Not logged inCSS-Forum
Forum CSS-Online Help Search Login
CSS-Shop Impressum Datenschutz
Up Topic Hauptforen / CSS-Forum / Endspielschwäche von Lc0 ?
- - By Reinhold Stibi Date 2023-05-06 12:31 Edited 2023-05-06 12:46
[Event "Wertungspartie, 5 min"]
[Site "Maschinenraum"]
[Date "2023.05.06"]
[Round "?"]
[White "Hippo100, Stockfish dev-20230"]
[Black "Ivers, Lc0 v0.30.0-rc1"]
[Result "1-0"]
[ECO "C54"]
[WhiteElo "3027"]
[BlackElo "3044"]
[Annotator "0.18;0.08"]
[PlyCount "149"]
[EventDate "2023.05.06"]
[SourceTitle "playchess.com"]
[TimeControl "300"]

{Lc0 v0.30.0-rc1 (2 cores): 24.7 plies; 157kN/s AMD Ryzen 9 5950X 16-Core
Processor       3394MHz, (16 cores, 32 threads), R-K1.ctg, 2048 MB} 1. e4 {B 0}
e5 {0.08/15 3} 2. Nf3 {B 0} Nc6 {0.09/18 3} 3. Bc4 {B 0 (Bb5)} Bc5 {0.06/11 5}
4. O-O {B 0 (c3)} Nf6 {0.07/13 8} 5. d3 {B 0} d6 {0.07/10 8} 6. c3 {B 0} h6 {
0.08/10 7} 7. b4 {B 0 (Re1)} Bb6 {0.09/12 2} 8. Nbd2 {B 0 (a4)} Ne7 {0.09/13 9}
9. a4 {0.18/37 37 (Re1)} a5 {0.07/17 0} 10. Re1 {0.12/35 10 (b5)} O-O {0.05/19
3} 11. b5 {0.21/33 3} Ng6 {0.05/18 1} 12. d4 {B 0 (h3)} Re8 {0.04/20 7} 13. Qb3
{B 0 (h3)} Qd7 {0.01/16 12} 14. Bb2 {B 0} Nf4 {0.05/18 2} 15. h3 {B 0} Nxg2 {
0.16/19 5} 16. Kxg2 {B 0} Qxh3+ {0.20/21 1} 17. Kg1 {B 0} Qh5 {0.20/23 10} 18.
Bf1 {B 0} Qg6+ {0.25/23 1} 19. Bg2 {B 0} Nh5 {0.27/24 5} 20. Nf1 {B 0} Bh3 {
0.34/25 11} 21. Ng3 {B 0 (Nh4)} Nxg3 {0.23/22 5} 22. Bxh3 {B 0} Ne2+ {0.22/22 3
} 23. Kh1 {B 0} Qxe4 {0.26/25 5} 24. Bg2 {B 0} exd4 {0.28/25 3} 25. cxd4 {B 0}
Qg4 {0.27/25 1} 26. Ng1 {0.43/32 5} Nf4 {0.29/26 3} 27. Qf3 {0.44/40 10} Qxf3 {
0.26/28 0} 28. Bxf3 {0.53/43 48} Nd3 {0.23/38 0} 29. Rxe8+ {0.44/35 3} Rxe8 {
0.23/33 0} 30. Rb1 {0.52/33 2} Nxf2+ {0.24/31 0} 31. Kg2 {0.50/36 3} Ne4 {
