Discussion of anything and everything relating to chess playing software and machines.
Moderators: hgm , Rebel , chrisw
cdani
Posts: 2204 Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra
Post
by cdani » Mon Jul 27, 2020 8:53 am
cdani wrote: ↑ Sun Jul 26, 2020 10:08 pm
cdani wrote: ↑ Sun Jul 26, 2020 8:32 pm
2051 seems the best:
Someone found at Discord that experimental-20200722-1621.bin is even better. Testing...
2051 prevails. 20+0.1
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200725-2051 : 2857.3 1.8 4659.5 9249 50.4%
2 stnnuesx20200722-1621 : 2854.7 1.8 4589.5 9249 49.6%
Laskos
Posts: 10948 Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos
Post
by Laskos » Mon Jul 27, 2020 11:19 am
cdani wrote: ↑ Mon Jul 27, 2020 8:53 am
cdani wrote: ↑ Sun Jul 26, 2020 10:08 pm
cdani wrote: ↑ Sun Jul 26, 2020 8:32 pm
2051 seems the best:
Someone found at Discord that experimental-20200722-1621.bin is even better. Testing...
2051 prevails. 20+0.1
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200725-2051 : 2857.3 1.8 4659.5 9249 50.4%
2 stnnuesx20200722-1621 : 2854.7 1.8 4589.5 9249 49.6%
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
cdani
Posts: 2204 Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra
Post
by cdani » Tue Jul 28, 2020 12:20 am
Laskos wrote: ↑ Mon Jul 27, 2020 11:19 am
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
Yes. It seems equal to 2051. 20+0.1:
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200727-0928 : 2856.4 1.8 4629.5 9239 50.1%
2 stnnuesergio20200725-2051 : 2855.6 1.8 4609.5 9239 49.9%
Laskos
Posts: 10948 Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos
Post
by Laskos » Tue Jul 28, 2020 12:26 am
cdani wrote: ↑ Tue Jul 28, 2020 12:20 am
Laskos wrote: ↑ Mon Jul 27, 2020 11:19 am
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
Yes. It seems equal to 2051. 20+0.1:
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200727-0928 : 2856.4 1.8 4629.5 9239 50.1%
2 stnnuesergio20200725-2051 : 2855.6 1.8 4609.5 9239 49.9%
In 1000 games matches, today there were 3 excellent nets:
27.07_0928
27.07_1540
28.07_0207
Especially the last one seems superior almost outside error margins even in only 1000 games. Still maybe a lucky run.
MikeB
Posts: 4889 Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania
Post
by MikeB » Tue Jul 28, 2020 5:04 am
Laskos wrote: ↑ Tue Jul 28, 2020 12:26 am
cdani wrote: ↑ Tue Jul 28, 2020 12:20 am
Laskos wrote: ↑ Mon Jul 27, 2020 11:19 am
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
Yes. It seems equal to 2051. 20+0.1:
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200727-0928 : 2856.4 1.8 4629.5 9239 50.1%
2 stnnuesergio20200725-2051 : 2855.6 1.8 4609.5 9239 49.9%
In 1000 games matches, today there were 3 excellent nets:
27.07_0928
27.07_1540
28.07_0207
Especially the last one seems superior almost outside error margins even in only 1000 games. Still maybe a lucky run.
Ditto here with respect to 0207 - new leader for me.
Code: Select all
#########################################################################################################
### Summary ###
#########################################################################################################
PGN File: c:/cluster.mfb/pgn/07272100.pgn
Time Control: base+inc: 10+0.33
Games: 6000
Threads:
Hash: 128
Current date : time (EDST)
Date: 07/27/20 : 22:43:32
Projected-> Time: 1h:44m:0s
Run -> Time: 1h:43m:3s
6000 game(s) loaded
Rank Name Rating Δ + - # Σ Σ% W L D W% =% OppR
---------------------------------------------------------------------------------------------------------
1 Ho-0207 3511 0.0 6 6 3000 1565.0 52.2 438 308 2254 14.6 75.1 3496
2 Ho-1540 3504 7.6 6 6 3000 1521.0 50.7 397 355 2248 13.2 74.9 3499
3 Ho-0928 3496 7.1 6 6 3000 1479.0 49.3 351 393 2256 11.7 75.2 3501
4 Ho-1135 3489 7.7 6 6 3000 1435.0 47.8 302 432 2266 10.1 75.5 3504
---------------------------------------------------------------------------------------------------------
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-0207 97 99 99
Ho-1540 2 96 99
Ho-0928 0 3 97
Ho-1135 0 0 2
#########################################################################################################
### End ###
#########################################################################################################
Laskos
Posts: 10948 Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos
Post
by Laskos » Tue Jul 28, 2020 8:35 am
MikeB wrote: ↑ Tue Jul 28, 2020 5:04 am
Laskos wrote: ↑ Tue Jul 28, 2020 12:26 am
cdani wrote: ↑ Tue Jul 28, 2020 12:20 am
Laskos wrote: ↑ Mon Jul 27, 2020 11:19 am
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
Yes. It seems equal to 2051. 20+0.1:
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200727-0928 : 2856.4 1.8 4629.5 9239 50.1%
2 stnnuesergio20200725-2051 : 2855.6 1.8 4609.5 9239 49.9%
In 1000 games matches, today there were 3 excellent nets:
27.07_0928
27.07_1540
28.07_0207
Especially the last one seems superior almost outside error margins even in only 1000 games. Still maybe a lucky run.
