Latest Night Nurse released

[dkappe]

The latest Night Nurse, trained on Bad Gyal XL data. The scaling of the net is already corrected, so it can be used directly with the latest SF or Cfish binary without using NetRunner.

https://www.patreon.com/posts/night-nurse-0-4b-46830152

[dkappe]

Wading through a mountain of draws even at 1 min. + 1 sec., you can get a few attractive wins.

[pgn]
[Event "?"]
[Site "?"]
[Date "2021.01.29"]
[Round "11"]
[White "ninu-0.4b"]
[Black "sfdev"]
[Result "1-0"]
[ECO "E94"]
[GameDuration "00:04:17"]
[GameEndTime "2021-01-29T12:34:56.359 CST"]
[GameStartTime "2021-01-29T12:30:39.143 CST"]
[Opening "King's Indian"]
[PlyCount "169"]
[Termination "adjudication"]
[TimeControl "60+1"]
[Variation "Orthodox Variation"]

1. d4 {book} Nf6 {book} 2. c4 {book} g6 {book} 3. Nc3 {book} Bg7 {book}
4. e4 {book} d6 {book} 5. Nf3 {book} O-O {book} 6. Be2 {book} e5 {book}
7. O-O {book} Na6 {book} 8. Be3 {book} Ng4 {book} 9. Bg5 {book} Qe8 {book}
10. dxe5 {+1.06/24 8.7s} Nxe5 {-0.64/24 7.5s} 11. Nd4 {+0.77/22 2.3s}
Be6 {-0.53/21 1.2s} 12. b3 {+0.74/25 7.1s} Nc6 {-0.52/22 2.3s}
13. Be3 {+0.84/22 1.7s} Nc5 {-0.56/21 1.4s} 14. Rb1 {+0.74/25 6.2s}
Bxd4 {-0.50/22 3.4s} 15. Bxd4 {+0.80/22 1.3s} f5 {-0.46/21 0.95s}
16. exf5 {+0.69/22 1.3s} Bxf5 {-0.33/23 1.2s} 17. Rc1 {+0.65/21 1.2s}
Nxd4 {-0.37/24 1.5s} 18. Qxd4 {+0.55/22 2.4s} Qe5 {-0.32/24 1.4s}
19. Qd2 {+0.81/21 1.4s} Ne4 {-0.25/24 1.9s} 20. Nxe4 {+0.64/22 1.8s}
Bxe4 {-0.30/25 1.7s} 21. c5 {+0.76/23 3.0s} d5 {-0.21/22 1.8s}
22. c6 {+0.85/22 0.85s} b6 {-0.23/22 2.4s} 23. Rfe1 {+0.92/21 1.4s}
Rad8 {-0.34/25 3.2s} 24. Bf1 {+1.10/22 1.8s} Qd6 {-0.46/23 3.6s}
25. b4 {+1.07/21 1.2s} Rfe8 {-1.03/26 12s} 26. b5 {+1.11/22 3.2s}
Bf5 {-0.78/24 1.1s} 27. Rcd1 {+1.15/22 3.3s} Kf7 {-0.80/23 1.3s}
28. Qh6 {+1.18/22 2.9s} Kg8 {-0.87/27 2.5s} 29. Qd2 {+1.03/24 2.2s}
Kf7 {-0.82/26 2.1s} 30. Qh6 {+1.21/26 8.7s} Kg8 {-0.68/25 3.5s}
31. h4 {+1.28/23 1.2s} Bc2 {-0.68/24 4.6s} 32. Rc1 {+1.11/24 2.6s}
Be4 {-0.73/22 1.8s} 33. Qd2 {+1.20/24 1.9s} Kg7 {-0.76/22 2.3s}
34. Rcd1 {+1.38/24 1.3s} h5 {-1.59/28 14s} 35. f3 {+1.55/25 3.3s}
Qc5+ {-1.74/24 1.3s} 36. Qf2 {+1.44/23 0.77s} Bf5 {-1.10/22 0.26s}
37. Rxe8 {+1.60/27 1.7s} Rxe8 {-1.26/26 0.77s} 38. Qxc5 {+1.62/27 1.1s}
bxc5 {-1.44/24 0.94s} 39. Rxd5 {+1.51/28 1.5s} Re1 {-1.37/23 0.52s}
40. Kf2 {+1.60/27 0.88s} Ra1 {-1.40/25 0.95s} 41. Rxc5 {+1.73/26 0.92s}
Rxa2+ {-1.33/23 0.46s} 42. Ke3 {+1.74/25 2.7s} Kf6 {-1.13/22 0.48s}
43. Rc3 {+1.59/24 4.3s} Ke5 {-1.36/25 2.7s} 44. g3 {+1.63/23 1.1s}
Ra4 {-1.37/25 1.8s} 45. Be2 {+1.92/21 0.67s} Be6 {-1.25/22 0.26s}
46. Bd3 {+1.87/23 0.81s} Bf5 {-1.22/21 0.56s} 47. Be2 {+1.94/22 0.48s}
Be6 {-1.56/24 1.0s} 48. Rc2 {+1.85/22 0.81s} Ra3+ {-1.65/24 2.3s}
49. Kf2 {+1.72/23 0.79s} Bd5 {-1.60/22 0.28s} 50. Rd2 {+1.91/24 0.69s}
Kd6 {-1.66/22 0.42s} 51. Rd3 {+1.93/25 3.5s} Ra5 {-1.57/27 2.8s}
52. Re3 {+2.02/22 0.43s} Ra2 {-1.76/23 0.46s} 53. Rc3 {+2.05/23 0.51s}
Ke5 {-1.85/24 0.54s} 54. Ke3 {+1.89/24 0.85s} Be6 {-1.78/26 1.5s}
55. Bd3 {+1.95/21 0.69s} Bf7 {-1.69/24 1.5s} 56. Be4 {+1.86/22 1.7s}
Ra5 {-1.63/23 0.28s} 57. Rc5+ {+2.25/20 0.37s} Kf6 {-1.65/24 2.0s}
58. Bd5 {+2.55/19 0.55s} Bxd5 {-1.28/22 0.31s} 59. Rxd5 {+2.66/21 0.63s}
Ke6 {-1.26/20 0.37s} 60. Kd4 {+2.59/23 0.60s} Ra4+ {-2.96/21 1.3s}
61. Kc5 {+2.61/22 0.62s} Ra5 {-2.81/23 0.90s} 62. Rd7 {+3.11/23 1.7s}
a6 {-2.98/23 0.53s} 63. Rg7 {+3.26/21 0.41s} Rxb5+ {-2.51/22 1.1s}
64. Kc4 {+3.22/25 1.5s} Rb8 {-3.89/25 4.0s} 65. Rxg6+ {+3.27/24 0.60s}
Ke7 {-3.50/22 0.11s} 66. Rg7+ {+3.55/24 0.98s} Kd6 {-4.30/24 0.85s}
67. Rg6+ {+3.53/23 0.74s} Ke7 {-4.11/23 0.28s} 68. Rg5 {+3.80/20 0.69s}
Rh8 {-3.74/23 0.44s} 69. Kc5 {+3.76/23 0.93s} Rh7 {-4.50/24 1.1s}
70. Rd5 {+6.97/22 0.69s} Ke6 {-8.75/23 1.3s} 71. Rd7 {+8.71/20 0.77s}
Rh8 {-9.56/20 1.2s} 72. Rxc7 {+8.87/19 1.1s} Rf8 {-9.87/19 0.84s}
73. Rh7 {+9.39/18 0.96s} a5 {-10.02/20 1.7s} 74. Rh6+ {+10.19/18 0.96s}
Kf7 {-10.33/22 2.6s} 75. Rxh5 {+10.61/18 0.71s} Kg6 {-10.59/19 1.3s}
76. Rg5+ {+10.70/18 0.82s} Kh6 {-10.62/19 0.45s} 77. Kb6 {+11.18/21 0.80s}
a4 {-10.62/19 0.26s} 78. c7 {+16.87/23 0.76s} a3 {-24.29/24 2.1s}
79. Ra5 {+51.03/23 0.90s} a2 {-24.29/15 0.18s} 80. Rxa2 {+152.58/25 1.2s}
Kg7 {-152.59/22 1.4s} 81. Ra5 {+152.61/37 0.82s} Re8 {-152.61/24 0.69s}
82. Rc5 {+152.62/39 1.6s} Re6+ {-152.61/27 0.24s} 83. Rc6 {+152.63/41 0.63s}
Re8 {-152.63/33 0.24s} 84. c8=Q {+M29/37 1.5s} Rxc8 {-152.64/36 0.36s}
85. Rxc8 {+M25/40 0.80s, White wins by adjudication: SyzygyTB} 1-0
[/pgn]

