Sergio Vieri second net is out

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
perejaslav
Posts: 215
Joined: Sat Mar 18, 2006 3:01 am
Location: Cold

Re: Sergio Vieri second net is out

Post by perejaslav » Wed Jul 29, 2020 9:30 am

Code: Select all


# PLAYER                                       :  RATING  ERROR    POINTS  PLAYED   (%)  CFS(%)      W       D       L  D(%)
   1 SF-NNUE-20200728-1442.bin                 :  3664.3    4.5    5898.0   10000    59     100   3236    5324    1440    53
   2 SF-NNUE-20200728-0207.bin                 :  3656.1    4.6    5786.0   10000    58      50   3076    5420    1504    54
   3 SF-NNUE-20200728-1817.bin                 :  3656.1    4.6    5785.5   10000    58      62   3081    5409    1510    54
   4 SF-NNUE-20200728-0633.bin                 :  3655.0    4.7    5770.0   10000    58      66   3061    5418    1521    54
   5 SF-NNUE-20200728-1104.bin                 :  3653.6    4.5    5751.5   10000    58      89   3043    5417    1540    54
   6 SF-NNUE-20200727-2151.bin                 :  3649.9    5.0    5699.5   10000    57      68   2980    5439    1581    54
   7 SF-NNUE-20200729-0335.bin                 :  3643.9   24.1     273.5     487    56      59    143     261      83    54
   8 SF-NNUE-20200727-1540.bin                 :  3641.0    4.5    5576.0   10000    56      71   2882    5388    1730    54
   9 SF-NNUE-experimental-20200722-1621.bin    :  3639.4    4.3    5553.5   10000    56      77   2744    5619    1637    56
  10 SF-NNUE-20200727-0928.bin                 :  3636.9    4.3    5519.5   10000    55      54   2779    5481    1740    55
  11 SF-NNUE-20200725-1313.bin                 :  3636.6    4.4    5515.5   10000    55      53   2756    5519    1725    55
  12 SF-NNUE-20200722-2141.bin                 :  3636.4    4.5    5512.5   10000    55      54   2765    5495    1740    55
  13 SF-NNUE-experimental-20200722-1130.bin    :  3636.1    4.4    5507.5   10000    55      61   2757    5501    1742    55
  14 SF-NNUE-20200723-0511.bin                 :  3635.1    4.7    5493.5   10000    55      52   2721    5545    1734    55
  15 SF-NNUE-20200722-1517.bin                 :  3634.9    4.6    5491.5   10000    55      55   2679    5625    1696    56
  16 SF-NNUE-20200725-2051.bin                 :  3634.5    4.9    5485.5   10000    55      54   2749    5473    1778    55
  17 SF-NNUE-20200722-0944.bin                 :  3634.1    4.5    5480.5   10000    55      57   2688    5585    1727    56
  18 SF-NNUE-experimental-20200722-2115.bin    :  3633.5    4.7    5472.0   10000    55      51   2646    5652    1702    57
  19 SF-NNUE-20200724-1732.bin                 :  3633.4    5.2    5470.5   10000    55      51   2753    5435    1812    54
  20 SF-NNUE-20200724-0123.bin                 :  3633.3    4.4    5469.5   10000    55      51   2675    5589    1736    56
  21 SF-NNUE-experimental-20200725-2237.bin    :  3633.2    4.4    5468.0   10000    55      53   2729    5478    1793    55
  22 SF-NNUE-20200727-0332.bin                 :  3633.0    4.7    5465.0   10000    55      53   2765    5400    1835    54
  23 SF-NNUE-20200723-1134.bin                 :  3632.7    4.4    5461.0   10000    55      51   2650    5622    1728    56
  24 SF-NNUE-20200726-0504.bin                 :  3632.7    4.3    5460.0   10000    55      61   2686    5548    1766    55
  25 SF-NNUE-20200724-1240.bin                 :  3631.9    4.3    5449.0   10000    54      52   2690    5518    1792    55
  26 SF-NNUE-20200724-2343.bin                 :  3631.7    4.5    5446.5   10000    54      60   2674    5545    1781    55
  27 SF-NNUE-20200726-1135.bin                 :  3630.8    4.8    5434.5   10000    54      50   2689    5491    1820    55
  28 SF-NNUE-20200724-0640.bin                 :  3630.8    4.4    5434.0   10000    54      50   2677    5514    1809    55
  29 SF-NNUE-20200722-0359.bin                 :  3630.8    4.6    5433.5   10000    54      61   2641    5585    1774    56
  30 SF-NNUE-20200721-2323.bin                 :  3629.8    4.3    5420.0   10000    54      57   2617    5606    1777    56
  31 SF-NNUE-20200724-2344.bin                 :  3629.2    4.2    5412.0   10000    54      51   2673    5478    1849    55
  32 SF-NNUE-experimental-20200724-2244.bin    :  3629.1    4.8    5410.5   10000    54      51   2691    5439    1870    54
  33 SF-NNUE-experimental-20200723-1334.bin    :  3629.0    4.7    5409.0   10000    54      52   2678    5462    1860    55
  34 SF-NNUE-20200725-0545.bin                 :  3628.9    4.3    5406.5   10000    54      66   2675    5463    1862    55
  35 SF-NNUE-experimental-20200724-0650.bin    :  3627.6    4.3    5389.0   10000    54      83   2635    5508    1857    55
  36 SF-NNUE-experimental-20200722-0835.bin    :  3624.8    3.9    5349.5   10000    53      61   2591    5517    1892    55
  37 SF-NNUE-experimental-20200724-1215.bin    :  3623.9    4.8    5336.5   10000    53      94   2592    5489    1919    55
  38 SF-NNUE-20200721-1432.bin                 :  3619.2    4.2    5270.5   10000    53      52   2405    5731    1864    57
  39 SF-NNUE-20200721-1907.bin                 :  3619.0    4.9    5268.5   10000    53      55   2448    5641    1911    56
  40 SF-NNUE-20200721-0909.bin                 :  3618.6    4.7    5263.0   10000    53      99   2428    5670    1902    57
  41 SF-NNUE-20200720-1017.bin                 :  3610.9    4.5    5154.0   10000    52     100   2300    5708    1992    57
  42 Stockfish 220720 64 POPCNT                :  3600.0   ----  180835.0  400487    45     ---  70331  221008  109148    55