0.23/33 0} 32. Re1 {0.18/39 8} d5 {0.23/26 0} 33. Nh3 {0.12/40 5} c6 {0.22/27 0
} 34. bxc6 {0.12/39 3} bxc6 {0.23/27 0} 35. Nf2 {0.14/38 3} f5 {0.23/21 0} 36.
Bh5 {0.06/39 3} Re6 {0.23/26 8} 37. Nxe4 {0.10/40 0} g6 {0.22/25 7} 38. Bc3 {
0.09/39 0} dxe4 {0.23/26 11} 39. Be2 {0.08/41 0} Rd6 {0.23/33 0} 40. Bc4+ {
0.04/39 3} Kf8 {0.27/32 13} 41. Rb1 {0.07/43 0} Bxd4 {0.27/22 0} 42. Bxa5 {
0.00/52 3} Rd7 {0.31/34 23} 43. Be1 {0.00/34 11 (Bb4+)} g5 {0.14/24 17} 44. a5
{0.34/29 2} f4 {0.15/27 3} 45. a6 {0.40/31 0} f3+ {0.17/29 1} 46. Kh3 {0.48/28
5} h5 {0.72/29 8} 47. Rb7 {1.91/28 2} Ke7 {0.92/29 3} 48. Bb4+ {3.21/27 2} Kd8
{1.06/30 2} 49. Bc3 {5.64/27 17} g4+ {1.41/30 0} 50. Kg3 {6.02/22 3} Ba7 {
1.43/30 0} 51. Bf6+ {6.11/21 2} Ke8 {1.76/28 4} 52. Bg5 {6.41/21 0} e3 {
1.80/26 2} 53. Rxa7 {6.98/21 2} Rxa7 {1.87/25 1} 54. Bxe3 {7.80/20 1} Ra8 {
1.89/23 3} 55. a7 {8.64/18 1} Kd7 {2.28/22 36} 56. Ba6 {199.49/34 0} Kc7 {
2.37/19 0} 57. Kh4 {199.61/31 2} g3 {2.44/18 0} 58. Kxg3 {199.65/30 2} h4+ {
2.57/17 1} 59. Kxh4 {199.74/60 2} Rf8 {2.86/17 1} 60. Kg4 {199.79/58 2 (Bf2)}
f2 {3.52/14 3} 61. Bxf2 {199.84/49 0} Rg8+ {4.26/15 2} 62. Kf3 {199.86/52 0}
Rf8+ {4.64/14 0} 63. Ke4 {199.88/54 1} Ra8 {5.68/13 1} 64. Kd3 {199.92/53 16}
c5 {8.56/12 0} 65. Bxc5 {#26/43 1} Kc6 {9.04/10 0} 66. Be3 {5/1 0 (Bd4)} Re8 {
11.24/8 2} 67. Kc4 {#23/40 2 (Kc2)} Ra8 {8.20/8 0} 68. Kb4 {5/0 0 (Kb3)} Rd8 {
6.01/7 0} 69. Bf1 {#19/47 1 (Kc3)} Kb7 {14.85/6 0} 70. Bg2+ {#15/48 1} Kc7 {
31.13/3 0} 71. Bf4+ {#13/56 1 (Kb5)} Kb6 {18.49/7 0} 72. a8=Q {#12/48 1 (Bb8)}
Rxa8 {#120/1 0} 73. Bxa8 {10/1 0} Ka7 {20.49/6 0} 74. Bc6 {9/1 0 (Be4)} Kb6 {
24.07/6 0} 75. Bb5 {8/0 0 (Bd7) Ivers,Lc0 v0.30.0-rc1 resigns (Lag: Av=0.13s,
max=0.5s)} 1-0