[bastiball]

Thanks for the new net of Night Nurse !!!

[JohnW]

Please excuse my ignorance but can you explain how you use this? Using the Fritz gui and CFish 11 I don't see any parameter setting to enter the file name. In Stockfish if I tried replacing the nnue file setting with this bin file Stockfish just crashes.

Thanks for the help..

[dkappe]

Please excuse my ignorance but can you explain how you use this? Using the Fritz gui and CFish 11 I don't see any parameter setting to enter the file name. In Stockfish if I tried replacing the nnue file setting with this bin file Stockfish just crashes.

Thanks for the help..

I think you need at least Cfish 12 or Stockfish 12. The uci param is EvalFile.

[MikeB]

The latest Night Nurse, trained on Bad Gyal XL data. The scaling of the net is already corrected, so it can be used directly with the latest SF or Cfish binary without using NetRunner.

https://www.patreon.com/posts/night-nurse-0-4b-46830152

+1 ... finally... a good reason to re-release the Harmon chess app (droidfish clone), with new piece sets from Bryan Whitby ( thank you Bryan!) and another piece set based on the Isle of Lewis Chessmen ...

https://www.nms.ac.uk/explore-our-colle ... ss-pieces/

[Javier Ros]

The latest Night Nurse, trained on Bad Gyal XL data. The scaling of the net is already corrected, so it can be used directly with the latest SF or Cfish binary without using NetRunner.

https://www.patreon.com/posts/night-nurse-0-4b-46830152

Thank you for the release.

Is possible to run it in Android?

[dkappe]

Thank you for the release.

Is possible to run it in Android?

See MikeB’s post above.