Merge branch 'master' of https://github.com/Ricardicus/recurrent-neural-net

Author: Ricardicus

README.md

@@ -6,14 +6,12 @@ This is necessary in order to model long-term dependencies such as can be found
 This program will learn to produce text similar to the one that
 it has been training on using a LSTM network implemented in C. The repo is inspired by Andrej Karpathys <i>char-rnn</i>: https://github.com/karpathy/char-rnn but instead implemented in C to be used in more constrained environments.
 
-# How do I use it? 
+# Build 
 
 Do you have a mac or a Linux machine? 
 In that case it is super easy to download and run.
 
-However, I don't really know how to compile this program in Windows since I am not using Windows myself. I have used a program called MinGW to compile my own C programs in the past. But maybe check this out if you are on Windows: https://sourceforge.net/projects/mingw/files/
-
-Otherwise:
+## Mac or Linux (UNIX types)
 
 Download the program. 
 Open a terminal window and type:
@@ -36,6 +34,22 @@ Then run the program:
 where datafile is a file with the traning data and it will start training on it. You can see the progress 
 over time. 
 
+## Windows
+
+I have only built it using Visual Studio Native Tools Command Prompt for VS2017. 
+
+Install CMake and Nmake (and git). 
+Then do:
+
+<pre>
+git clone https://github.com/Ricardicus/recurrent-neural-net/
+cd recurrent-neural-net
+cmake -G"NMake Makefiles"
+nmake
+</pre>
+
+# Configure
+
 Check out the file "std_conf.h".
 
 In std_conf.h you can edit the program. You can edit the hyperparameters such as learning rate etc, set the number of layers (2/3 is best I think), set how often it should output data etc. If you edit this file, you edit the source code and you will need to rebuild the program with the command "make". You can also use input arguments to set some of the behaviour.