Meta Configuration of Package DeepLearningToga-1.5.21a

<package name="DeepLearningToga-1.5.21a" project="Games">
  <title>A very strong, UCI compatible Chess engine ("learning" multiprocessor version)</title>
  <description>TogaII is an improved version of the well known UCI chess engine fruit and was based on fruit's last OSS version 2.1.
As togaII uses UCI (universal chess interface, UCI) you will need a uci-compliant user interface to play against toga2.
You can also use polyglot as an "UCI to xboard"-adapter to use togaII with xboard-compatible programs.
TogaII's ELO-rating is about 2800 points.

This version is derived from 1.4.1 and aimed at multi processor machines and has a learning mode implemented.

By default the number of cpu's is detected automatically, so you can also use this on a single core machine.
The ini-file for polyglot is set to 2 cores by default, which is safe as it will be ignored if you are running
DeepLearningToga via polyglot (i.e. with xboard) on a single core machine.
If you have a machine with more than two cores and want to use all of them, just set this

Threads = 2

to the amount of cores you have.

In scid use "Configure UCI engine" option when configuring the engine (Tools =&gt; Analysis Engine =&gt; New/Edit) and set 
"Number of Threads" accordingly.

Learning mode is disabled by default in the ini-File for polyglot:

Use Learn = false
Generate Learn = true
Information File = toga_info.finf

If "Use Learn" is set to "true", learning will be enabled. Beware, that this will create a growing file 
"/home/Your_Username/.DeepLearningToga-1.5.21a/toga_info.finf", check the size of that file regularly!

The startup scripts "xDeepLearningToga-1.5.21a" and "DeepLearningToga-1.5.21a" will create the file "toga_info.finf" in $HOME/.DeepLearningToga-1.5.21a/ automatically.

For scid set the appropriate parameters with the "Configure UCI engine" option when configuring the engine 
(Tools =&gt; Analysis Engine =&gt; New/Edit) and use _absolute_ paths (/home/Your_Username/.DeepLearningToga-1.5.21a/)

It is recommended to use the same path as for xboard/polyglot to combine the learning results.

See "/usr/share/doc/packages/DeepLearningToga-1.5.21a/readme.txt" for more details.

Authors:
--------
		Fabien Letouzey (Fruit 2.1), Thomas Gaksch (Toga II), Daniel Shawul (egbb code), Chris Formula (SE enhancements), Jerry Donald (JD enhancements).
		William H. Mowery Jr. &lt;WHMoweryJr@yahoo.com&gt; (this version)
		Oliver Korff &lt;ok@xynyx.de&gt; (manual page)
		Marc Lacrosse (opening book "Performance.bin")

</description>
  <url>http://alpha.uhasselt.be/Research/Algebra/Toga/linux_versions/</url>
</package>