diff --git a/.gitignore b/.gitignore
index 7642f5086..8df040bc4 100644
--- a/.gitignore
+++ b/.gitignore
@@ -44,3 +44,5 @@ examples/bigdata
 # generated Dockerfile
 docker/sparkling-water/base/Dockerfile
 
+# data generated by one of the tests
+spark-tfidf-data/
diff --git a/build.gradle b/build.gradle
index ace79c443..99eeb3012 100644
--- a/build.gradle
+++ b/build.gradle
@@ -89,7 +89,7 @@ configure(subprojects) { project ->
     repositories {
         // Should be enabled only in development mode
         if (h2oBuild == '99999') mavenLocal()
-        
+
         mavenCentral()
 
         maven {
diff --git a/core/build.gradle b/core/build.gradle
index 9fa0d3ff6..9e5289efb 100644
--- a/core/build.gradle
+++ b/core/build.gradle
@@ -26,7 +26,6 @@ dependencies {
   }
 
   // Spark components
-
   // - core
   compile "org.apache.spark:spark-core_${scalaBaseVersion}:${sparkVersion}"
   // - SQL component
@@ -98,5 +97,4 @@ integTest {
     // Pass list of jars required for testing
     systemProperty "sparkling.assembly.jar", "${project(":sparkling-water-assembly").configurations.shadow.artifacts.file.join(',')}"
     systemProperty "sparkling.itest.jar", "${integTestJar.archivePath}"
-}
-
+}
\ No newline at end of file
diff --git a/core/src/main/scala/org/apache/spark/h2o/H2OContext.scala b/core/src/main/scala/org/apache/spark/h2o/H2OContext.scala
index ded46a59e..ae6064116 100644
--- a/core/src/main/scala/org/apache/spark/h2o/H2OContext.scala
+++ b/core/src/main/scala/org/apache/spark/h2o/H2OContext.scala
@@ -58,9 +58,7 @@ import scala.util.control.NoStackTrace
   * @param conf H2O configuration
   */
 class H2OContext private (@transient val sparkContext: SparkContext, @transient conf: H2OConf) extends org.apache.spark.Logging
-  with Serializable with SparkDataFrameConverter with SupportedRDDConverter with H2OContextUtils{
-  self =>
-
+  with Serializable with SparkDataFrameConverter with SupportedRDDConverter with H2OContextUtils { self =>
 
   /** IP of H2O client */
   private var localClientIp: String = _
@@ -85,7 +83,7 @@ class H2OContext private (@transient val sparkContext: SparkContext, @transient
     * otherwise it creates new H2O cluster living in Spark
     */
   def init(): H2OContext = {
-    if(!isRunningOnCorrectSpark(sparkContext)){
+    if (!isRunningOnCorrectSpark(sparkContext)) {
       throw new WrongSparkVersion(s"You are trying to use Sparkling Water built for Spark $buildSparkMajorVersion," +
         s" but your $$SPARK_HOME(=${sparkContext.getSparkHome().getOrElse("SPARK_HOME is not defined!")}) property" +
         s" points to Spark of version ${sparkContext.version}. Please ensure correct Spark is provided and" +
@@ -125,13 +123,13 @@ class H2OContext private (@transient val sparkContext: SparkContext, @transient
 
   /** Transform DataFrame to H2OFrame */
   def asH2OFrame(df : DataFrame): H2OFrame = asH2OFrame(df, None)
-  def asH2OFrame(df : DataFrame, frameName: Option[String]) : H2OFrame = toH2OFrame(this, df, frameName)
-  def asH2OFrame(df : DataFrame, frameName: String) : H2OFrame = asH2OFrame(df, Option(frameName))
+  def asH2OFrame(df : DataFrame, frameName: Option[String]): H2OFrame = toH2OFrame(this, df, frameName)
+  def asH2OFrame(df : DataFrame, frameName: String): H2OFrame = asH2OFrame(df, Option(frameName))
 
   /** Transform DataFrame to H2OFrame key */
   def toH2OFrameKey(df : DataFrame): Key[Frame] = toH2OFrameKey(df, None)
-  def toH2OFrameKey(df : DataFrame, frameName: Option[String]) : Key[Frame] = asH2OFrame(df, frameName)._key
-  def toH2OFrameKey(df : DataFrame, frameName: String) : Key[Frame] = toH2OFrameKey(df, Option(frameName))
+  def toH2OFrameKey(df : DataFrame, frameName: Option[String]): Key[Frame] = asH2OFrame(df, frameName)._key
+  def toH2OFrameKey(df : DataFrame, frameName: String): Key[Frame] = toH2OFrameKey(df, Option(frameName))
 
   /** Create a new H2OFrame based on existing Frame referenced by its key.*/
   def asH2OFrame(s: String): H2OFrame = new H2OFrame(s)
@@ -145,7 +143,7 @@ class H2OContext private (@transient val sparkContext: SparkContext, @transient
     * in case we are RDD[T] where T is class defined in REPL. This is because class T is created as inner class
     * and we are not able to create instance of class T without outer scope - which is impossible to get.
     * */
-  def asRDD[A <: Product: TypeTag: ClassTag](fr : H2OFrame) : RDD[A] = toRDD[A, H2OFrame](this, fr)
+  def asRDD[A <: Product : TypeTag : ClassTag](fr: H2OFrame): RDD[A] = toRDD[A, H2OFrame](this, fr)
 
   /** A generic convert of Frame into Product RDD type
     *
@@ -161,9 +159,9 @@ class H2OContext private (@transient val sparkContext: SparkContext, @transient
 
   /** Convert given H2O frame into DataFrame type */
   @deprecated("1.3", "Use asDataFrame")
-  def asSchemaRDD[T <: Frame](fr : T, copyMetadata: Boolean = true)(implicit sqlContext: SQLContext) : DataFrame = toDataFrame(this, fr, copyMetadata)
-  def asDataFrame[T <: Frame](fr : T, copyMetadata: Boolean = true)(implicit sqlContext: SQLContext) : DataFrame = toDataFrame(this, fr, copyMetadata)
-  def asDataFrame(s : String, copyMetadata: Boolean)(implicit sqlContext: SQLContext) : DataFrame = toDataFrame(this, new H2OFrame(s), copyMetadata)
+  def asSchemaRDD[T <: Frame](fr: T, copyMetadata: Boolean = true)(implicit sqlContext: SQLContext): DataFrame = toDataFrame(this, fr, copyMetadata)
+  def asDataFrame[T <: Frame](fr: T, copyMetadata: Boolean = true)(implicit sqlContext: SQLContext): DataFrame = toDataFrame(this, fr, copyMetadata)
+  def asDataFrame(s: String, copyMetadata: Boolean)(implicit sqlContext: SQLContext): DataFrame = toDataFrame(this, new H2OFrame(s), copyMetadata)
 
   def h2oLocalClient = this.localClientIp + ":" + this.localClientPort
 
@@ -210,7 +208,7 @@ class H2OContext private (@transient val sparkContext: SparkContext, @transient
   // scalastyle:on
 }
 
-object H2OContext extends Logging{
+object H2OContext extends Logging {
 
   private[H2OContext] def setInstantiatedContext(h2oContext: H2OContext): Unit = {
     synchronized {
diff --git a/core/src/main/scala/org/apache/spark/h2o/backends/internal/InternalH2OBackend.scala b/core/src/main/scala/org/apache/spark/h2o/backends/internal/InternalH2OBackend.scala
index d29947663..1887472a8 100644
--- a/core/src/main/scala/org/apache/spark/h2o/backends/internal/InternalH2OBackend.scala
+++ b/core/src/main/scala/org/apache/spark/h2o/backends/internal/InternalH2OBackend.scala
@@ -109,7 +109,7 @@ class InternalH2OBackend(@transient val hc: H2OContext) extends SparklingBackend
     H2O.waitForCloudSize(executors.length, hc.getConf.cloudTimeout)
 
     // Register web API for client
-    RestAPIManager.registerClientWebAPI(hc)
+    RestAPIManager(hc).registerAll()
     H2O.finalizeRegistration()
     executors
   }
diff --git a/core/src/main/scala/water/api/RestAPIManager.scala b/core/src/main/scala/water/api/RestAPIManager.scala
index 0a78564bf..93f06b858 100644
--- a/core/src/main/scala/water/api/RestAPIManager.scala
+++ b/core/src/main/scala/water/api/RestAPIManager.scala
@@ -17,6 +17,8 @@
 
 package water.api
 
+import java.util.ServiceLoader
+
 import org.apache.spark.SparkContext
 import org.apache.spark.h2o.H2OContext
 import water.api.DataFrames.DataFramesHandler
@@ -24,9 +26,34 @@ import water.api.H2OFrames.H2OFramesHandler
 import water.api.RDDs.RDDsHandler
 import water.api.scalaInt.ScalaCodeHandler
 