White advantage = 35.51 +/- 0.36
Draw rate (equal opponents) = 56.49 % +/- 0.08

10"+0.1"

Stockfish 220720 64 POPCNT = Stockfish 170720 64 POPCNT

User avatar
cdani
Posts: 2166
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Re: Sergio Vieri second net is out

Post by cdani » Wed Jul 29, 2020 12:38 pm

Another little test, 20+0.1

Code: Select all

   # PLAYER                 : RATING  ERROR   POINTS  PLAYED    (%)
   1 stnnue20200728-2138    : 2856.9    3.6    990.5    1971   50.3%
   2 stnnue20200729-1218    : 2855.1    3.6    980.5    1971   49.7%

User avatar
MikeB
Posts: 4201
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Sergio Vieri second net is out

Post by MikeB » Wed Jul 29, 2020 1:31 pm

MikeB wrote:
Wed Jul 29, 2020 6:24 am
Laskos wrote:
Wed Jul 29, 2020 4:56 am
cdani wrote:
Wed Jul 29, 2020 4:51 am
cdani wrote:
Tue Jul 28, 2020 10:06 pm
cdani wrote:
Tue Jul 28, 2020 8:30 pm
And I think that the latest net
20200729-0335.bin
is 25-30 elo stronger than 20200728-1442!!!!!!!! Testing now.
I talked too fast :oops: I'm testing 20200729-0109.bin and 20200729-0335.bin and don't seem to be better than 20200728-1442...
Finally these two where better but not much:

Code: Select all

20+0.1
   # PLAYER                 : RATING  ERROR   POINTS  PLAYED    (%)
   1 stnnue20200729-0109    : 2857.5    2.5   2101.5    4168   50.4%
   2 stnnue20200728-1442    : 2854.5    2.5   2066.5    4168   49.6%

20+0.1
   # PLAYER                 : RATING  ERROR   POINTS  PLAYED    (%)
   1 stnnue20200729-0335    : 2858.4    2.8   1868.5    3687   50.7%
   2 stnnue20200728-1442    : 2853.6    2.8   1818.5    3687   49.3%
Can confirm 0109 as best for me up to now, but only 1000 games. Haven't checked all of them, it seems the latest 0912 is not stronger.
20,000 game set kicked off...

