Difference between revisions of "Building KataHex"

From HexWiki
Jump to: navigation, search
(created)
 
(creation)
Line 1: Line 1:
 
There is more than one version of KataHex out there, so these instructions have variants
 
There is more than one version of KataHex out there, so these instructions have variants
  
=== Ubuntu Linux ===
+
----
 +
== Ubuntu Linux ==
 
You may have to use <code>apt-get</code> or <code>synaptic</code> to install packages
 
You may have to use <code>apt-get</code> or <code>synaptic</code> to install packages
  
Line 10: Line 11:
 
<code>librust-tcmalloc-sys-dev</code>
 
<code>librust-tcmalloc-sys-dev</code>
  
The following BASH commands then may do the trick, but will produce code that may only work on the exact CPU model you use for the build.  See comments in the KataHex source code for details.
+
The following BASH commands then may do the trick, but will produce code that may only work on the exact CPU model you use for the build.  See comments in the KataHex source code, as well as the KataGo source code (on with KataHex is based, and which contains more details).
  
 
<pre>
 
<pre>
Line 19: Line 20:
 
==== CUDA ====
 
==== CUDA ====
 
I have not yet been able to build with the CUDA backend
 
I have not yet been able to build with the CUDA backend
 +
 +
==== TENSORRT ====
 +
I have not yet been able to build with the TENSORRT backend
 +
 +
==== OPENCL ====
 +
I have not yet been able to build with the OPENCL backend

Revision as of 19:40, 18 January 2026

There is more than one version of KataHex out there, so these instructions have variants


Ubuntu Linux

You may have to use apt-get or synaptic to install packages

EIGEN (CPU only)

You may have to install packages libtcmalloc-minimal4t64, librust-tcmalloc-dev, and librust-tcmalloc-sys-dev

The following BASH commands then may do the trick, but will produce code that may only work on the exact CPU model you use for the build. See comments in the KataHex source code, as well as the KataGo source code (on with KataHex is based, and which contains more details).

 cmake . -DUSE_BACKEND=EIGEN -DUSE_AVX2=1 -DUSE_TCMALLOC=1 -DCMAKE_CXX_FLAGS='-march=native'
 make -j 4    

CUDA

I have not yet been able to build with the CUDA backend

TENSORRT

I have not yet been able to build with the TENSORRT backend

OPENCL

I have not yet been able to build with the OPENCL backend