+trait RestApi {
+  def register(h2oContext: H2OContext): Unit
+}
+
+private[api] class RestAPIManager(hc: H2OContext) {
+  private val loader: ServiceLoader[RestApi] = ServiceLoader.load(classOf[RestApi])
+
+  def registerAll(): Unit = {
+    // Register first the core
+    register(CoreRestApi)
+    // Then additional APIs
+    import scala.collection.JavaConversions._
+    loader.reload()
+    loader.foreach(api => register(api))
+  }
+
+  def register(api: RestApi): Unit = {
+    api.register(hc)
+  }
+}
 
 object RestAPIManager {
-  def registerClientWebAPI(h2oContext: H2OContext): Unit = {
+  def apply(hc: H2OContext) = new RestAPIManager(hc)
+}
+
+private object CoreRestApi extends RestApi {
+
+  def register(h2oContext: H2OContext): Unit = {
     if(h2oContext.getConf.isH2OReplEnabled){
       registerScalaIntEndp(h2oContext.sparkContext, h2oContext)
     }
@@ -44,8 +71,8 @@ object RestAPIManager {
     }
 
     RequestServer.registerEndpoint("getDataFrame", "POST", "/3/h2oframes/{h2oframe_id}/dataframe",
-      classOf[H2OFramesHandler], "toDataFrame", "Transform H2OFrame with given ID to Spark's DataFrame",
-      h2oFramesFactory)
+                                   classOf[H2OFramesHandler], "toDataFrame", "Transform H2OFrame with given ID to Spark's DataFrame",
+                                   h2oFramesFactory)
 
   }
 
@@ -56,14 +83,14 @@ object RestAPIManager {
     def rddsFactory = new HandlerFactory {
       override def create(aClass: Class[_ <: Handler]): Handler = rddsHandler
     }
-   RequestServer.registerEndpoint("listRDDs", "GET", "/3/RDDs", classOf[RDDsHandler], "list",
-     "Return all RDDs within Spark cloud", rddsFactory)
+    RequestServer.registerEndpoint("listRDDs", "GET", "/3/RDDs", classOf[RDDsHandler], "list",
+                                   "Return all RDDs within Spark cloud", rddsFactory)
 
     RequestServer.registerEndpoint("getRDD", "POST", "/3/RDDs/{rdd_id}", classOf[RDDsHandler],
-      "getRDD", "Get RDD with the given ID from Spark cloud", rddsFactory)
+                                   "getRDD", "Get RDD with the given ID from Spark cloud", rddsFactory)
 
     RequestServer.registerEndpoint("rddToH2OFrame", "POST", "/3/RDDs/{rdd_id}/h2oframe",
-      classOf[RDDsHandler], "toH2OFrame", "Transform RDD with the given ID to H2OFrame", rddsFactory)
+                                   classOf[RDDsHandler], "toH2OFrame", "Transform RDD with the given ID to H2OFrame", rddsFactory)
 
   }
 
@@ -76,14 +103,14 @@ object RestAPIManager {
     }
 
     RequestServer.registerEndpoint("listDataFrames", "GET", "/3/dataframes",
-      classOf[DataFramesHandler], "list", "Return all Spark's DataFrames", dataFramesfactory)
+                                   classOf[DataFramesHandler], "list", "Return all Spark's DataFrames", dataFramesfactory)
 
     RequestServer.registerEndpoint("getDataFrame", "POST", "/3/dataframes/{dataframe_id}",
-      classOf[DataFramesHandler], "getDataFrame", "Get Spark's DataFrame with the given ID", dataFramesfactory)
+                                   classOf[DataFramesHandler], "getDataFrame", "Get Spark's DataFrame with the given ID", dataFramesfactory)
 
     RequestServer.registerEndpoint("dataFrametoH2OFrame", "POST",
-      "/3/dataframes/{dataframe_id}/h2oframe", classOf[DataFramesHandler], "toH2OFrame",
-      "Transform Spark's DataFrame with the given ID to H2OFrame", dataFramesfactory)
+                                   "/3/dataframes/{dataframe_id}/h2oframe", classOf[DataFramesHandler], "toH2OFrame",
+                                   "Transform Spark's DataFrame with the given ID to H2OFrame", dataFramesfactory)
 
   }
 
@@ -93,18 +120,18 @@ object RestAPIManager {
       override def create(aClass: Class[_ <: Handler]): Handler = scalaCodeHandler
     }
     RequestServer.registerEndpoint("interpretScalaCode", "POST" ,"/3/scalaint/{session_id}",
-      classOf[ScalaCodeHandler], "interpret", "Interpret the code and return the result",
-      scalaCodeFactory)
+                                   classOf[ScalaCodeHandler], "interpret", "Interpret the code and return the result",
+                                   scalaCodeFactory)
 
     RequestServer.registerEndpoint("initScalaSession", "POST", "/3/scalaint",
-      classOf[ScalaCodeHandler], "initSession", "Return session id for communication with scala interpreter",
-      scalaCodeFactory)
+                                   classOf[ScalaCodeHandler], "initSession", "Return session id for communication with scala interpreter",
+                                   scalaCodeFactory)
 
     RequestServer.registerEndpoint("getScalaSessions", "GET" ,"/3/scalaint",
-      classOf[ScalaCodeHandler], "getSessions", "Return all active session IDs", scalaCodeFactory)
+                                   classOf[ScalaCodeHandler], "getSessions", "Return all active session IDs", scalaCodeFactory)
 