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290204.pgn
Time Control: Time Control-> base+inc: 20+0.50
Games: 20000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 02:13:51

Projected-> Time: 9h:36m:20s
     Run -> Time: 0h:8m:53s

Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR 
---------------------------------------------------------------------------------------------------------

   1 Ho-2138   3514   0.0   46   46    96   50.5  52.6   25   20   51  26.0  53.1  3496 
   2 Ho-0109   3511   3.1   46   46    98   51.0  52.0   26   22   50  26.5  51.0  3497 
   3 Ho-0912   3506   5.8   46   46    98   50.0  51.0   24   22   52  24.5  53.1  3499 
   4 Ho-0335   3495  10.7   46   46    98   48.0  49.0   22   24   52  22.4  53.1  3501 
   5 Ho-0629   3474  21.4   47   47    98   44.5  45.4   22   31   45  22.4  45.9  3507 
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-2138     54 61 74 91
Ho-0109  45    57 71 89
Ho-0912  38 42    64 85
Ho-0335  25 28 35    76
Ho-0629   8 10 14 23   

Updated Results about every 3 minutes via watcher.sh
Results: https://www.dropbox.com/s/r5c6nwn5rgtmlnj/elo.txt?dl=0

PGN File: https://www.dropbox.com/s/bemhmhgrqrtrw ... 4.pgn?dl=0
Update - I am going to kill this run at 15,000 games - 0629 is the designated leader for me - ymmv . Will leave pgn on drop box for those that are interested - will start a new "shorter" run soon.

Final Results:

Code: Select all

#########################################################################################################
###                                              Summary                                              ###
#########################################################################################################

PGN File: c:/cluster.mfb/pgn/07290204.pgn
Time Control: Time Control-> base+inc: 20+0.50
Games: 20000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 09:30:45

Projected-> Time: 9h:36m:20s
     Run -> Time: 7h:25m:47s

15082 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-0629   3502   0.0    6    6  6033 3035.5  50.3 1472 1434 3127  24.4  51.8  3500
   2 Ho-0109   3501   1.0    6    6  6033 3025.0  50.1 1458 1441 3134  24.2  51.9  3500
   3 Ho-0335   3500   0.6    6    6  6032 3018.5  50.0 1427 1422 3183  23.7  52.8  3500
   4 Ho-0912   3500   0.0    6    6  6034 3019.0  50.0 1448 1444 3142  24.0  52.1  3500
   5 Ho-2138   3497   3.2    6    6  6032 2984.0  49.5 1416 1480 3136  23.5  52.0  3501
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-0629     60 66 66 89
Ho-0109  39    55 56 83
Ho-0335  33 44    50 80
Ho-0912  33 43 49    79
Ho-2138  10 16 19 20

#########################################################################################################
###                                                End                                                ###
#########################################################################################################
Image

User avatar
MikeB
Posts: 4201
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Sergio Vieri second net is out

Post by MikeB » Wed Jul 29, 2020 2:11 pm

started next round - this might be a good one - current leader 0629 is way behind all of the other new nets:
https://www.dropbox.com/s/r5c6nwn5rgtmlnj/elo.txt?dl=0

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 10:07:00

Projected-> Time: 4h:11m:0s
     Run -> Time: 0h:25m:56s

937 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-1743   3505   0.0   17   17   374  190.5  50.9   44   37  293  11.8  78.3  3499
   2 Ho-1218   3504   0.6   17   17   375  190.5  50.8   47   41  287  12.5  76.5  3499
   3 Ho-1910   3504   0.1   17   17   374  190.0  50.8   46   40  288  12.3  77.0  3499
   4 Ho-1500   3499   5.7   17   17   376  187.0  49.7   48   50  278  12.8  73.9  3500
   5 Ho-0629   3487  11.2   17   17   375  179.0  47.7   39   56  280  10.4  74.7  3503
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-1743     52 52 72 94
Ho-1218  47    50 70 94
Ho-1910  47 49    69 94
Ho-1500  27 29 30    84
Ho-0629   5  5  5 15

