Awesome

BIDMach is a very fast machine learning library. Check the latest <a href="https://github.com/BIDData/BIDMach/wiki/Benchmarks">benchmarks</a>

The github distribution contains source code only. You also need a jdk 8, an installation of NVIDIA CUDA 8.0 (if you want to use a GPU) and CUDNN 5 if you plan to use deep networks. For building you need <a href="https://maven.apache.org/docs/history.html">maven 3.X</a>.

After doing <code>git clone</code>, cd to the BIDMach directory, and build and install the jars with <code>mvn install</code>. You can then run bidmach with ./bidmach. More details on installing and running are available <a href="https://github.com/BIDData/BIDMach/wiki/Installing-and-Running">here</a>.

The main project page is <a href="http://bid2.berkeley.edu/bid-data-project/">here</a>.

Documentation is <a href="https://github.com/BIDData/BIDMach/wiki">here in the wiki</a>

New BIDMach has a <a href="https://groups.google.com/forum/#!forum/bidmach-users-group">discussion group</a> on Google Groups.

BIDMach is a sister project of BIDMat, a matrix library, which is <a href="https://github.com/BIDData/BIDMat">also on github</a>

BIDData also has a project for deep reinforcement learning. <a href="https://github.com/BIDData/BIDMach_RL">BIDMach_RL</a> contains state-of-the-art implementations of several reinforcement learning algorithms.