     RequestServer.registerEndpoint("destroyScalaSession", "DELETE", "/3/scalaint/{session_id}",
-      classOf[ScalaCodeHandler], "destroySession", "Return session id for communication with scala interpreter",
-      scalaCodeFactory)
+                                   classOf[ScalaCodeHandler], "destroySession", "Return session id for communication with scala interpreter",
+                                   scalaCodeFactory)
   }
 }
diff --git a/examples/flows/Spark_SVM_Model.flow b/examples/flows/Spark_SVM_Model.flow
new file mode 100644
index 000000000..235eb17e1
--- /dev/null
+++ b/examples/flows/Spark_SVM_Model.flow
@@ -0,0 +1,45 @@
+{
+  "version": "1.0.0",
+  "cells": [
+    {
+      "type": "sca",
+      "input": "import org.apache.spark.mllib.linalg.Vectors\nimport scala.util.Random\nval df = sc.parallelize(1 to 50).map(v => {\n  val values = Array.fill(5){0}.map(x => Random.nextDouble())\n  // Use this for regression instead of binomial. Threshold has to be set to NaN during model building\n  //val label = Math.round(Random.nextDouble())\n  val label = if(Math.round(Random.nextDouble()) > 0.5) \"1\" else \"0\"\n  (label, Vectors.dense(values))\n  }).toDF(\"Label\", \"Vector\")\n\nval hc = H2OContext.getOrCreate(sc)\nhc.asH2OFrame(df, \"bubbles\")\nval weightsDF = sc.parallelize(Array(Tuple1(Vectors.dense(1,1,1,1,1)))).toDF(\"Vector\")\nhc.asH2OFrame(weightsDF, \"weights\")"
+    },
+    {
+      "type": "cs",
+      "input": "changeColumnType frame: \"bubbles\", column: \"Label\", type: 'enum'"
+    },
+    {
+      "type": "cs",
+      "input": "buildModel \"svm\""
+    },
+    {
+      "type": "cs",
+      "input": "buildModel 'svm', {\"model_id\":\"test1\",\"training_frame\":\"bubbles\",\"response_column\":\"Label\",\"initial_weights_frame\":\"weights\",\"nfolds\":0,\"add_intercept\":false,\"step_size\":1,\"reg_param\":0.01,\"convergence_tol\":0.001,\"mini_batch_fraction\":1,\"threshold\":0,\"updater\":\"L2\",\"gradient\":\"Hinge\",\"ignored_columns\":[],\"ignore_const_cols\":true}"
+    },
+    {
+      "type": "cs",
+      "input": "getModel \"test1\""
+    },
+    {
+      "type": "cs",
+      "input": "predict model: \"test1\""
+    },
+    {
+      "type": "cs",
+      "input": "predict model: \"test1\", frame: \"bubbles\", predictions_frame: \"prediction-test1\""
+    },
+    {
+      "type": "cs",
+      "input": "bindFrames \"combined-prediction-test1\", [ \"prediction-test1\", \"bubbles\" ]"
+    },
+    {
+      "type": "cs",
+      "input": "getFrameSummary \"combined-prediction-test1\""
+    },
+    {
+      "type": "cs",
+      "input": "getFrameData \"combined-prediction-test1\""
+    }
+  ]
+}
\ No newline at end of file
diff --git a/examples/smalldata/bcwd.csv b/examples/smalldata/bcwd.csv
new file mode 100644
index 000000000..56d4032dc
--- /dev/null
+++ b/examples/smalldata/bcwd.csv
@@ -0,0 +1,700 @@
+clump_thickness,uniformity_cell_size,uniformity_cell_shape,marginal_adhesion,single_epithelial_cell_size,bare_nuclei,bland_chromatin,normal_nucleoli,mitoses,label
+5,1,1,1,2,1,3,1,1,B
+5,4,4,5,7,10,3,2,1,B
+3,1,1,1,2,2,3,1,1,B
+6,8,8,1,3,4,3,7,1,B
+4,1,1,3,2,1,3,1,1,B
+8,10,10,8,7,10,9,7,1,M
+1,1,1,1,2,10,3,1,1,B
+2,1,2,1,2,1,3,1,1,B
+2,1,1,1,2,1,1,1,5,B
+4,2,1,1,2,1,2,1,1,B
+1,1,1,1,1,1,3,1,1,B
+2,1,1,1,2,1,2,1,1,B
+5,3,3,3,2,3,4,4,1,M
+1,1,1,1,2,3,3,1,1,B
+8,7,5,10,7,9,5,5,4,M
+7,4,6,4,6,1,4,3,1,M
+4,1,1,1,2,1,2,1,1,B
+4,1,1,1,2,1,3,1,1,B
+10,7,7,6,4,10,4,1,2,M
+6,1,1,1,2,1,3,1,1,B
+7,3,2,10,5,10,5,4,4,M
+10,5,5,3,6,7,7,10,1,M
+3,1,1,1,2,1,2,1,1,B
+8,4,5,1,2,5,7,3,1,M
+1,1,1,1,2,1,3,1,1,B
+5,2,3,4,2,7,3,6,1,M
+3,2,1,1,1,1,2,1,1,B
+5,1,1,1,2,1,2,1,1,B
+2,1,1,1,2,1,2,1,1,B
+1,1,3,1,2,1,1,1,1,B
+3,1,1,1,1,1,2,1,1,B
+2,1,1,1,2,1,3,1,1,B
+10,7,7,3,8,5,7,4,3,M
+2,1,1,2,2,1,3,1,1,B
+3,1,2,1,2,1,2,1,1,B
+2,1,1,1,2,1,2,1,1,B
+10,10,10,8,6,1,8,9,1,M
+6,2,1,1,1,1,7,1,1,B
+5,4,4,9,2,10,5,6,1,M
+2,5,3,3,6,7,7,5,1,M
+6,6,6,9,6,5,7,8,1,B
+10,4,3,1,3,3,6,5,2,M
+6,10,10,2,8,10,7,3,3,M
+5,6,5,6,10,1,3,1,1,M
+10,10,10,4,8,1,8,10,1,M
+1,1,1,1,2,1,2,1,2,B
+3,7,7,4,4,9,4,8,1,M
+1,1,1,1,2,1,2,1,1,B
+4,1,1,3,2,1,3,1,1,B
+7,8,7,2,4,8,3,8,2,M
+9,5,8,1,2,3,2,1,5,M
+5,3,3,4,2,4,3,4,1,M
+10,3,6,2,3,5,4,10,2,M
+5,5,5,8,10,8,7,3,7,M
+10,5,5,6,8,8,7,1,1,M
+10,6,6,3,4,5,3,6,1,M
+8,10,10,1,3,6,3,9,1,M
+8,2,4,1,5,1,5,4,4,M
+5,2,3,1,6,10,5,1,1,M
+9,5,5,2,2,2,5,1,1,M
+5,3,5,5,3,3,4,10,1,M
+1,1,1,1,2,2,2,1,1,B
+9,10,10,1,10,8,3,3,1,M
+6,3,4,1,5,2,3,9,1,M
+1,1,1,1,2,1,2,1,1,B
+10,4,2,1,3,2,4,3,10,M
+4,1,1,1,2,1,3,1,1,B
+5,3,4,1,8,10,4,9,1,M
+8,3,8,3,4,9,8,9,8,M
+1,1,1,1,2,1,3,2,1,B
+5,1,3,1,2,1,2,1,1,B
+6,10,2,8,10,2,7,8,10,M
+1,3,3,2,2,1,7,2,1,B
+9,4,5,10,6,10,4,8,1,M
+10,6,4,1,3,4,3,2,3,M
+1,1,2,1,2,2,4,2,1,B
+1,1,4,1,2,1,2,1,1,B
+5,3,1,2,2,1,2,1,1,B
+3,1,1,1,2,3,3,1,1,B
+2,1,1,1,3,1,2,1,1,B
+2,2,2,1,1,1,7,1,1,B
+4,1,1,2,2,1,2,1,1,B
+5,2,1,1,2,1,3,1,1,B
+3,1,1,1,2,2,7,1,1,B
+3,5,7,8,8,9,7,10,7,M
+5,10,6,1,10,4,4,10,10,M
+3,3,6,4,5,8,4,4,1,M
+3,6,6,6,5,10,6,8,3,M
+4,1,1,1,2,1,3,1,1,B
+2,1,1,2,3,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+3,1,1,2,2,1,1,1,1,B
+4,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,2,1,1,B
+2,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,3,1,1,B
+2,1,1,2,2,1,1,1,1,B
+5,1,1,1,2,1,3,1,1,B
+9,6,9,2,10,6,2,9,10,M
+7,5,6,10,5,10,7,9,4,M
+10,3,5,1,10,5,3,10,2,M
+2,3,4,4,2,5,2,5,1,M
+4,1,2,1,2,1,3,1,1,B
+8,2,3,1,6,3,7,1,1,M
+10,10,10,10,10,1,8,8,8,M
+7,3,4,4,3,3,3,2,7,M
+10,10,10,8,2,10,4,1,1,M
+1,6,8,10,8,10,5,7,1,M
+1,1,1,1,2,1,2,3,1,B
+6,5,4,4,3,9,7,8,3,M
+1,3,1,2,2,2,5,3,2,B
