Here's a quick guide how to run this example. Note that the scripts here assume
that the executable is located in ".../mantaflow/build/", and most scripts were
primarily developed for Unix systems (Windows will work, but require path
modifications). Also several of the scripts use relative paths, so make sure
to run everything from the "example1_smoke_tiled" directory.

First generate data by calling
	>>> manta manta_genSimData.py
For windows, e.g.: ../../build/Release/manta.exe manta_genSimData.py
You can also use the tf_genManySims.py script to generate 10 data sets in one go.

Then use it to train a first model with tensorflow. This will take ca. 2 min.
E.g., with: 
	>>> python tf_train_pressure.py  out 0 fromSim 1000 toSim -1 trainingEpochs 10000
now you should have a trained model checkpoint for test_0001 in the data
directory (../data by default).

You can then use the model to generate an output with the command below. It
assumes the trained model was generated in ../data/test_0001 , and that the
sim data to apply it to the same sim data set:
	>>> python tf_train_pressure.py  out 1 fromSim 1000 toSim -1 loadModelTest 1 

For an automated test run, you can call runTests.sh, which runs through a
few different versions. By default it requires sim test data sets 3000-3002,
which you can download from the mantaflow web page.