Ditto here with respect to 0207 - new leader for me.
Code: Select all
#########################################################################################################
### Summary ###
#########################################################################################################
PGN File: c:/cluster.mfb/pgn/07272100.pgn
Time Control: base+inc: 10+0.33
Games: 6000
Threads:
Hash: 128
Current date : time (EDST)
Date: 07/27/20 : 22:43:32
Projected-> Time: 1h:44m:0s
Run -> Time: 1h:43m:3s
6000 game(s) loaded
Rank Name Rating Δ + - # Σ Σ% W L D W% =% OppR
---------------------------------------------------------------------------------------------------------
1 Ho-0207 3511 0.0 6 6 3000 1565.0 52.2 438 308 2254 14.6 75.1 3496
2 Ho-1540 3504 7.6 6 6 3000 1521.0 50.7 397 355 2248 13.2 74.9 3499
3 Ho-0928 3496 7.1 6 6 3000 1479.0 49.3 351 393 2256 11.7 75.2 3501
4 Ho-1135 3489 7.7 6 6 3000 1435.0 47.8 302 432 2266 10.1 75.5 3504
---------------------------------------------------------------------------------------------------------
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-0207 97 99 99
Ho-1540 2 96 99
Ho-0928 0 3 97
Ho-1135 0 0 2
#########################################################################################################
### End ###
#########################################################################################################
Now the latest one 20200728-0633.bin seems the strongest, almost outside error margins again in 1000 games. Now there are substantial Elo gains versus old 2141, 10-12 Elo points or so.
cdani
Posts: 2204 Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra
Post
by cdani » Tue Jul 28, 2020 9:48 am
Laskos wrote: ↑ Tue Jul 28, 2020 12:26 am
cdani wrote: ↑ Tue Jul 28, 2020 12:20 am
Laskos wrote: ↑ Mon Jul 27, 2020 11:19 am
The new one 20200727-0928.bin seems on a par with the best in 1000 games.
Yes. It seems equal to 2051. 20+0.1:
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnuesergio20200727-0928 : 2856.4 1.8 4629.5 9239 50.1%
2 stnnuesergio20200725-2051 : 2855.6 1.8 4609.5 9239 49.9%
In 1000 games matches, today there were 3 excellent nets:
27.07_0928
27.07_1540
28.07_0207
Especially the last one seems superior almost outside error margins even in only 1000 games. Still maybe a lucky run.
For me too:
20+0.1
Code: Select all
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnue20200728-0207 : 2863.1 3.4 1141.0 2195 52.0%
2 stnnue20200725-2051 : 2848.9 3.4 1054.0 2195 48.0%
cdani
Posts: 2204 Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra
Post
by cdani » Tue Jul 28, 2020 6:22 pm
Laskos wrote: ↑ Tue Jul 28, 2020 8:35 am
Now the latest one 20200728-0633.bin seems the strongest, almost outside error margins again in 1000 games. Now there are substantial Elo gains versus old 2141, 10-12 Elo points or so.
Fist I tested a bit the two new nets:
Code: Select all
20+0.1
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnue20200728-1442 : 2857.1 2.8 1916.5 3810 50.3%
2 stnnue20200728-0633 : 2854.9 2.8 1893.5 3810 49.7%
And then I took the possible best against previous best:
Code: Select all
20+0.1
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnue20200728-1442 : 2860.3 3.2 1552.0 3030 51.2%
2 stnnue20200728-0207 : 2851.7 3.2 1478.0 3030 48.8%
So we have a new best, 20200728-1442, and with big margin!!
cdani
Posts: 2204 Joined: Sat Jan 18, 2014 10:24 am
Location: Andorra
Post
by cdani » Tue Jul 28, 2020 6:27 pm
Now in the download site there is a graphic of rating (fixed d12 games).
https://www.comp.nus.edu.sg/~sergio-v/nnue/
Laskos
Posts: 10948 Joined: Wed Jul 26, 2006 10:21 pm
Full name: Kai Laskos
Post
by Laskos » Tue Jul 28, 2020 7:01 pm
cdani wrote: ↑ Tue Jul 28, 2020 6:22 pm
Laskos wrote: ↑ Tue Jul 28, 2020 8:35 am
Now the latest one 20200728-0633.bin seems the strongest, almost outside error margins again in 1000 games. Now there are substantial Elo gains versus old 2141, 10-12 Elo points or so.
Fist I tested a bit the two new nets:
Code: Select all
20+0.1
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnue20200728-1442 : 2857.1 2.8 1916.5 3810 50.3%
2 stnnue20200728-0633 : 2854.9 2.8 1893.5 3810 49.7%
And then I took the possible best against previous best:
Code: Select all
20+0.1
# PLAYER : RATING ERROR POINTS PLAYED (%)
1 stnnue20200728-1442 : 2860.3 3.2 1552.0 3030 51.2%
2 stnnue20200728-0207 : 2851.7 3.2 1478.0 3030 48.8%
So we have a new best, 20200728-1442, and with big margin!!
Wow, hard to keep pace with the improvements. Suddenly the rating increased by some 20 Elo points since the old best 2141 (which was already very strong).