+8,6,4,3,5,9,3,1,1,M
+10,3,3,10,2,10,7,3,3,M
+10,10,10,3,10,8,8,1,1,M
+3,3,2,1,2,3,3,1,1,B
+1,1,1,1,2,5,1,1,1,B
+8,3,3,1,2,2,3,2,1,B
+4,5,5,10,4,10,7,5,8,M
+1,1,1,1,4,3,1,1,1,B
+3,2,1,1,2,2,3,1,1,B
+1,1,2,2,2,1,3,1,1,B
+4,2,1,1,2,2,3,1,1,B
+10,10,10,2,10,10,5,3,3,M
+5,3,5,1,8,10,5,3,1,M
+5,4,6,7,9,7,8,10,1,M
+1,1,1,1,2,1,2,1,1,B
+7,5,3,7,4,10,7,5,5,M
+3,1,1,1,2,1,3,1,1,B
+8,3,5,4,5,10,1,6,2,M
+1,1,1,1,10,1,1,1,1,B
+5,1,3,1,2,1,2,1,1,B
+2,1,1,1,2,1,3,1,1,B
+5,10,8,10,8,10,3,6,3,M
+3,1,1,1,2,1,2,2,1,B
+3,1,1,1,3,1,2,1,1,B
+5,1,1,1,2,2,3,3,1,B
+4,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,1,1,1,B
+4,1,2,1,2,1,2,1,1,B
+1,1,1,1,1,5,2,1,1,B
+3,1,1,1,2,1,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+9,5,5,4,4,5,4,3,3,M
+1,1,1,1,2,5,1,1,1,B
+2,1,1,1,2,1,2,1,1,B
+1,1,3,1,2,5,2,1,1,B
+3,4,5,2,6,8,4,1,1,M
+1,1,1,1,3,2,2,1,1,B
+3,1,1,3,8,1,5,8,1,B
+8,8,7,4,10,10,7,8,7,M
+1,1,1,1,1,1,3,1,1,B
+7,2,4,1,6,10,5,4,3,M
+10,10,8,6,4,5,8,10,1,M
+4,1,1,1,2,3,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,5,5,6,3,10,3,1,1,M
+1,2,2,1,2,1,2,1,1,B
+2,1,1,1,2,1,3,1,1,B
+1,1,2,1,3,5,1,1,1,B
+9,9,10,3,6,10,7,10,6,M
+10,7,7,4,5,10,5,7,2,M
+4,1,1,1,2,1,3,2,1,B
+3,1,1,1,2,1,3,1,1,B
+1,1,1,2,1,3,1,1,7,B
+5,1,1,1,2,5,3,1,1,B
+4,1,1,1,2,2,3,2,1,B
+5,6,7,8,8,10,3,10,3,M
+10,8,10,10,6,1,3,1,10,M
+3,1,1,1,2,1,3,1,1,B
+1,1,1,2,1,1,1,1,1,B
+3,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,2,1,1,B
+6,10,10,10,8,10,10,10,7,M
+8,6,5,4,3,10,6,1,1,M
+5,8,7,7,10,10,5,7,1,M
+2,1,1,1,2,1,3,1,1,B
+5,10,10,3,8,1,5,10,3,M
+4,1,1,1,2,1,3,1,1,B
+5,3,3,3,6,10,3,1,1,M
+1,1,1,1,1,1,3,1,1,B
+1,1,1,1,2,1,1,1,1,B
+6,1,1,1,2,1,3,1,1,B
+5,8,8,8,5,10,7,8,1,M
+8,7,6,4,4,10,5,1,1,M
+2,1,1,1,1,1,3,1,1,B
+1,5,8,6,5,8,7,10,1,M
+10,5,6,10,6,10,7,7,10,M
+5,8,4,10,5,8,9,10,1,M
+1,2,3,1,2,1,3,1,1,B
+10,10,10,8,6,8,7,10,1,M
+7,5,10,10,10,10,4,10,3,M
+5,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+3,1,1,1,2,1,3,1,1,B
+4,1,1,1,2,1,3,1,1,B
+8,4,4,5,4,7,7,8,2,B
+5,1,1,4,2,1,3,1,1,B
+1,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,2,1,1,B
+9,7,7,5,5,10,7,8,3,M
+10,8,8,4,10,10,8,1,1,M
+1,1,1,1,2,1,3,1,1,B
+5,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,3,1,1,B
+5,10,10,9,6,10,7,10,5,M
+10,10,9,3,7,5,3,5,1,M
+1,1,1,1,1,1,3,1,1,B
+1,1,1,1,1,1,3,1,1,B
+5,1,1,1,1,1,3,1,1,B
+8,10,10,10,5,10,8,10,6,M
+8,10,8,8,4,8,7,7,1,M
+1,1,1,1,2,1,3,1,1,B
+10,10,10,10,7,10,7,10,4,M
+10,10,10,10,3,10,10,6,1,M
+8,7,8,7,5,5,5,10,2,M
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+6,10,7,7,6,4,8,10,2,M
+6,1,3,1,2,1,3,1,1,B
+1,1,1,2,2,1,3,1,1,B
+10,6,4,3,10,10,9,10,1,M
+4,1,1,3,1,5,2,1,1,M
+7,5,6,3,3,8,7,4,1,M
+10,5,5,6,3,10,7,9,2,M
+1,1,1,1,2,1,2,1,1,B
+10,5,7,4,4,10,8,9,1,M
+8,9,9,5,3,5,7,7,1,M
+1,1,1,1,1,1,3,1,1,B
+10,10,10,3,10,10,9,10,1,M
+7,4,7,4,3,7,7,6,1,M
+6,8,7,5,6,8,8,9,2,M
+8,4,6,3,3,1,4,3,1,B
+10,4,5,5,5,10,4,1,1,M
+3,3,2,1,3,1,3,6,1,B
+3,1,4,1,2,5,3,1,1,B
+10,8,8,2,8,10,4,8,10,M
+9,8,8,5,6,2,4,10,4,M
+8,10,10,8,6,9,3,10,10,M
+10,4,3,2,3,10,5,3,2,M
+5,1,3,3,2,2,2,3,1,B
+3,1,1,3,1,1,3,1,1,B
+2,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,5,5,1,1,B
+1,1,1,1,2,1,3,1,1,B
+5,1,1,2,2,2,3,1,1,B
+8,10,10,8,5,10,7,8,1,M
+8,4,4,1,2,9,3,3,1,M
+4,1,1,1,2,1,3,6,1,B
+3,1,1,1,2,5,3,1,1,B
+1,2,2,1,2,1,1,1,1,B
+10,4,4,10,2,10,5,3,3,M
+6,3,3,5,3,10,3,5,3,B
+6,10,10,2,8,10,7,3,3,M
+9,10,10,1,10,8,3,3,1,M
+5,6,6,2,4,10,3,6,1,M
+3,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,3,1,1,B
+5,7,7,1,5,8,3,4,1,B
+10,5,8,10,3,10,5,1,3,M
+5,10,10,6,10,10,10,6,5,M
+8,8,9,4,5,10,7,8,1,M
+10,4,4,10,6,10,5,5,1,M
+7,9,4,10,10,3,5,3,3,M
+5,1,4,1,2,1,3,2,1,B
+10,10,6,3,3,10,4,3,2,M
+3,3,5,2,3,10,7,1,1,M
+10,8,8,2,3,4,8,7,8,M
+1,1,1,1,2,1,3,1,1,B
+8,4,7,1,3,10,3,9,2,M
+5,1,1,1,2,1,3,1,1,B
+3,3,5,2,3,10,7,1,1,M
+7,2,4,1,3,4,3,3,1,M
+3,1,1,1,2,1,3,2,1,B
+3,1,3,1,2,5,2,1,1,B
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+10,5,7,3,3,7,3,3,8,M
+3,1,1,1,2,1,3,1,1,B
+2,1,1,2,2,1,3,1,1,B
+1,4,3,10,4,10,5,6,1,M
+10,4,6,1,2,10,5,3,1,M
+7,4,5,10,2,10,3,8,2,M
+8,10,10,10,8,10,10,7,3,M
+10,10,10,10,10,10,4,10,10,M
+3,1,1,1,3,1,2,1,1,B
+6,1,3,1,4,5,5,10,1,M
+5,6,6,8,6,10,4,10,4,M
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,3,1,1,B
+8,8,8,1,2,5,6,10,1,M
+10,4,4,6,2,10,2,3,1,M
+1,1,1,1,2,5,2,1,1,B
+5,5,7,8,6,10,7,4,1,M
+5,3,4,3,4,5,4,7,1,B
+5,4,3,1,2,5,2,3,1,B
+8,2,1,1,5,1,1,1,1,B
+9,1,2,6,4,10,7,7,2,M
+8,4,10,5,4,4,7,10,1,M
+1,1,1,1,2,1,3,1,1,B
+10,10,10,7,9,10,7,10,10,M
+1,1,1,1,2,1,3,1,1,B
+8,3,4,9,3,10,3,3,1,M
+10,8,4,4,4,10,3,10,4,M
+1,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,3,1,1,B
+7,8,7,6,4,3,8,8,4,M
+3,1,1,1,2,5,5,1,1,B
+2,1,1,1,3,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+8,6,4,10,10,1,3,5,1,M
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,1,1,2,1,1,B
+4,6,5,6,7,5,4,9,1,B
+5,5,5,2,5,10,4,3,1,M
+6,8,7,8,6,8,8,9,1,M
+1,1,1,1,5,1,3,1,1,B
+4,4,4,4,6,5,7,3,1,B
+7,6,3,2,5,10,7,4,6,M
+3,1,1,1,2,5,3,1,1,B
+3,1,1,1,2,1,3,1,1,B
+5,4,6,10,2,10,4,1,1,M
+1,1,1,1,2,1,3,1,1,B
+3,2,2,1,2,1,2,3,1,B
+10,1,1,1,2,10,5,4,1,M
+1,1,1,1,2,1,2,1,1,B
+8,10,3,2,6,4,3,10,1,M
+10,4,6,4,5,10,7,1,1,M
+10,4,7,2,2,8,6,1,1,M
+5,1,1,1,2,1,3,1,2,B
+5,2,2,2,2,1,2,2,1,B
+5,4,6,6,4,10,4,3,1,M
+8,6,7,3,3,10,3,4,2,M
+1,1,1,1,2,1,1,1,1,B
+6,5,5,8,4,10,3,4,1,M
+1,1,1,1,2,1,3,1,1,B
+1,1,1,1,1,1,2,1,1,B
+8,5,5,5,2,10,4,3,1,M
+10,3,3,1,2,10,7,6,1,M
+1,1,1,1,2,1,3,1,1,B
+2,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+7,6,4,8,10,10,9,5,3,M
+1,1,1,1,2,1,1,1,1,B
+5,2,2,2,3,1,1,3,1,B
+1,1,1,1,1,1,1,3,1,B
+3,4,4,10,5,1,3,3,1,M
+4,2,3,5,3,8,7,6,1,M
+5,1,1,3,2,1,1,1,1,B
