CuriosAI sai issues - Githubissues

CuriosAI / sai

SAI: a fork of Leela Zero with variable komi.

GNU General Public License v3.0

103 stars 11 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

The 9x9 weighs was missed.

#151 CGLemon opened 9 months ago
2
Add support for scoreLead label.

#150 CGLemon closed 9 months ago
3
Hope to support ONNX Runtime (Training version & Inferencing version) and DirectML

#149 Looong01 opened 1 year ago
1
Just suggestion: go to 40b or close...

#148 y-ich opened 1 year ago
6
autogtp build error

#147 broody closed 2 years ago
2
suggest use katago games to increase self play games

#146 l1t1 opened 3 years ago
2
Is it possible to do 64-bit Linux binary CPU only release?

#145 GD-fix opened 3 years ago
4
Doesn't compile CPU only version

#144 GD-fix closed 3 years ago
1
Fix the shared mutex stall (Need more tests).

#143 CGLemon closed 1 year ago
2
sai 0.18.1 vs 0.17.6 (my impression)

#142 Cabu closed 3 years ago
16
sai 0.18.1 run with problem on window7 32bits

#141 l1t1 closed 3 years ago
20
What effect are the Lambda and Mu?

#140 CGLemon opened 3 years ago
2
Fix prob cache bug

#139 CGLemon closed 3 years ago
2
Implement shared mutex

#138 CGLemon closed 3 years ago
12
network: 61be4874d51e397e8e58f raise *ERROR*: Could not talk to engine after launching.

#137 l1t1 closed 3 years ago
3
all the elo of nets drop to 500

#136 l1t1 closed 3 years ago
2
katago training so fast

#135 l1t1 closed 3 years ago
4
Latest news from telegram channel

#134 Vandertic opened 3 years ago
16
to get the similar elo, the training games is about 1:2 of leelazero now

#133 l1t1 opened 3 years ago
0
removing the old networks

#132 maxinjapan opened 3 years ago
6
what is the meaning of the experiment type of matchs

#131 l1t1 closed 3 years ago
5
sai 510 p1600 has 1/10 chance to beat lz elf v2 p800

#130 l1t1 opened 3 years ago
1
consider to switch 20x256 net at number 512?

#129 l1t1 opened 3 years ago
6
self-play games lost comments when I interrupt autogtp with the q command

#128 nanzi closed 3 years ago
1
will sai use quantize on 20x256 nets?

#127 l1t1 closed 3 years ago
2
[Question] License of trained weights

#126 y-ich closed 3 years ago
6
sai self-play games of each day has exceeded leelazero

#125 l1t1 opened 3 years ago
0
suggest begin to train the 20*192 size net

#124 l1t1 opened 3 years ago
5
SAI periodically disconnects in sabaki

#123 cryptsport opened 3 years ago
27
force update

#122 kennyfs opened 3 years ago
2
does the recent 15*192 nets catch up with latest 12*256 nets

#121 l1t1 opened 3 years ago
2
net 357 has more than 6000 self-play games now

#120 l1t1 opened 4 years ago
2
speed up train paper

#119 l1t1 closed 4 years ago
1
fix typo, link to wiki

#118 wonderingabout closed 4 years ago
0
rewrite SAI documentation

#117 wonderingabout closed 4 years ago
2
a way to find good net size

#116 l1t1 closed 4 years ago
2
Server offline

#115 Vandertic closed 4 years ago
0
how can i use net_to_model.py？

#114 tang3 opened 4 years ago
1
try training with python parse.py 16 128 /tmp/test.txt.0 found error

#113 tang3 closed 4 years ago
7
Progress stalled? Currently checking

#112 Vandertic opened 4 years ago
6
Temporary problem with training

#111 Vandertic closed 4 years ago
0
Is g130 network files broken?

#110 nanzi closed 4 years ago
4
is net 298 training more self paly games

#109 l1t1 closed 4 years ago
1
the game amount of promote matches of this round and some last round is only 20+

#108 l1t1 closed 4 years ago
1
the game amount of promote match of this round is only 20+

#107 l1t1 closed 4 years ago
1
question about the new training process

#106 l1t1 opened 4 years ago
0
Make lz-analyze output every move's areas

#105 yzyray closed 4 years ago
2
will you continue to train 9x192 nets after moving to 12x256

#104 l1t1 opened 4 years ago
2
new net no.260 promoted too early

#103 l1t1 opened 4 years ago
5
http://sai.unich.it/submit-match 404 error

#102 Cabu opened 4 years ago
12