937 game(s) loaded

loops/scheduled: 8/100
using a more standard chess opening suite in this run, draws will run near 75-80%, so the wins will become more important
same dopdox link for results , if pgn is desired , please advise
Image

User avatar
MikeB
Posts: 4201
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Sergio Vieri second net is out

Post by MikeB » Wed Jul 29, 2020 2:42 pm

MikeB wrote:
Wed Jul 29, 2020 2:11 pm
started next round - this might be a good one - current leader 0629 is way behind all of the other new nets:
https://www.dropbox.com/s/r5c6nwn5rgtmlnj/elo.txt?dl=0

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 10:07:00

Projected-> Time: 4h:11m:0s
     Run -> Time: 0h:25m:56s

937 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-1743   3505   0.0   17   17   374  190.5  50.9   44   37  293  11.8  78.3  3499
   2 Ho-1218   3504   0.6   17   17   375  190.5  50.8   47   41  287  12.5  76.5  3499
   3 Ho-1910   3504   0.1   17   17   374  190.0  50.8   46   40  288  12.3  77.0  3499
   4 Ho-1500   3499   5.7   17   17   376  187.0  49.7   48   50  278  12.8  73.9  3500
   5 Ho-0629   3487  11.2   17   17   375  179.0  47.7   39   56  280  10.4  74.7  3503
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-1743     52 52 72 94
Ho-1218  47    50 70 94
Ho-1910  47 49    69 94
Ho-1500  27 29 30    84
Ho-0629   5  5  5 15

937 game(s) loaded

loops/scheduled: 8/100
using a more standard chess opening suite in this run, draws will run near 75-80%, so the wins will become more important
same dopdox link for results , if pgn is desired , please advise
0629 has made quite a comeback here , considering it's rough start

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 10:38:08

Projected-> Time: 4h:11m:0s
     Run -> Time: 0h:57m:4s

2192 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-1218   3503   0.0   10   10   876  442.5  50.5   98   89  689  11.2  78.7  3499
   2 Ho-0629   3500   2.5   10   10   876  438.5  50.1   90   89  697  10.3  79.6  3500
   3 Ho-1500   3500   0.3   10   10   878  439.0  50.0  101  101  676  11.5  77.0  3500
   4 Ho-1743   3499   0.9   10   10   876  436.5  49.8   87   90  699   9.9  79.8  3500
   5 Ho-1910   3498   1.4   10   10   878  435.5  49.6   93  100  685  10.6  78.0  3501
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-1218     64 66 71 77
Ho-0629  35    51 57 65
Ho-1500  33 48    55 63
Ho-1743  28 42 44    58
Ho-1910  22 34 36 41

2192 game(s) loaded

loops/scheduled: 20/100

waiting: 150
  ...seconds remaining:   100
Image

Zenmastur
Posts: 906
Joined: Sat May 31, 2014 6:28 am

Re: Sergio Vieri second net is out

Post by Zenmastur » Wed Jul 29, 2020 4:32 pm

Rebel wrote:
Wed Jul 29, 2020 7:43 am
Laskos wrote:
Wed Jul 29, 2020 4:56 am
Can confirm 0109 as best for me up to now, but only 1000 games. Haven't checked all of them, it seems the latest 0912 is not stronger.
I am running 0109 right now, increased the number of games to 5000. After 500 games a 64% score, that's +98 elo :D

We will see.
I tested 1218 with PO-AVX2 @ 20+0.2 5000 games against SF11-0717-modern with a 4-move book the results were:

Code: Select all

Games Completed = 5000 of 5000 (Avg game length = 67.616 sec)
Settings = RR/256MB/20000ms+200ms/M 1050cp for 2 moves, D 200 moves/PGN:(1)
Time = 21267 sec elapsed, 0 sec remaining
 1.  SFSF-NNUE-1218           	2962.5/5000	1216-291-3493  	
 2.  SF11-0717                	2037.5/5000	291-1216-3493  	
  
Only 2 defining forces have ever offered to die for you.....Jesus Christ and the American Soldier. One died for your soul, the other for your freedom.