+2,1,1,1,2,1,3,1,1,B
+3,4,5,3,7,3,4,6,1,B
+2,7,10,10,7,10,4,9,4,M
+1,1,1,1,2,1,2,1,1,B
+4,1,1,1,3,1,2,2,1,B
+5,3,3,1,3,3,3,3,3,M
+8,10,10,7,10,10,7,3,8,M
+8,10,5,3,8,4,4,10,3,M
+10,3,5,4,3,7,3,5,3,M
+6,10,10,10,10,10,8,10,10,M
+3,10,3,10,6,10,5,1,4,M
+3,2,2,1,4,3,2,1,1,B
+4,4,4,2,2,3,2,1,1,B
+2,1,1,1,2,1,3,1,1,B
+2,1,1,1,2,1,2,1,1,B
+6,10,10,10,8,10,7,10,7,M
+5,8,8,10,5,10,8,10,3,M
+1,1,3,1,2,1,1,1,1,B
+1,1,3,1,1,1,2,1,1,B
+4,3,2,1,3,1,2,1,1,B
+1,1,3,1,2,1,1,1,1,B
+4,1,2,1,2,1,2,1,1,B
+5,1,1,2,2,1,2,1,1,B
+3,1,2,1,2,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,1,1,2,1,1,B
+3,1,1,4,3,1,2,2,1,B
+5,3,4,1,4,1,3,1,1,B
+1,1,1,1,2,1,1,1,1,B
+10,6,3,6,4,10,7,8,4,M
+3,2,2,2,2,1,3,2,1,B
+2,1,1,1,2,1,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+3,3,2,2,3,1,1,2,3,B
+7,6,6,3,2,10,7,1,1,M
+5,3,3,2,3,1,3,1,1,B
+2,1,1,1,2,1,2,2,1,B
+5,1,1,1,3,2,2,2,1,B
+1,1,1,2,2,1,2,1,1,B
+10,8,7,4,3,10,7,9,1,M
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,1,1,1,1,1,B
+1,2,3,1,2,1,2,1,1,B
+3,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,3,1,1,B
+4,1,1,1,2,1,1,1,1,B
+3,2,1,1,2,1,2,2,1,B
+1,2,3,1,2,1,1,1,1,B
+3,10,8,7,6,9,9,3,8,M
+3,1,1,1,2,1,1,1,1,B
+5,3,3,1,2,1,2,1,1,B
+3,1,1,1,2,4,1,1,1,B
+1,2,1,3,2,1,1,2,1,B
+1,1,1,1,2,1,2,1,1,B
+4,2,2,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+2,3,2,2,2,2,3,1,1,B
+3,1,2,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,1,5,2,1,1,B
+10,10,10,6,8,4,8,5,1,M
+5,1,2,1,2,1,3,1,1,B
+8,5,6,2,3,10,6,6,1,M
+3,3,2,6,3,3,3,5,1,B
+8,7,8,5,10,10,7,2,1,M
+1,1,1,1,2,1,2,1,1,B
+5,2,2,2,2,2,3,2,2,B
+2,3,1,1,5,1,1,1,1,B
+3,2,2,3,2,3,3,1,1,B
+10,10,10,7,10,10,8,2,1,M
+4,3,3,1,2,1,3,3,1,B
+5,1,3,1,2,1,2,1,1,B
+3,1,1,1,2,1,1,1,1,B
+9,10,10,10,10,10,10,10,1,M
+5,3,6,1,2,1,1,1,1,B
+8,7,8,2,4,2,5,10,1,M
+1,1,1,1,2,1,2,1,1,B
+2,1,1,1,2,1,2,1,1,B
+1,3,1,1,2,1,2,2,1,B
+5,1,1,3,4,1,3,2,1,B
+5,1,1,1,2,1,2,2,1,B
+3,2,2,3,2,1,1,1,1,B
+6,9,7,5,5,8,4,2,1,B
+10,8,10,1,3,10,5,1,1,M
+10,10,10,1,6,1,2,8,1,M
+4,1,1,1,2,1,1,1,1,B
+4,1,3,3,2,1,1,1,1,B
+5,1,1,1,2,1,1,1,1,B
+10,4,3,10,4,10,10,1,1,M
+5,2,2,4,2,4,1,1,1,B
+1,1,1,3,2,3,1,1,1,B
+1,1,1,1,2,2,1,1,1,B
+5,1,1,6,3,1,2,1,1,B
+2,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,1,1,1,2,1,1,1,1,B
+1,1,1,1,1,1,1,1,1,B
+5,7,9,8,6,10,8,10,1,M
+4,1,1,3,1,1,2,1,1,B
+5,1,1,1,2,1,1,1,1,B
+3,1,1,3,2,1,1,1,1,B
+4,5,5,8,6,10,10,7,1,M
+2,3,1,1,3,1,1,1,1,B
+10,2,2,1,2,6,1,1,2,M
+10,6,5,8,5,10,8,6,1,M
+8,8,9,6,6,3,10,10,1,M
+5,1,2,1,2,1,1,1,1,B
+5,1,3,1,2,1,1,1,1,B
+5,1,1,3,2,1,1,1,1,B
+3,1,1,1,2,5,1,1,1,B
+6,1,1,3,2,1,1,1,1,B
+4,1,1,1,2,1,1,2,1,B
+4,1,1,1,2,1,1,1,1,B
+10,9,8,7,6,4,7,10,3,M
+10,6,6,2,4,10,9,7,1,M
+6,6,6,5,4,10,7,6,2,M
+4,1,1,1,2,1,1,1,1,B
+1,1,2,1,2,1,2,1,1,B
+3,1,1,1,1,1,2,1,1,B
+6,1,1,3,2,1,1,1,1,B
+6,1,1,1,1,1,1,1,1,B
+4,1,1,1,2,1,1,1,1,B
+5,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,1,1,1,B
+4,1,2,1,2,1,1,1,1,B
+4,1,1,1,2,1,1,1,1,B
+5,2,1,1,2,1,1,1,1,B
+4,8,7,10,4,10,7,5,1,M
+5,1,1,1,1,1,1,1,1,B
+5,3,2,4,2,1,1,1,1,B
+9,10,10,10,10,5,10,10,10,M
+8,7,8,5,5,10,9,10,1,M
+5,1,2,1,2,1,1,1,1,B
+1,1,1,3,1,3,1,1,1,B
+3,1,1,1,1,1,2,1,1,B
+10,10,10,10,6,10,8,1,5,M
+3,6,4,10,3,3,3,4,1,M
+6,3,2,1,3,4,4,1,1,M
+1,1,1,1,2,1,1,1,1,B
+5,8,9,4,3,10,7,1,1,M
+4,1,1,1,1,1,2,1,1,B
+5,10,10,10,6,10,6,5,2,M
+5,1,2,10,4,5,2,1,1,B
+3,1,1,1,1,1,2,1,1,B
+1,1,1,1,1,1,1,1,1,B
+4,2,1,1,2,1,1,1,1,B
+4,1,1,1,2,1,2,1,1,B
+4,1,1,1,2,1,2,1,1,B
+6,1,1,1,2,1,3,1,1,B
+4,1,1,1,2,1,2,1,1,B
+4,1,1,2,2,1,2,1,1,B
+4,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,1,1,1,B
+3,3,1,1,2,1,1,1,1,B
+8,10,10,10,7,5,4,8,7,M
+1,1,1,1,2,4,1,1,1,B
+5,1,1,1,2,1,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,1,1,1,1,B
+3,1,1,1,1,1,2,1,1,B
+6,6,7,10,3,10,8,10,2,M
+4,10,4,7,3,10,9,10,1,M
+1,1,1,1,1,1,1,1,1,B
+1,1,1,1,1,1,2,1,1,B
+3,1,2,2,2,1,1,1,1,B
+4,7,8,3,4,10,9,1,1,M
+1,1,1,1,3,1,1,1,1,B
+4,1,1,1,3,1,1,1,1,B
+10,4,5,4,3,5,7,3,1,M
+7,5,6,10,4,10,5,3,1,M
+3,1,1,1,2,1,2,1,1,B
+3,1,1,2,2,1,1,1,1,B
+4,1,1,1,2,1,1,1,1,B
+4,1,1,1,2,1,3,1,1,B
+6,1,3,2,2,1,1,1,1,B
+4,1,1,1,1,1,2,1,1,B
+7,4,4,3,4,10,6,9,1,M
+4,2,2,1,2,1,2,1,1,B
+1,1,1,1,1,1,3,1,1,B
+3,1,1,1,2,1,2,1,1,B
+2,1,1,1,2,1,2,1,1,B
+1,1,3,2,2,1,3,1,1,B
+5,1,1,1,2,1,3,1,1,B
+5,1,2,1,2,1,3,1,1,B
+4,1,1,1,2,1,2,1,1,B
+6,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,2,2,1,1,B
+3,1,1,1,2,1,1,1,1,B
+5,3,1,1,2,1,1,1,1,B
+4,1,1,1,2,1,2,1,1,B
+2,1,3,2,2,1,2,1,1,B
+5,1,1,1,2,1,2,1,1,B
+6,10,10,10,4,10,7,10,1,M
+2,1,1,1,1,1,1,1,1,B
+3,1,1,1,1,1,1,1,1,B
+7,8,3,7,4,5,7,8,2,M
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+3,2,2,2,2,1,4,2,1,B
+4,4,2,1,2,5,2,1,2,B
+3,1,1,1,2,1,1,1,1,B
+4,3,1,1,2,1,4,8,1,B
+5,2,2,2,1,1,2,1,1,B
+5,1,1,3,2,1,1,1,1,B
+2,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,1,3,1,1,B
+5,1,1,1,2,1,3,1,1,B
+1,1,1,1,2,1,3,1,1,B
+3,1,1,1,2,1,2,1,1,B
+4,1,1,1,2,1,3,2,1,B
+5,7,10,10,5,10,10,10,1,M
+3,1,2,1,2,1,3,1,1,B
+4,1,1,1,2,3,2,1,1,B
+8,4,4,1,6,10,2,5,2,M
+10,10,8,10,6,5,10,3,1,M
+8,10,4,4,8,10,8,2,1,M
+7,6,10,5,3,10,9,10,2,M
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+10,9,7,3,4,2,7,7,1,M
+5,1,2,1,2,1,3,1,1,B
+5,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,3,1,1,B
+5,1,2,1,2,1,2,1,1,B
+5,7,10,6,5,10,7,5,1,M
+6,10,5,5,4,10,6,10,1,M
+3,1,1,1,2,1,1,1,1,B
+5,1,1,6,3,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+8,10,10,10,6,10,10,10,1,M
+5,1,1,1,2,1,2,2,1,B
+9,8,8,9,6,3,4,1,1,M
+5,1,1,1,2,1,1,1,1,B
+4,10,8,5,4,1,10,1,1,M
+2,5,7,6,4,10,7,6,1,M
+10,3,4,5,3,10,4,1,1,M
+5,1,2,1,2,1,1,1,1,B
+4,8,6,3,4,10,7,1,1,M
+5,1,1,1,2,1,2,1,1,B
+4,1,2,1,2,1,2,1,1,B
+5,1,3,1,2,1,3,1,1,B
+3,1,1,1,2,1,2,1,1,B