46... h5 ?  +0.72  29  8 Sek.  Man bedenke dass dieser grober Fehler bei Tiefe 29  und nach 8 Sek. gemacht wurde,
                                           also  nicht in Zeitnot.

Richtig: 46 ...La7 !   + 0.03

48.Tb7  + 1.91  28   2 Sek.
Parent - - By Reinhold Stibi Date 2023-05-06 16:09 Edited 2023-05-06 16:18
5k2/3r4/P1p4p/6p1/2Bbp3/5p1K/8/1R2B3 b - - 0 1

Analysis by Lc0 v0.30.0-dag+git.1842e13d N 365  =TCEC 2023 Version  mit zwei RTX 2070 im Verbund
                                                                            entspricht einer RTX 3070 Ti


46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.Lf6+ Ke8 52.Lg5 e3 53.Txa7 Txa7 54.Lxe3 Ta8 55.a7 Kd7 56.La6 Kc7 57.Kh4 g3 58.Kxg3 Tg8+ 59.Kh4 Tf8 60.Kxh5 f2 61.Lxf2 Ta8 62.Kg6 Tf8 63.Lc5 Td8 64.Kf5
  Weiß steht besser: +/- (1.00)  Tiefe: 20/40   00:00:05  29kN, tb=111
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.Lf6+ Ke8 52.Lg5 e3 53.Txa7 Txa7 54.Lxe3 Ta8 55.a7 Kd7 56.La6 Kc7 57.Kh4 g3 58.Kxg3 Tg8+ 59.Kh4 Tf8 60.Kxh5 f2 61.Lxf2 Ta8 62.Kg6 Tf8 63.Lc5 Td8 64.Kf5
  Weiß steht besser: +/- (1.02)  Tiefe: 20/41   00:00:05  33kN, tb=122
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.Lf6+ Ke8 52.Lg5 e3 53.Txa7 Txa7 54.Lxe3 Ta8 55.a7 Kd7 56.La6 Kc7 57.Kh4 g3 58.Kxg3 Tg8+ 59.Kh4 Tf8 60.Kxh5 f2 61.Lxf2 Ta8 62.Kg6 Tf8 63.Lc5 Td8 64.Kf5
  Weiß steht besser: +/- (1.02)  Tiefe: 20/42   00:00:05  33kN, tb=123
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.Lf6+ Ke8 52.Lg5 e3 53.Txa7 Txa7 54.Lxe3 Ta8 55.a7 Kd7 56.La6 Kc7 57.Kh4 g3 58.Kxg3 Tg8+ 59.Kh4 Tf8 60.Kxh5 f2 61.Lxf2 Ta8 62.Kg6 Tf8 63.Lg3+ Kb6 64.Lb8 Tg8+
  Weiß steht besser: +/- (1.05)  Tiefe: 20/42   00:00:06  38kN, tb=138
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.Lf6+ Ke8 52.Lg5 e3 53.Txa7 Txa7 54.Lxe3 Ta8 55.a7 Kd7 56.La6 Kc7 57.Kh4 g3 58.Kxg3 Tg8+ 59.Kh4 Tf8 60.Kxh5 f2 61.Lxf2 Ta8 62.Kg6 Tg8+ 63.Kf5 Tf8+ 64.Ke4 Td8 65.Lc5
  Weiß steht besser: +/- (1.09)  Tiefe: 20/42   00:00:08  56kN, tb=235
46...La7 47.Tb7 Ke8 48.Kg4 Txb7 49.axb7 h5+ 50.Kxg5 e3 51.Lg3 h4 52.Kxh4 e2 53.Ld3 Kd8 54.La6 Lb8 55.Le1 Kd7 56.Kg4 f2 57.Lxf2 Kc7 58.Kf3 e1D 59.Lxe1 Kb6 60.Lg3 Kxa6
  Weiß steht etwas besser: +/= (0.41)  Tiefe: 20/42   00:00:11  82kN, tb=477

46...La7 47.Tb7 Ke8 48.Kg4 Txb7 49.axb7 h5+ 50.Kxg5 e3 51.Lg3 h4 52.Kxh4 e2 53.Ld3 Kd8 54.La6 Lb8 55.Le1 Kd7 56.Kg4 f2 57.Lxf2 Kc7 58.Lg3+ Kb6 59.Lxe2 Lxg3
  Weiß hat minimalen Vorteil: = (0.30)  Tiefe: 20/42   00:00:16  126kN, tb=869
46...La7 47.Tb7 Ke8 48.Kg4 Txb7 49.axb7 h5+ 50.Kxg5 e3 51.Lg3 h4 52.Kxh4 e2 53.Ld3 Kd8 54.La6 Lb8 55.Le1 Kd7 56.Kg4 f2 57.Lxf2 Kc7 58.Kf3 e1D 59.Lxe1 Kb6 60.Lg3 Kxa6
  Weiß hat minimalen Vorteil: = (0.29)  Tiefe: 20/42   00:00:21  162kN, tb=948
46...La7 47.Tb7 Ke8 48.Kg4 Txb7 49.axb7 h5+ 50.Kxg5 e3 51.Lg3 h4 52.Kxh4 e2 53.Ld3 Kd8 54.La6 Lb8 55.Le1 Kd7 56.Kg4 f2 57.Lxf2 Kc7 58.Kf3 e1D 59.Lxe1 Kb6 60.Lg3 Kxa6
  Weiß hat minimalen Vorteil: = (0.27)  Tiefe: 20/42   00:00:26  201kN, tb=1151
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 g4+ 50.Kg3 f2+ 51.Kg2 c5 52.Lxe4 Td6 53.Lc3 Txa6 54.Ld5 c4 55.Lxc4 Ta3 56.Lf6 Tf3 57.Lh4 Tc3 58.Lf1 Lc5 59.Tb7 Ld6 60.Th7 Th3 61.Lxf2 Th2+ 62.Kg1 g3
  Weiß hat minimalen Vorteil: = (0.26)  Tiefe: 20/42   00:00:29  223kN, tb=1252

Für den richtigen Zug braucht Lc0 11 Sek. !