User avatar
MikeB
Posts: 4201
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Sergio Vieri second net is out

Post by MikeB » Wed Jul 29, 2020 6:08 pm

MikeB wrote:
Wed Jul 29, 2020 2:42 pm
MikeB wrote:
Wed Jul 29, 2020 2:11 pm
started next round - this might be a good one - current leader 0629 is way behind all of the other new nets:
https://www.dropbox.com/s/r5c6nwn5rgtmlnj/elo.txt?dl=0

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 10:07:00

Projected-> Time: 4h:11m:0s
     Run -> Time: 0h:25m:56s

937 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-1743   3505   0.0   17   17   374  190.5  50.9   44   37  293  11.8  78.3  3499
   2 Ho-1218   3504   0.6   17   17   375  190.5  50.8   47   41  287  12.5  76.5  3499
   3 Ho-1910   3504   0.1   17   17   374  190.0  50.8   46   40  288  12.3  77.0  3499
   4 Ho-1500   3499   5.7   17   17   376  187.0  49.7   48   50  278  12.8  73.9  3500
   5 Ho-0629   3487  11.2   17   17   375  179.0  47.7   39   56  280  10.4  74.7  3503
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-1743     52 52 72 94
Ho-1218  47    50 70 94
Ho-1910  47 49    69 94
Ho-1500  27 29 30    84
Ho-0629   5  5  5 15

937 game(s) loaded

loops/scheduled: 8/100
using a more standard chess opening suite in this run, draws will run near 75-80%, so the wins will become more important
same dopdox link for results , if pgn is desired , please advise
0629 has made quite a comeback here , considering it's rough start

Code: Select all

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 10:38:08

Projected-> Time: 4h:11m:0s
     Run -> Time: 0h:57m:4s

2192 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-1218   3503   0.0   10   10   876  442.5  50.5   98   89  689  11.2  78.7  3499
   2 Ho-0629   3500   2.5   10   10   876  438.5  50.1   90   89  697  10.3  79.6  3500
   3 Ho-1500   3500   0.3   10   10   878  439.0  50.0  101  101  676  11.5  77.0  3500
   4 Ho-1743   3499   0.9   10   10   876  436.5  49.8   87   90  699   9.9  79.8  3500
   5 Ho-1910   3498   1.4   10   10   878  435.5  49.6   93  100  685  10.6  78.0  3501
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-1218     64 66 71 77
Ho-0629  35    51 57 65
Ho-1500  33 48    55 63
Ho-1743  28 42 44    58
Ho-1910  22 34 36 41

2192 game(s) loaded

loops/scheduled: 20/100

waiting: 150
  ...seconds remaining:   100
WOW 0629 came back and won

below is the entire output from cutechess-cli (client), running headless and in stealth mode...overshot the projected time by 84 seconds or 0.6%

Code: Select all

/c/Users/MichaelB7/home/Github/cutechess/projects/cli/run.sh: 
PGN File: c:/cluster.mfb/pgn/07290941.pgn
Games: 10000
Projected-> Time: 4h:11m:0s

Time Control-> base+inc: 20+0.40
Start running the chess match ...

#########################################################################################################
###                                              Summary                                              ###
#########################################################################################################

PGN File: c:/cluster.mfb/pgn/07290941.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 10000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 13:53:28

Projected-> Time: 4h:11m:0s
     Run -> Time: 4h:12m:24s

10000 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-0629   3505   0.0    5    5  4000 2035.0  50.9  444  374 3182  11.1  79.5  3499
   2 Ho-1743   3501   3.6    5    5  4000 2009.0  50.2  410  392 3198  10.2  80.0  3500
   3 Ho-1218   3501   0.7    5    5  4000 2003.5  50.1  422  415 3163  10.6  79.1  3500
   4 Ho-1500   3500   1.0    5    5  4000 1996.5  49.9  406  413 3181  10.2  79.5  3500
   5 Ho-1910   3494   5.6    5    5  4000 1956.0  48.9  394  482 3124   9.8  78.1  3502
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho Ho Ho
Ho-0629     87 91 95 99
Ho-1743  12    58 71 99
Ho-1218   8 41    63 98
Ho-1500   4 28 36    96
Ho-1910   0  0  1  3

#########################################################################################################
###                                                End                                                ###
#########################################################################################################
up next , a revisit with 2138. since it may have had an unlucky run with me, and the latest 2214, which is expected to be weaker.