+5,2,4,1,1,1,1,1,1,B
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,1,1,2,1,1,B
+4,1,1,1,2,1,2,1,1,B
+5,4,6,8,4,1,8,10,1,M
+5,3,2,8,5,10,8,1,2,M
+10,5,10,3,5,8,7,8,3,M
+4,1,1,2,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,10,10,10,10,10,10,1,1,M
+5,1,1,1,2,1,1,1,1,B
+10,4,3,10,3,10,7,1,2,M
+5,10,10,10,5,2,8,5,1,M
+8,10,10,10,6,10,10,10,10,M
+2,3,1,1,2,1,2,1,1,B
+2,1,1,1,1,1,2,1,1,B
+4,1,3,1,2,1,2,1,1,B
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,1,5,1,1,1,B
+4,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,2,1,1,B
+6,3,3,3,3,2,6,1,1,B
+7,1,2,3,2,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,1,1,2,1,1,2,1,1,B
+3,1,3,1,3,4,1,1,1,B
+4,6,6,5,7,6,7,7,3,M
+2,1,1,1,2,5,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+4,1,1,1,2,1,1,1,1,B
+6,2,3,1,2,1,1,1,1,B
+5,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+8,7,4,4,5,3,5,10,1,M
+3,1,1,1,2,1,1,1,1,B
+3,1,4,1,2,1,1,1,1,B
+10,10,7,8,7,1,10,10,3,M
+4,2,4,3,2,2,2,1,1,B
+4,1,1,1,2,1,1,1,1,B
+5,1,1,3,2,1,1,1,1,B
+4,1,1,3,2,1,1,1,1,B
+3,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,2,1,1,B
+1,2,2,1,2,1,1,1,1,B
+1,1,1,3,2,1,1,1,1,B
+5,10,10,10,10,2,10,10,10,M
+3,1,1,1,2,1,2,1,1,B
+3,1,1,2,3,4,1,1,1,B
+1,2,1,3,2,1,2,1,1,B
+5,1,1,1,2,1,2,2,1,B
+4,1,1,1,2,1,2,1,1,B
+3,1,1,1,2,1,3,1,1,B
+3,1,1,1,2,1,2,1,1,B
+5,1,1,1,2,1,2,1,1,B
+5,4,5,1,8,1,3,6,1,B
+7,8,8,7,3,10,7,2,3,M
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,2,1,1,B
+4,1,1,1,2,1,3,1,1,B
+1,1,3,1,2,1,2,1,1,B
+1,1,3,1,2,1,2,1,1,B
+3,1,1,3,2,1,2,1,1,B
+1,1,1,1,2,1,1,1,1,B
+5,2,2,2,2,1,1,1,2,B
+3,1,1,1,2,1,3,1,1,B
+5,7,4,1,6,1,7,10,3,M
+5,10,10,8,5,5,7,10,1,M
+3,10,7,8,5,8,7,4,1,M
+3,2,1,2,2,1,3,1,1,B
+2,1,1,1,2,1,3,1,1,B
+5,3,2,1,3,1,1,1,1,B
+1,1,1,1,2,1,2,1,1,B
+4,1,4,1,2,1,1,1,1,B
+1,1,2,1,2,1,2,1,1,B
+5,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+10,10,10,10,5,10,10,10,7,M
+5,10,10,10,4,10,5,6,3,M
+5,1,1,1,2,1,3,2,1,B
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,2,3,1,B
+4,1,1,1,2,1,1,1,1,B
+1,1,1,1,2,1,1,1,8,B
+1,1,1,3,2,1,1,1,1,B
+5,10,10,5,4,5,4,4,1,M
+3,1,1,1,2,1,1,1,1,B
+3,1,1,1,2,1,2,1,2,B
+3,1,1,1,3,2,1,1,1,B
+2,1,1,1,2,1,1,1,1,B
+5,10,10,3,7,3,8,10,2,M
+4,8,6,4,3,4,10,6,1,M
+4,8,8,5,4,5,10,4,1,M
diff --git a/examples/src/main/scala/org/apache/spark/examples/h2o/SparkSVMDemo.scala b/examples/src/main/scala/org/apache/spark/examples/h2o/SparkSVMDemo.scala
new file mode 100644
index 000000000..76f00de96
--- /dev/null
+++ b/examples/src/main/scala/org/apache/spark/examples/h2o/SparkSVMDemo.scala
@@ -0,0 +1,70 @@
+/*
+* Licensed to the Apache Software Foundation (ASF) under one or more
+* contributor license agreements.  See the NOTICE file distributed with
+* this work for additional information regarding copyright ownership.
+* The ASF licenses this file to You under the Apache License, Version 2.0
+* (the "License"); you may not use this file except in compliance with
+* the License.  You may obtain a copy of the License at
+*
+*    http://www.apache.org/licenses/LICENSE-2.0
+*
+* Unless required by applicable law or agreed to in writing, software
+* distributed under the License is distributed on an "AS IS" BASIS,
+* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+* See the License for the specific language governing permissions and
+* limitations under the License.
+*/
+package org.apache.spark.examples.h2o
+
+import java.io.File
+
+import org.apache.spark.{SparkContext, SparkFiles}
+import org.apache.spark.h2o.H2OContext
+import org.apache.spark.ml.spark.models.svm.SVM
+import org.apache.spark.ml.spark.models.svm.SVMParameters
+import org.apache.spark.sql.SQLContext
+import water.fvec.H2OFrame
+import water.support.SparkContextSupport
+
+object SparkSVMDemo extends SparkContextSupport {
+
+  def main(args: Array[String]) {
+    val conf = configure("Sparkling Water: Spark SVM demo.")
+    val sc = new SparkContext(conf)
+
+    val h2oContext = H2OContext.getOrCreate(sc)
+    implicit val sqLContext = SQLContext.getOrCreate(sc)
+
+    // Setup environment
+    addFiles(sc, absPath("examples/smalldata/bcwd.csv"))
+
+    // Load H2O from CSV file (i.e., access directly H2O cloud)
+    // Use super-fast advanced H2O CSV parser !!!
+    val breastCancerData = new H2OFrame(new File(SparkFiles.get("bcwd.csv")))
+
+    // Training data
+    breastCancerData.replace(breastCancerData.numCols()-1, breastCancerData.lastVec().toCategoricalVec)
+    breastCancerData.update()
+
+    // Configure Deep Learning algorithm
+    val parms = new SVMParameters
+    parms._train = breastCancerData.key
+    parms._response_column = "label"
+
+    val svm = new SVM(parms, h2oContext)
+
+    val svmModel = svm.trainModel.get
+
+    // Use model for scoring
+    val predictionH2OFrame = svmModel.score(breastCancerData)
+    val predictionsFromModel = h2oContext.asDataFrame(predictionH2OFrame).collect
+    println(predictionsFromModel.mkString("\n===> Model predictions: ", ",\n", ", ...\n"))
+
+    // Stop Spark cluster and destroy all executors
+    if (System.getProperty("spark.ext.h2o.preserve.executors")==null) {
+      sc.stop()
+    }
+    // Shutdown H2O
+    h2oContext.stop()
+  }
+}
diff --git a/ml/build.gradle b/ml/build.gradle
index b28d7b9d5..573e2eb0e 100644
--- a/ml/build.gradle
+++ b/ml/build.gradle
@@ -14,5 +14,6 @@ dependencies {
   // And use scalatest for Scala testing
   testCompile "org.scalatest:scalatest_${scalaBaseVersion}:2.2.1"
   testCompile "junit:junit:4.11"
-}
+  testCompile project(':sparkling-water-core').sourceSets.test.output
 