5k2/3r4/P1p4p/6p1/2Bbp3/5p1K/8/1R2B3 b - - 0 1

Analysis by Stockfish dev-20230503-72d542f0  Ca avx2 nnue   mit Ryzen 3970X reduziert auf 16 CPUs


46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 Kd8 57.Kxf3 e1T 58.Lxe1 Kc7 59.Lf2 Ta8 60.Ke4 Td8 61.La7 Ta8 62.Lf2
  Die Stellung ist ausgeglichen: = (0.00)  Tiefe: 32/46   00:00:01  13752kN, tb=65807
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 Kd8 57.Kxf3 e1T 58.Lxe1 Kc7 59.Lf2 Ta8 60.Ke4 Td8 61.La7 Ta8 62.Le3 Td8 63.Lf2 Tf8
  Die Stellung ist ausgeglichen: = (0.00)  Tiefe: 33/46   00:00:01  15623kN, tb=69566
46...La7
  Weiß steht besser: +/- (0.98 ++)  Tiefe: 34/46   00:00:04  79374kN, tb=264630

46...La7
  Weiß steht etwas besser: +/= (0.52 ++)  Tiefe: 34/46   00:00:04  83624kN, tb=281039
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.Tb7+ Kd6 54.La5 g4+ 55.Kxg4 f2 56.Tb1 h5+ 57.Kf4 e3 58.Kf3 Tc2 59.Kxe3 c4+ 60.Ke4 h4 61.Lb4+ Kc7 62.Le6 Kc6 63.Ld5+ Kc7 64.Kf3 c3 65.Tf1 h3 66.Le4 Ta2
  Weiß hat minimalen Vorteil: = (0.29)  Tiefe: 34/46   00:00:05  111MN, tb=388162
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.Tb7+ Kd6 54.La5 g4+ 55.Kxg4 f2 56.Tb1 h5+ 57.Kf4 e3 58.Kf3 Tc2 59.Kxe3 c4+ 60.Ke4 h4 61.Lb4+ Kc7 62.Le6 Kc6 63.Ld5+ Kc7 64.Kf3 c3 65.Tf1 h3 66.Le4 Ta2
  Weiß steht etwas besser: +/= (0.40 ++)  Tiefe: 35/45   00:00:07  137MN, tb=438641
Tiefe: 36/60   00:00:11  226MN, tb=641099
Parent - By Reinhold Stibi Date 2023-05-06 16:26
5k2/3r4/P1p4p/6p1/2Bbp3/5p1K/8/1R2B3 b - - 0 1

Analysis by Stockfish dev-20230503-72d542f0  Ca avx2 nnue   mit Ryzen 3970X 32 Kerner/64 Threads
                                                                                           also volle Pulle


46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 Td7 57.Kxf3 e1D 58.Lxe1 Kd6 59.Lf2 Kc7 60.La7 Th7 61.Ke2 Th3 62.Ke1 Th4 63.Ld3
  Die Stellung ist ausgeglichen: = (0.14)  Tiefe: 19/36   00:00:00  7336kN, tb=18888
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 e1D 57.Lxe1 Kd6 58.Lf2 Tg7+ 59.Kf4 Kc7 60.Kxf3 Te7 61.La7 Te1 62.Le3 Ta1
  Die Stellung ist ausgeglichen: = (0.12)  Tiefe: 20/36   00:00:00  7987kN, tb=21124
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 e1D 57.Lxe1 Kd6 58.Lf2 Tg7+ 59.Kxf3 Kc7 60.Lc5 Th7 61.Kg4 Th8 62.Kg3 Td8 63.Lf2 Tc8 64.Ld4 Ta8
  Die Stellung ist ausgeglichen: = (0.11)  Tiefe: 21/41   00:00:00  9308kN, tb=26844
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 e1D 57.Lxe1 Kd6 58.Lf2 Tg7+ 59.Kxf3 Kc7 60.Le3 Th7 61.Kg3 Th8 62.Kf2 Th4 63.Ld3 Th8 64.Ke2 Td8 65.Lc4
  Die Stellung ist ausgeglichen: = (0.10)  Tiefe: 22/41   00:00:00  11662kN, tb=39754
46...h5 47.Tb7 Ke7 48.Lb4+ Kd8 49.Lc3 g4+ 50.Kg3 La7 51.La5+ Ke7 52.Lc7 e3 53.Txa7 h4+ 54.Kxh4 e2 55.Lg3 Txa7 56.Kxg4 e1D 57.Lxe1 Kd6 58.Lf2 Tg7+ 59.Kf4 Kc7 60.La7 f2 61.Lxf2 Th7 62.Ke3 Th3+ 63.Kd4 Kb8 64.Ke4 Ta3 65.Lc5
  Die Stellung ist ausgeglichen: = (0.08)  Tiefe: 23/43   00:00:00  13250kN, tb=44954
46...La7 47.Lb4+ Ke8 48.Le6 Td4 49.Lg4 Td3 50.Lf5 Te3 51.Lg6+ Kd7 52.Lf7 c5 53.Ld2 Td3 54.Tb7+ Kd6 55.La5 g4+ 56.Kh2 e3 57.Lc4 Ta3 58.Lc7+ Kc6 59.Lb5+ Kd5 60.Txa7 Ta2+ 61.Kg3 e2 62.Kf2
  Weiß hat minimalen Vorteil: = (0.29)  Tiefe: 24/43   00:00:01  97679kN, tb=262234

