 19, 2021 3:40 am
I only want to say that "NNUE Era" and "Copy-Paste SF NNUE Era" are two basicaly different rhings. The first is beautiful while the second is unworthy. In many discussions here, these two concepts are not separated, which is a great mistake.
Are you saying that all NNUE nets play the same?
I am saying the opposite. It's good to have many original nets with different play styles. I highly appreciate all what you do for training NiNu and Toga nets. Here you can find some of my thoughts about this problem: http://talkchess.com/forum3/viewtopic.p ... ng#p874723
dkappe wrote: ↑Tue Jan 19, 2021 2:18 am
In a few months, everyone will be looking for ways to goose the nets — through size, architecture, etc. Those that use existing tools like tensorflow and pytorch instead of developing them from scratch will win the race(s).
Writing gradient descent optimizers by hand never was a viable long term strategy to begin with. With two major frameworks (TF and PT) and high-level interfaces (Keras, PT Lightning) on top of them, the playing field is level again. A GD optimizer is just another tool, like a compiler, that let's the developer focus on his domain rather than the tech stack underlying it.
Engin wrote: ↑Thu Jan 21, 2021 4:52 pm
every programmers are using alpha beta, null move, LMR now, why the hell can not using NNUE idea now ? fck... GPL rules !
Clearly the complain is more than "Hes using NNUE!". Its "Hes using Stockfish's NNUE code verbatim with no changes". Nothing to do with the GPL.
#WeAreAllDraude #JusticeForDraude #RememberDraude #LeptirBigUltra "Those who can't do, clone instead" - Eduard ( A real life friend, not this forum's Eduard )