Life In 19x19 http://www.lifein19x19.com/ |
|
KataGo Distributed Training and new networks http://www.lifein19x19.com/viewtopic.php?f=18&t=17944 |
Page 14 of 23 |
Author: | And [ Thu Jul 07, 2022 3:45 am ] | |||
Post subject: | Re: KataGo Distributed Training and new networks | |||
10 sec/move katago b40-s118 - b20-s530 15:5 (b40 won 7 Black, 8 White. KataGo - Eigen AVX2, numSearchThreads = 1) b20 won the shortest and longest game with white:
|
Author: | And [ Thu Jul 07, 2022 3:52 am ] | |||
Post subject: | Re: KataGo Distributed Training and new networks | |||
b40 also won the shortest and longest game with white:
|
Author: | And [ Fri Jul 22, 2022 6:22 am ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
KataGo b60c320-s632 maxPlayouts 1, numSearchThreads 1 - CS Zero 9d H5
|
Author: | And [ Wed Sep 28, 2022 2:12 am ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
KataGo b60c320-s647 maxPlayouts 1, numSearchThreads 1 - CS Zero 9d
|
Author: | And [ Thu Nov 03, 2022 11:58 am ] |
Post subject: | Re: KataGo Distributed Training and new networks |
new 18 block network https://cdn.discordapp.com/attachments/ ... r05.bin.gz does not work on KataGo 1.11.0: "This neural net requires a newer KataGo version." https://discord.com/channels/4170221623 ... 2637639690 |
Author: | And [ Mon Nov 07, 2022 8:50 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
New Go-playing trick defeats world-class Go AI—but loses to human amateurs Last week, a group of AI researchers published a paper outlining a method to defeat KataGo by using adversarial techniques that take advantage of KataGo's blind spots https://arstechnica.com/information-tec ... -amateurs/ |
Author: | Harleqin [ Thu Nov 10, 2022 1:48 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
Ugh. According to the article, this just exploits that they implemented the rules as »two passes end the game«, but KataGo by itself doesn't know this. I wonder whether the authors even understand Go. |
Author: | And [ Thu Nov 10, 2022 4:01 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
I think the end is the most interesting: "The research shows that AI systems that seem to perform at a human level are often doing so in a very alien way, and so can fail in ways that are surprising to humans," explains Gleave. "This result is entertaining in Go, but similar failures in safety-critical systems could be dangerous." Imagine a self-driving car AI that encounters a wildly unlikely scenario it doesn't expect, allowing a human to trick it into performing dangerous behaviors, for example. "[This research] underscores the need for better automated testing of AI systems to find worst-case failure modes," says Gleave, "not just test average-case performance." A half-decade after AI finally triumphed over the best human Go players, the ancient game continues its influential role in machine learning. Insights into the weaknesses of Go-playing AI, once broadly applied, may even end up saving lives. |
Author: | And [ Sat Nov 19, 2022 4:28 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
who compiled the version for Windows with support for the new 18b network, upload please! |
Author: | PC_Screen [ Mon Nov 21, 2022 4:22 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
Drop this katago.exe file in a working katago v1.11 OpenCL folder and have it replace the old katago.exe and it should work. The new 18b net is actually a bit faster and stronger than the old 40b net despite being "smaller" (it's a new architecture) and it's still mid-training so it should grow stronger over the following weeks/months. https://drive.google.com/file/d/1-uWOZS ... SDRQt/view |
Author: | And [ Wed Nov 23, 2022 10:00 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
@PC_Screen thank you! it works, but KataGo throws an error on the second move (in Sabaki and on the command line). probably not enough video memory. Code: Uncaught exception: CL_OUT_OF_RESOURCES <possibly ran out of GPU memory?> EDIT yes, there is not enough video memory - 7x7 works could you make an eigenavx2 version? |
Author: | hope366 [ Thu Nov 24, 2022 5:02 am ] |
Post subject: | Re: KataGo Distributed Training and new networks |
Thanks for uploading the engine files. I tried "katago.exe" which was put on Google Drive. I've tried with Lizzie and LizzieYzy, but I get an error and the engine doesn't work. The error message said something like "If you have ucrtbased.dll, it will work", so I got it from the internet and put it in the folder where katago.exe is located. I wasn't expecting much, but this worked. It's working with Lizzie and LizzieYzy, but it's noticeably slower compared to 40 blocks. It is about 60 to 70% of the speed of 40 blocks. I thought it was a problem related to FP16, so I tried changing the FP16 setting in the configuration file from "auto" to "true", but there was no change in speed. Any good ideas? |
Author: | And [ Thu Nov 24, 2022 9:34 am ] |
Post subject: | Re: KataGo Distributed Training and new networks |
I managed to run KataGo 18b on 19x19, and the speed has increased significantly! how I did it: 1 started tunning with a 10b network 2 renamed the resulting file to "tune8...x19_y19_c384_mv11.txt " and that's it! EDIT will it work for you? ... |
Author: | PC_Screen [ Thu Nov 24, 2022 11:50 am ] |
Post subject: | Re: KataGo Distributed Training and new networks |
Katago Eigen, same process, drop it inside the katago 1.11 eigen folder and replace the old katago.exe. https://drive.google.com/file/d/18XJHc8 ... sp=sharing As for the speed being lower, try running the benchmark in the command prompt (./katago.exe benchmark -model <NEURALNET>.bin.gz) and adjust the number of search threads based on it |
Author: | And [ Thu Nov 24, 2022 1:29 pm ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
KataGo (18b, playouts 1) - CS Zero 9d
|
Author: | hope366 [ Thu Nov 24, 2022 5:00 pm ] |
Post subject: | Re: KataGo Distributed Training and new networks |
>EDIT will it work for you? ... I tried that method, but unfortunately I didn't see any speed improvement. I also ran a benchmark test and it's slower than 40 blocks due to false FP16 relationships. CUDA version or TensorRT version will enable FP16 and speed will be faster. The katago.exe provided by @PC_Screen seems to be the OpenCL version. |
Author: | And [ Thu Nov 24, 2022 9:06 pm ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
second game KataGo (18b, playouts 1) - CS Zero 9d
|
Author: | And [ Fri Nov 25, 2022 8:44 am ] |
Post subject: | Re: KataGo Distributed Training and new networks |
Similarly replaced the tuning file for the 40b network, the time to move 40b is 20% longer than that of 18b (playouts 1, gt 610) |
Author: | And [ Fri Nov 25, 2022 9:25 am ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
KataGo 18b, playouts 10 - 40b, playouts 10
|
Author: | And [ Fri Nov 25, 2022 10:07 am ] | ||
Post subject: | Re: KataGo Distributed Training and new networks | ||
second game KataGo 18b, playouts 10 - 40b, playouts 10 (40b-s118)
|
Page 14 of 23 | All times are UTC - 8 hours [ DST ] |
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |