Adding some preliminary benchmark results to the front page. Encourag…

…ing, but some work to do on the monitoring (slow), which was manually disabled for those benchmarks.

README.rst

@@ -32,6 +32,23 @@ The datasets are randomized each time, but the output should be an image that lo
 .. image:: docs/plot_activation.png
 
 
+Benchmarks
+----------
+
+Here are the results of testing 10 epochs of training for two-thirds of the original MNIST data, on Ubuntu 14.04 and a GeForce GTX 650 (Memory: 1024Mb, Cores: 384).  You can run ``examples/bench_mnist.py`` to get the results.
+
+==========  ============  ===============  ===================
+   MNIST      sknn.mlp      nolearn.dbn      nolearn.lasagne
+==========  ============  ===============  ===================
+ Accuracy    **98.00%**       97.80%             97.75%
+ Training     **36s**          274s                68s
+==========  ============  ===============  ===================
+
+All the networks have 300 hidden units of the default type, and were given the same data with monitoring disabled. (For ``sknn`` the monitoring was commented out manually as of 2015/04/10.)  The remaining third of the MNIST dataset was only used to test the score once training terminated.
+
+**WARNING**: These results are not surprising, as ``pylearn2`` is developed by one of the best and most famous Deep Learning labs in the world.  However, they are not definitive and those numbers are very sensitive to parameter changes.
+
+
 Upcoming Features v0.1
 ----------------------