46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.Tb7+ Kd6 54.La5 g4+ 55.Kxg4 f2 56.Tb1 Te3 57.Ld2 Te2 58.Lf4+ Kc6 59.Lg3 Tb2 60.Tc1 Tc2 61.Le8+ Kd5 62.Td1+ Kc4 63.Lc6 Te2 64.Tc1+ Kb4 65.Ld5 c4 66.Ld6+ Ka4 67.Lxc4 Te1 68.Tc3 Tg1+ 69.Kf5 f1D+ 70.Lxf1
  Weiß hat minimalen Vorteil: = (0.25)  Tiefe: 25/49   00:00:02  122MN, tb=304304
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.La5 Kc6 54.Lc4 Tc2 55.Lb5+ Kd6 56.Td1+ Ke6 57.Ld7+ Ke5 58.Lc7+ Kf6 59.Td6+ Ke7 60.Lf5 c4 61.Td7+ Kf6 62.Lxe4 f2 63.Lg2 Ke6 64.Td6+ Ke7 65.Txh6 Tc1 66.La5 g4+ 67.Kg3
  Weiß hat minimalen Vorteil: = (0.28)  Tiefe: 26/53   00:00:02  137MN, tb=324426
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.La5 Kc6 54.Lc4 Tc2 55.Lb5+ Kd6 56.Td1+ Ke6 57.Ld7+ Ke5 58.Lc7+ Kf6 59.Td6+ Ke7 60.Lf5 c4 61.Td7+ Kf6 62.Lxe4 f2 63.Lg2 Ke6 64.Td6+ Ke7 65.Lf1 c3 66.Txh6 Tb2 67.Th7+ Ke6 68.Lc4+ Kf5
  Weiß steht etwas besser: +/= (0.31)  Tiefe: 27/53   00:00:02  153MN, tb=341900
46...La7 47.Lb4+ Ke8 48.Le6 Td3 49.Lf5 Te3 50.Lg6+ Kd7 51.Lf7 c5 52.Ld2 Te2 53.La5 Kc6 54.Lc4 Tc2 55.Lb5+ Kd6 56.Td1+ Ke6 57.Ld7+ Ke5 58.Lc7+ Kf6 59.Td6+ Ke7 60.Lf5 c4 61.Td7+ Kf6 62.Lxe4 f2 63.Lg2 Ke6 64.Td6+ Ke7 65.Lf1 Ta2 66.Txh6 c3 67.Ld6+ Kf7 68.Lb4 c2 69.Ld2 Ta1 70.Kg2 Lc5 71.Tc6 c1T 72.Lxc1 Txc1
  Weiß hat minimalen Vorteil: = (0.27)  Tiefe: 28/61   00:00:03  186MN, tb=395404

Lösung in 1 Sek. ist schon prächtig im Gegensatz zu Lc0 mit 11 Sek.

Parent - - By Torsten Cuber Date 2023-05-06 17:48 Upvotes 1
Lieber Reinhold, nur zur Info:
Auf meinem XIAOMI 11 Smartphone findet Stockfish 15 den richtigen Zug 46.- La7 in 0:07 Minuten, d.h in 7 Sekunden. Zum Vergleich: Der Super-Taktiker Crystal 5 findet dies in 0:26 Minuten (26 Sekunden).
Parent - - By Reinhold Stibi Date 2023-05-06 19:22
Danke Torsten, deine Ergebnisse unterstreichen mit den Schwachpunkt von Lc0.

Dies ist eine Stellung, die der Mensch auch gut nachvollziehen kann.
Parent - By Reinhold Stibi Date 2023-05-07 08:47
Auf Schach.de spielt hauptsächlich der Norweger Ivers mit Lc0 , momentan wie so oft, als Einziger.

Hat schon wieder eine Partie mit Lc0 v0.30 verloren und ist in der Tabelle nun weit hinten.

Mit mittelstarker Hardware scheint Lc0 gegen Stockfish nicht so recht dagegenhalten zu können.
Up Topic Hauptforen / CSS-Forum / Endspielschwäche von Lc0 ?

Powered by mwForum 2.29.3 © 1999-2014 Markus Wichitill