+}
diff --git a/ml/src/main/scala/hex/Register.scala b/ml/src/main/scala/hex/Register.scala
new file mode 100644
index 000000000..64102e29e
--- /dev/null
+++ b/ml/src/main/scala/hex/Register.scala
@@ -0,0 +1,59 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package hex
+
+import org.apache.spark.h2o.H2OContext
+import org.apache.spark.ml.spark.models.svm.SVM
+import water.H2O
+import water.api.{GridSearchHandler, ModelBuilderHandler, RestApi}
+
+class Register extends RestApi {
+
+  override def register(hc: H2OContext) = {
+
+    val models = Seq(new SVM(true, hc))
+
+    for (algo <- models) {
+      val base: String = algo.getClass.getSimpleName
+      val lbase: String = base.toLowerCase
+      val bh_clz = classOf[ModelBuilderHandler[_, _, _]]
+      val version: Int = 3
+      H2O.register(
+        "POST /" + version + "/ModelBuilders/" + lbase,
+        bh_clz,
+        "train",
+        "train_" + lbase,
+        "Train a " + base + " model."
+      )
+      H2O.register(
+        "POST /" + version + "/ModelBuilders/" + lbase + "/parameters",
+        bh_clz,
+        "validate_parameters",
+        "validate_" + lbase,
+        "Validate a set of " + base + " model builder parameters."
+      )
+      H2O.register(
+        "POST /99/Grid/" + lbase,
+        classOf[GridSearchHandler[_,_,_,_]],
+        "train",
+        "grid_search_" + lbase,
+        "Run grid search for " + base + " model."
+      )
+    }
+  }
+}
diff --git a/ml/src/main/scala/hex/SVMDriverUtil.java b/ml/src/main/scala/hex/SVMDriverUtil.java
new file mode 100644
index 000000000..b9e599ce0
--- /dev/null
+++ b/ml/src/main/scala/hex/SVMDriverUtil.java
@@ -0,0 +1,66 @@
+package hex;
+
+import org.apache.spark.api.java.JavaRDD;
+import org.apache.spark.api.java.function.Function;
+import org.apache.spark.ml.spark.models.svm.SVMModel;
+import org.apache.spark.mllib.regression.LabeledPoint;
+import org.apache.spark.rdd.RDD;
+import scala.Tuple2;
+import water.fvec.Frame;
+
+public class SVMDriverUtil {
+
+    public static ModelMetrics computeMetrics(SVMModel model, RDD<LabeledPoint> training,
+                                              final org.apache.spark.mllib.classification.SVMModel trainedModel,
+                                              Frame f,
+                                              String[] responseDomains) {
+
+        // Compute Spark evaluations
+        JavaRDD<Tuple2<Double, Double>> predictionAndLabels = training.toJavaRDD().map(
+                new Function<LabeledPoint, Tuple2<Double, Double>>() {
+                    public Tuple2<Double, Double> call(LabeledPoint p) {
+                        Double prediction = trainedModel.predict(p.features());
+                        return new Tuple2<>(prediction, p.label());
+                    }
+                }
+        );
+
+        double mse;
+
+        // Set the metrics
+        switch (model._output.getModelCategory()) {
+            case Binomial:
+                ModelMetricsBinomial.MetricBuilderBinomial builderB = new ModelMetricsBinomial.MetricBuilderBinomial(responseDomains);
+                for (Tuple2<Double, Double> predAct : predictionAndLabels.collect()) {
+                    Double pred = predAct._1;
+                    builderB.perRow(
+                            new double[] {pred, pred == 1 ? 0 : 1, pred == 1 ? 1 : 0},
+                            new float[] {predAct._2.floatValue()},
+                            model
+                    );
+                }
+                mse = builderB._sumsqe / builderB._nclasses;
+
+                return new ModelMetricsBinomial(
+                        model, f, builderB._count, mse, responseDomains,
+                        builderB.weightedSigma(), new AUC2(builderB._auc), builderB._logloss, null
+                );
+            default:
+                ModelMetricsRegression.MetricBuilderRegression builderR = new ModelMetricsRegression.MetricBuilderRegression();
+                for (Tuple2<Double, Double> predAct : predictionAndLabels.collect()) {
+                    Double pred = predAct._1;
+                    builderR.perRow(
+                            new double[] {pred, pred == 1 ? 0 : 1, pred == 1 ? 1 : 0},
+                            new float[] {predAct._2.floatValue()},
+                            model
+                    );
+                }
+                mse = builderR._sumsqe / builderR._nclasses;
+
+                return new ModelMetricsRegression(
+                        model, f, builderR._count, mse, builderR.weightedSigma(), builderR._sumdeviance / builderR._wcount, 0
+                );
+        }
+    }
+
+}
diff --git a/ml/src/main/scala/hex/schemas/SVMModelV3.scala b/ml/src/main/scala/hex/schemas/SVMModelV3.scala
new file mode 100644
index 000000000..59af258f8
--- /dev/null
+++ b/ml/src/main/scala/hex/schemas/SVMModelV3.scala
@@ -0,0 +1,48 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package hex.schemas
+
+import SVMV3.SVMParametersV3
+import hex.schemas.SVMModelV3.SVMModelOutputV3
+import org.apache.spark.ml.spark.models.svm.{SVMModel, SVMParameters}
+import water.api.schemas3.{ModelOutputSchemaV3, ModelSchemaV3}
+import water.api.API
+
+class SVMModelV3 extends ModelSchemaV3[SVMModel,
+  SVMModelV3,
+  SVMParameters,
+  SVMParametersV3,
+  SVMModel.SVMOutput,
+  SVMModelV3.SVMModelOutputV3] {
+
+  override def createParametersSchema(): SVMParametersV3 = { new SVMParametersV3() }
+  override def createOutputSchema(): SVMModelOutputV3 = { new SVMModelOutputV3() }
+  
+}
+
+object SVMModelV3 {
+
+  final class SVMModelOutputV3 extends ModelOutputSchemaV3[SVMModel.SVMOutput, SVMModelOutputV3] {
+    // Output fields
+    @API(help = "Iterations executed") var iterations: Int = 0
+    @API(help = "Interceptor") var interceptor: Double = 0
+    @API(help = "Weights") var weights: Array[Double] = Array()
+  }
+
+}
+  
diff --git a/ml/src/main/scala/hex/schemas/SVMV3.java b/ml/src/main/scala/hex/schemas/SVMV3.java
new file mode 100644
index 000000000..dc7f57d2c
--- /dev/null
+++ b/ml/src/main/scala/hex/schemas/SVMV3.java
@@ -0,0 +1,109 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package hex.schemas;
+
+import org.apache.spark.ml.spark.models.svm.*;
+import water.DKV;
+import water.Key;
+import water.Value;
+import water.api.API;
+import water.api.schemas3.KeyV3;
+import water.api.schemas3.ModelParametersSchemaV3;
+import water.fvec.Frame;
+
+// Seems like this has to be in Java since H2O's frameworks uses reflection's getFields...
+// I probably could mix Java and Scala here, leave SVMParametersV3 with fields as Java
+// and then make the same Scala class SVMParametersV3 which extends it but not sure if it's worth it...
+public class SVMV3 extends ModelBuilderSchema<SVM, SVMV3, SVMV3.SVMParametersV3> {
+
+    public static final class SVMParametersV3 extends
+            ModelParametersSchemaV3<SVMParameters, SVMParametersV3> {
+        public static String[] fields = new String[]{
+                "model_id",
+                "training_frame",
+                "response_column",
+                "initial_weights_frame",
+                "validation_frame",
+                "nfolds",
+                "add_intercept",
+
+                "step_size",
+                "reg_param",
+                "convergence_tol",
+                "mini_batch_fraction",
+                "threshold",
+                "updater",
+                "gradient",
+
+                "ignored_columns",
+                "ignore_const_cols"
+        };
+
+        @API(help="Add intercept.", direction=API.Direction.INPUT, gridable = true)
+        public boolean add_intercept = false;
+
+        @API(help="Set step size", direction=API.Direction.INPUT, gridable = true)
+        public double step_size = 1.0;
+
+        @API(help="Set regularization parameter", direction=API.Direction.INPUT, gridable = true)
+        public double reg_param = 0.01;
+
+        @API(help="Set convergence tolerance", direction=API.Direction.INPUT, gridable = true)
+        public double convergence_tol = 0.001;
+
+        @API(help="Set mini batch fraction", direction=API.Direction.INPUT, gridable = true)
+        public double mini_batch_fraction = 1.0;
+
+        // TODO what exactly does INOUT do?? Should this be only INPUT?
+        @API(help="Set threshold that separates positive predictions from negative ones. NaN for raw prediction.", direction=API.Direction.INOUT, gridable = true)
+        public double threshold = 0.0;
+
+        @API(help="Set the updater for SGD.", direction=API.Direction.INPUT, values = {"L2", "L1", "Simple"}, required = true, gridable = true)
+        public Updater updater = Updater.L2;
+
+        @API(help="Set the gradient computation type for SGD.", direction=API.Direction.INPUT, values = {"Hinge", "LeastSquares", "Logistic"}, required = true, gridable = true)
+        public Gradient gradient = Gradient.Hinge;
+
+        @API(help="Initial model weights.", direction=API.Direction.INOUT, gridable = true)
+        public KeyV3.FrameKeyV3 initial_weights_frame;
+
+        @Override
+        public SVMParametersV3 fillFromImpl(SVMParameters impl) {
+            super.fillFromImpl(impl);
+
+            if (null != impl._initial_weights) {
+                Value v = DKV.get(impl._initial_weights);
+                if (null != v) {
+                    initial_weights_frame = new KeyV3.FrameKeyV3(((Frame) v.get())._key);
+                }
+            }
+
+            return this;
+        }
+
+        @Override
+        public SVMParameters fillImpl(SVMParameters impl) {
+            super.fillImpl(impl);
+            impl._initial_weights =
+                    null == this.initial_weights_frame ? null : Key.<Frame>make(this.initial_weights_frame.name);
+            return impl;
+        }
+
+    }
+
+}
diff --git a/ml/src/main/scala/org/apache/spark/ml/h2o/H2OPipeline.scala b/ml/src/main/scala/org/apache/spark/ml/h2o/H2OPipeline.scala
index 0c888ca36..5420e67c6 100644
--- a/ml/src/main/scala/org/apache/spark/ml/h2o/H2OPipeline.scala
+++ b/ml/src/main/scala/org/apache/spark/ml/h2o/H2OPipeline.scala
@@ -19,8 +19,9 @@ package org.apache.spark.ml.h2o
 
