Adding documentation and RELEASE NOTES

RELEASE_NOTES.md

@@ -7,6 +7,7 @@
 - Scikit-learn support for Multinomial Logistic Regression
 - Support for min/max values other than defaults (i.e. 0.0 and 1.0) in MinMaxScalerModel
 - Support for custom transformers (StringMap, MathUnary) in Pyspark
+- Support for a performant implementation of the XGboost runtime (XGboost Predictor)
 
 ### Bug Fixes
 - Fix XGBoost sparse vector support

mleap-xgboost-runtime/README.md

@@ -0,0 +1,23 @@
+# MLeap XGBoost Runtime
+
+We provide two implementation of XGboost for use at runtime:
+- XGboost4j (from `ml.dmlc.xgboost4j`): this is the official C++ implementation.
+- XGBoost-Predictor (from `biz.k11i.xgboost-predictor`): this is a much faster implementation written directly in Java.
+
+By default, MLeap Bundles are de-serialized into XGboost4j Booster objects.
+In order to use the Predictor implementation, you may:
+
+1. Create a `resources/reference.conf` file in your project, like this:
+    ```
+    ml.combust.mleap.xgboost.ops = [
+      "ml.combust.mleap.xgboost.runtime.bundle.ops.XGBoostPredictorClassificationOp",
+      "ml.combust.mleap.xgboost.runtime.bundle.ops.XGBoostRegressionOp"
+    ]
+    ```
+2. add this to your project's pom file:
+    ```
+    <!-- Append our reference.conf into MLeap's reference.conf so our Ops are registered -->
+      <transformer implementation="org.apache.maven.plugins.shade.resource.AppendingTransformer">
+      <resource>reference.conf</resource>
+      </transformer>
+    ```