对官方引擎胜率60%的roy7引擎

lu01

https://github.com/leela-zero/le ... uecomment-473688679
结果是：
leelaz-max_lcb_root v leelaz-next (400/400 games)
board size: 19   komi: 7.5
                      wins              black          white        avg cpul
eelaz-max_lcb_root    247 61.75%       118 59.00%     129 64.50%    425.93
leelaz-next            153 38.25%       71  35.50%     82  41.00%    427.05
                                        189 47.25%     211 52.75%

作者要求不要用于autogtp

taishuan

這麼猛，那官方引擎0.17是不是要慢點出啊?太快放出來萬一被吊打不是很糗?

yunhuo

下载两个exe就行了，没别的文件？

lu01

放在原始文件的目录下

飘弈女孩

网页打不开啊

yunhuo

lu01 发表于 2019-3-18 12:13
放在原始文件的目录下

原始文件？就是leela官方文件目录里面吧

landa

啥配套文件也不用，那些文件是给aotugtp训练用的。

jiangbangjian

有哪们朋友试用过吗?是不是64位系统不能用啊?

Here is the Win32 binary:

leela-zero-roy7.zip

晕，换了权重能用了，还挑权重的,15b_204_352k_q怎么用不了

lu01

64操作系统能用32应用程序

hred9D

运行的参数需要改变吗？

taishuan

jiangbangjian 发表于 2019-3-18 19:24
有哪们朋友试用过吗?是不是64位系统不能用啊?Here is the Win32 binary:leela-zero-roy7.zip ...

真的挑權重，40b練成的15b權重不能用，elf v2可以用。

lu01

f438没问题

D:\tool\leela15>leelaz-max_lcb_root -w \f438.txt
Using OpenCL batch size of 5
Using 10 thread(s).
RNG seed: 10256206109071070972
Leela Zero 0.16  Copyright (C) 2017-2019  Gian-Carlo Pascutto and contributors
This program comes with ABSOLUTELY NO WARRANTY.
This is free software, and you are welcome to redistribute it
under certain conditions; see the COPYING file for details.

BLAS Core: built-in Eigen 3.3.5 library.
Detecting residual layers...v1...192 channels...15 blocks.
Initializing OpenCL (autodetecting precision).
Detected 1 OpenCL platforms.
Platform version: OpenCL 1.2 CUDA 9.1.84
Platform profile: FULL_PROFILE
Platform name:    NVIDIA CUDA
Platform vendor:  NVIDIA Corporation
Device ID:     0
Device name:   GeForce GT 730
Device type:   GPU
Device vendor: NVIDIA Corporation
Device driver: 391.35
Device speed:  901 MHz
Device cores:  2 CU
Device score:  1112
Selected platform: NVIDIA CUDA
Selected device: GeForce GT 730
with OpenCL 1.2 capability.
Half precision compute support: No.
Tensor Core support: No.
Detected 1 OpenCL platforms.
Platform version: OpenCL 1.2 CUDA 9.1.84
Platform profile: FULL_PROFILE
Platform name:    NVIDIA CUDA
Platform vendor:  NVIDIA Corporation
Device ID:     0
Device name:   GeForce GT 730
Device type:   GPU
Device vendor: NVIDIA Corporation
Device driver: 391.35
Device speed:  901 MHz
Device cores:  2 CU
Device score:  1112
Selected platform: NVIDIA CUDA
Selected device: GeForce GT 730
with OpenCL 1.2 capability.
Half precision compute support: No.
Tensor Core support: No.

Started OpenCL SGEMM tuner.
Will try 290 valid configurations.
(1/290) KWG=16 KWI=8 MDIMA=8 MDIMC=8 MWG=64 NDIMB=8 NDIMC=8 NWG=32 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=2 VWN=2 2.0760 ms (
159.8 GFLOPS)
(2/290) KWG=16 KWI=2 MDIMA=16 MDIMC=16 MWG=64 NDIMB=8 NDIMC=8 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=2 1.4157 ms
(234.4 GFLOPS)
(9/290) KWG=16 KWI=2 MDIMA=8 MDIMC=8 MWG=64 NDIMB=16 NDIMC=16 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=4 1.4073 ms
(235.8 GFLOPS)
(67/290) KWG=16 KWI=2 MDIMA=8 MDIMC=8 MWG=32 NDIMB=16 NDIMC=16 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=4 1.3594 m
s (244.1 GFLOPS)
Wavefront/Warp size: 32
Max workgroup size: 1024
Max workgroup dimensions: 1024 1024 64

