Add tempo also to NNUE eval.

[locutus2]

STC:
LLR: 2.93 (-2.94,2.94) {-0.50,1.50}
Total: 10608 W: 1507 L: 1358 D: 7743
Ptnml(0-2): 94, 945, 3074, 1100, 91
https://tests.stockfishchess.org/tests/view/5f2c5921b3ebe5cbfee85b8b

LTC:
LLR: 2.94 (-2.94,2.94) {0.25,1.75}
Total: 7536 W: 556 L: 448 D: 6532
Ptnml(0-2): 9, 383, 2881, 481, 14
https://tests.stockfishchess.org/tests/view/5f2c6f4461e3b6af64881e95

Bench: 4746616

[mstembera]

Congrats!
I wonder if side to move is encoded for NNUE? I know @nodchip mentioned that castling rights for example aren't right now.

[locutus2]

@mstembera
I haven't looked in the code but as i understand the description only the combinations of pieces with the kings are considered as input features.

[mstembera]

I'm not against this PR but it seems like in the longrun the holistic solution is to first encode missing features.

[sf-x]

I wonder if side to move is encoded for NNUE? I know @nodchip mentioned that castling rights for example aren't right now.

It isn't. The board is flipped horizontally if needed so that the same side is always to move.

[Lolligerhans]

I wonder if side to move is encoded for NNUE? I know @nodchip mentioned that castling rights for example aren't right now.

It isn't. The board is flipped horizontally if needed so that the same side is always to move.

Wouldn't that mean that it is implicitly encoded?

[vondele]

Thanks!