Code: Select all

PGN File: c:/cluster.mfb/pgn/07291356.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 6000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 14:02:07

Projected-> Time: 2h:30m:44s
     Run -> Time: 0h:5m:13s

169 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-0629   3506   0.0   43   43   112   57.5  51.3   30   27   55  26.8  49.1  3497
   2 Ho-2138   3500   6.5   42   42   113   56.5  50.0   27   27   59  23.9  52.2  3500
   3 Ho-2214   3494   5.3   41   41   113   55.0  48.7   22   25   66  19.5  58.4  3503
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho
Ho-0629     60 68
Ho-2138  39    58
Ho-2214  31 41

169 game(s) loaded

loops/scheduled: 2/120
same link as before for results..
Image

User avatar
cdani
Posts: 2166
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Re: Sergio Vieri second net is out

Post by cdani » Wed Jul 29, 2020 6:44 pm

Another little test, 20+0.1

Code: Select all

   # PLAYER                 : RATING  ERROR   POINTS  PLAYED    (%)
   1 stnnue20200729-1743    : 2856.7    3.0   1266.5    2524   50.2%
   2 stnnue20200728-2138    : 2855.3    3.0   1257.5    2524   49.8%

Jouni
Posts: 2138
Joined: Wed Mar 08, 2006 7:15 pm

Re: Sergio Vieri second net is out

Post by Jouni » Wed Jul 29, 2020 8:03 pm

Sergio's own graphics says 2138 was the best. After that about 10 ELO regression. Only info is "d16" whatever it means? BTW my own test at bullet:

Code: Select all

                                          
1   Stockfish+NNUE 190720 64 BMI2  +115  +209/=242/-49 66.00%  330.0/500
2   Stockfish190720 64 BMI2       -115  +49/=242/-209 34.00%  170.0/500

Wow!
Jouni

User avatar
MikeB
Posts: 4201
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Sergio Vieri second net is out

Post by MikeB » Wed Jul 29, 2020 8:43 pm

2106 still on top, beating the net that Sergio called "best" 2138, quite convincingly with 95% los after 6000 games . 2214 was pretty close, surprisingly based on Sergio's own rating.

Code: Select all

$ run.sh
Settings ...

CMD1=Honey-XI-NN.exe ENG1=Ho-0629 EVFILE1=./eval/20200729-0629.bin
CMD2=Honey-XI-NN.exe ENG2=Ho-2138 EVFILE2=./eval/20200728-2138.bin
CMD3=Honey-XI-NN.exe ENG3=Ho-2214 EVFILE3=./eval/20200729-2214.bin

PGN File: c:/cluster.mfb/pgn/07291356.pgn
Games: 6000
Projected-> Time: 2h:30m:44s
20+0.40

Time Control-> base+inc: 20+0.40
Start running the chess match ...

#########################################################################################################
###                                              Summary                                              ###
#########################################################################################################

PGN File: c:/cluster.mfb/pgn/07291356.pgn
Time Control: Time Control-> base+inc: 20+0.40
Games: 6000
Threads: 1
Hash: 128

Current date : time (EDST)
Date: 07/29/20 : 16:35:27

Projected-> Time: 2h:30m:44s
     Run -> Time: 2h:38m:33s

6000 game(s) loaded
Rank Name     Rating   Δ     +    -     #     Σ    Σ%     W    L    D   W%    =%   OppR
---------------------------------------------------------------------------------------------------------

   1 Ho-0629   3503   0.0    7    7  4000 2022.0  50.5  982  938 2080  24.6  52.0  3499
   2 Ho-2214   3502   0.5    7    7  4000 2017.0  50.4  951  917 2132  23.8  53.3  3499
   3 Ho-2138   3495   6.6    7    7  4000 1961.0  49.0  916  994 2090  22.9  52.2  3502
---------------------------------------------------------------------------------------------------------

  Δ = delta from the next higher rated opponent
  # = number of games played
  Σ = total score, 1 point for win, 1/2 point for draw

LOS:
         Ho Ho Ho
Ho-0629     54 95
Ho-2214  45    93
Ho-2138   4  6

#########################################################################################################
###                                                End                                                ###
#########################################################################################################
next up, regression tests 0629 vs the very first sergio net 1017 and cur-dev-stockfish, using a more normal opening file ( more draws)
Image

Post Reply