Search found 1336 matches

by xr_a_y
Wed Oct 14, 2020 1:57 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Black crushing white, weird ?
Replies: 7
Views: 711

Re: Black crushing white, weird ?

really? I thought that it's common to swap board from stm's point of view and then simply evaluate, I don't see how there could be any bias in that case, but I haven't studied NNUE at all. I always want symmetric evals in my engine anyway, for debugging purposes EDIT: of course, this would make the...
by xr_a_y
Wed Oct 14, 2020 12:52 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Black crushing white, weird ?
Replies: 7
Views: 711

Re: Black crushing white, weird ?

Same thing with Hert500 opening book => not a book bias Same thing with 11 ply instead of 10 => not a even/odd effect bias NOT the same thing (white clearly winning) with "napping nexus" instead of "nascent nutrient" => this is a net bias !! Conclusion nascent nutrient is good playing black ... less...
by xr_a_y
Wed Oct 14, 2020 12:09 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Black crushing white, weird ?
Replies: 7
Views: 711

Black crushing white, weird ?

Generating data for NNUE RL training with nascent nutrient I got this using the 2moves_LT_1000 opening book Score of minic_2.51_nascent_nutrient vs minic_2.51_nascent_nutrient2: 347 - 327 - 702 [0.507] ... minic_2.51_nascent_nutrient playing White: 111 - 213 - 364 [0.426] 688 ... minic_2.51_nascent_...
by xr_a_y
Wed Oct 14, 2020 8:50 am
Forum: Computer Chess Club: General Topics
Topic: Minic version 2
Replies: 157
Views: 41934

Re: Minic version 2

Minic 2.51 is available.

https://github.com/tryingsomestuff/Mini ... s/tag/2.51

NNUE performance is just crazy with this release, as shown in previous post.

Have fun.
by xr_a_y
Mon Oct 12, 2020 4:54 pm
Forum: Computer Chess Club: General Topics
Topic: Minic version 2
Replies: 157
Views: 41934

Re: Minic version 2

And threading seems ok. SF11 gauntlet at 10s+0.1 MinicNNUE using nn-97f742aaefcd Rank Name Elo +/- Games Score Draw 0 stockfish.11 1 minic_dev_uci_nnue_8t 136 38 196 68.6% 42.3% 2 minic_dev_uci_nnue_4t 52 35 197 57.4% 47.7% 3 minic_dev_uci_nnue_2t -44 33 197 43.7% 53.8% 4 stockfish.10 -44 30 197 43....
by xr_a_y
Mon Oct 12, 2020 2:51 pm
Forum: Computer Chess Club: General Topics
Topic: Minic version 2
Replies: 157
Views: 41934

Re: Minic version 2

Confirmed at longer TC, here 4min+1s, versus SF9 0 stockfish.9 1 minic_dev_nnue_nn-97f742aaefcd -3 38 130 49.6% 59.2% 2 minic_dev_nnue_napping_nexus -73 35 130 39.6% 63.8% 3 minic_dev_nnue_nascent_nutrient -184 48 130 25.8% 39.2% Have you already tried to re-evaluate your minic based data with Mini...
by xr_a_y
Mon Oct 12, 2020 5:54 am
Forum: Computer Chess Club: General Topics
Topic: Minic version 2
Replies: 157
Views: 41934

Re: Minic version 2

Confirmed at longer TC, here 4min+1s, versus SF9 0 stockfish.9 1 minic_dev_nnue_nn-97f742aaefcd -3 38 130 49.6% 59.2% 2 minic_dev_nnue_napping_nexus -73 35 130 39.6% 63.8% 3 minic_dev_nnue_nascent_nutrient -184 48 130 25.8% 39.2%
by xr_a_y
Sun Oct 11, 2020 11:31 am
Forum: Computer Chess Club: General Topics
Topic: Minic version 2
Replies: 157
Views: 41934

Re: Minic version 2

Some news about Minic. I recently discovered and fixed a very very bad design in MinicNNUE implementation. Don't want to dig too much into details but as a copy/make engine, I don't want to store position relative stuff in a separate "state" thing (often used later as a sort of stack in other engine...
by xr_a_y
Sat Oct 10, 2020 8:19 pm
Forum: Computer Chess Club: General Topics
Topic: Dangerous turn
Replies: 31
Views: 2834

Re: Dangerous turn

And Texel-tuned engines are often optimized to game outcome (whatever engine played the game), not the score itself. This doesn't matter much. Infact training against game outcome (Z) is better than score (Q) because Z can potentially incorporate long-term effects. A0 used Z entirely but you can al...
by xr_a_y
Sat Oct 10, 2020 8:03 pm
Forum: Computer Chess Club: General Topics
Topic: NNUE flip vs rotate in Minic/Igel/Rubichess/etc
Replies: 2
Views: 332

Re: NNUE flip vs rotate in Minic/Igel/Rubichess/etc

Minic is staying to "kVersion = 0x7AF32F16u" and won't be able to read recent nets.

I tried building net with new version of "orient" function, but without success ...