Started OpenCL SGEMM tuner.
Will try 290 valid configurations.
(1/290) KWG=16 KWI=8 MDIMA=8 MDIMC=8 MWG=64 NDIMB=8 NDIMC=8 NWG=32 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=2 VWN=2 2.0007 ms (
165.8 GFLOPS)
(2/290) KWG=16 KWI=2 MDIMA=16 MDIMC=16 MWG=64 NDIMB=8 NDIMC=8 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=2 1.5245 ms
(217.6 GFLOPS)
(69/290) KWG=16 KWI=2 MDIMA=16 MDIMC=16 MWG=64 NDIMB=8 NDIMC=8 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=4 1.4920 m
s (222.4 GFLOPS)
(103/290) KWG=16 KWI=8 MDIMA=16 MDIMC=16 MWG=64 NDIMB=8 NDIMC=8 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=4 1.4887
ms (222.9 GFLOPS)
(237/290) KWG=16 KWI=8 MDIMA=8 MDIMC=8 MWG=64 NDIMB=8 NDIMC=8 NWG=64 SA=1 SB=1 STRM=0 STRN=0 TCE=0 VWM=4 VWN=2 1.4603 ms
(227.2 GFLOPS)
Wavefront/Warp size: 32
Max workgroup size: 1024
Max workgroup dimensions: 1024 1024 64
Using OpenCL single precision (less than 5% slower than half).
Detected 1 OpenCL platforms.
Platform version: OpenCL 1.2 CUDA 9.1.84
Platform profile: FULL_PROFILE
Platform name:    NVIDIA CUDA
Platform vendor:  NVIDIA Corporation
Device ID:     0
Device name:   GeForce GT 730
Device type:   GPU
Device vendor: NVIDIA Corporation
Device driver: 391.35
Device speed:  901 MHz
Device cores:  2 CU
Device score:  1112
Selected platform: NVIDIA CUDA
Selected device: GeForce GT 730
with OpenCL 1.2 capability.
Half precision compute support: No.
Tensor Core support: No.
Loaded existing SGEMM tuning.
Wavefront/Warp size: 32
Max workgroup size: 1024
Max workgroup dimensions: 1024 1024 64
Setting max tree size to 4248 MiB and cache size to 472 MiB.

Passes: 0            Black (X) Prisoners: 0
Black (X) to move    White (O) Prisoners: 0

   a b c d e f g h j k l m n o p q r s t
19 . . . . . . . . . . . . . . . . . . . 19
18 . . . . . . . . . . . . . . . . . . . 18
17 . . . . . . . . . . . . . . . . . . . 17
16 . . . + . . . . . + . . . . . + . . . 16
15 . . . . . . . . . . . . . . . . . . . 15
14 . . . . . . . . . . . . . . . . . . . 14
13 . . . . . . . . . . . . . . . . . . . 13
12 . . . . . . . . . . . . . . . . . . . 12
11 . . . . . . . . . . . . . . . . . . . 11
10 . . . + . . . . . + . . . . . + . . . 10
9 . . . . . . . . . . . . . . . . . . .  9
8 . . . . . . . . . . . . . . . . . . .  8
7 . . . . . . . . . . . . . . . . . . .  7
6 . . . . . . . . . . . . . . . . . . .  6
5 . . . . . . . . . . . . . . . . . . .  5
4 . . . + . . . . . + . . . . . + . . .  4
3 . . . . . . . . . . . . . . . . . . .  3
2 . . . . . . . . . . . . . . . . . . .  2
1 . . . . . . . . . . . . . . . . . . .  1
   a b c d e f g h j k l m n o p q r s t

Hash: 9A930BE1616C538E Ko-Hash: A14C933E7669946D

Black time: 01:00:00

lu01

f438 的地址 https://leela.online-go.com/netw ... 671954a8df3cf707.gz

gfchy

64位的用leelaz-max_lcb_root.exe就可以

gfchy

64位的用leelaz-max_lcb_root.exe就可以