 import org.apache.spark.annotation.Since
 import org.apache.spark.ml.Pipeline.SharedReadWrite
-import org.apache.spark.ml.{PipelineStage, Transformer, PipelineModel, Pipeline}
+import org.apache.spark.ml.{Pipeline, PipelineModel, PipelineStage, Transformer}
 import _root_.org.apache.spark.sql.DataFrame
+import org.apache.spark.h2o.H2OFrame
 import org.apache.spark.ml.util.{Identifiable, MLReadable, MLReader}
 
 /**
diff --git a/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Gradient.java b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Gradient.java
new file mode 100644
index 000000000..c6be2e935
--- /dev/null
+++ b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Gradient.java
@@ -0,0 +1,37 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.ml.spark.models.svm;
+
+import org.apache.spark.mllib.optimization.HingeGradient;
+import org.apache.spark.mllib.optimization.LeastSquaresGradient;
+import org.apache.spark.mllib.optimization.LogisticGradient;
+
+public enum Gradient {
+    Hinge(new HingeGradient()),
+    LeastSquares(new LeastSquaresGradient()),
+    Logistic(new LogisticGradient());
+
+    private org.apache.spark.mllib.optimization.Gradient sparkGradient;
+
+    Gradient(org.apache.spark.mllib.optimization.Gradient sparkGradient) {
+        this.sparkGradient = sparkGradient;
+    }
+
+    public org.apache.spark.mllib.optimization.Gradient get() {
+        return sparkGradient;
+    }
+}
diff --git a/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVM.java b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVM.java
new file mode 100644
index 000000000..f570f94ce
--- /dev/null
+++ b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVM.java
@@ -0,0 +1,278 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.ml.spark.models.svm;
+
+import hex.*;
+
+import org.apache.spark.api.java.function.Function;
+import org.apache.spark.SparkContext;
+import org.apache.spark.h2o.H2OContext;
+import org.apache.spark.ml.spark.models.svm.SVMModel.SVMOutput;
+import org.apache.spark.mllib.classification.SVMWithSGD;
+import org.apache.spark.mllib.linalg.Vector;
+import org.apache.spark.mllib.linalg.Vectors;
+import org.apache.spark.mllib.regression.LabeledPoint;
+import org.apache.spark.rdd.RDD;
+import org.apache.spark.sql.Row;
+import org.apache.spark.sql.SQLContext;
+import org.apache.spark.sql.types.DataTypes;
+import org.apache.spark.sql.types.StructField;
+import water.Scope;
+import water.fvec.Frame;
+import water.fvec.H2OFrame;
+import water.fvec.Vec;
+import water.util.Log;
+
+import java.util.Arrays;
+
+public class SVM extends ModelBuilder<SVMModel, SVMParameters, SVMOutput> {
+
+    transient private final H2OContext hc;
+
+    public SVM(boolean startup_once, H2OContext hc) {
+        super(new SVMParameters(), startup_once);
+        this.hc = hc;
+    }
+
+    public SVM(SVMParameters parms, H2OContext hc) {
+        super(parms);
+        init(false);
+        this.hc = hc;
+    }
+
+    @Override
+    protected Driver trainModelImpl() {
+        return new SVMDriver();
+    }
+
+    @Override
+    public ModelCategory[] can_build() {
+        return new ModelCategory[]{
+                ModelCategory.Binomial,
+                ModelCategory.Regression
+        };
+    }
+
+    @Override
+    public boolean isSupervised() {
+        return true;
+    }
+
+    @Override
+    public void init(boolean expensive) {
+        super.init(expensive);
+
+        _parms.validate(this);
+
+        if (_train == null) return;
+
+        if (null != _parms._initial_weights) {
+            Frame user_points = _parms._initial_weights.get();
+            if (user_points.numCols() != _train.numCols() - numSpecialCols()) {
+                error("_initial_weights",
+                        "The user-specified initial weights must have the same number of columns " +
+                                "(" + (_train.numCols() - numSpecialCols()) + ") as the training observations");
+            }
+
+            if (user_points.hasNAs()) {
+                error("_initial_weights", "Initial weights cannot contain missing values.");
+            }
+        }
+
+        for (int i = 0; i < _train.numCols(); i++) {
+            Vec vec = _train.vec(i);
+            String vecName = _train.name(i);
+            if (vec.naCnt() > 0 && (null == _parms._ignored_columns || Arrays.binarySearch(_parms._ignored_columns, vecName) < 0)) {
+                error("_train", "Training frame cannot contain any missing values [" + vecName + "].");
+            }
+        }
+
+        for (Vec vec : _train.vecs()) {
+            if (!(vec.isNumeric() || vec.isTime() || vec.isCategorical())) {
+                error("_train", "SVM supports only frames with numeric values. But a " + vec.get_type_str() + " was found.");
+            }
+        }
+
+        if (null != _parms._response_column && null == _train.vec(_parms._response_column)) {
+            error("_train", "Training frame has to contain the response column.");
+        }
+
+
+        if (_train != null && _parms._response_column != null) {
+            String[] responseDomains = responseDomains();
+            if (null == responseDomains) {
+                if (!(Double.isNaN(_parms._threshold))) {
+                    error("_threshold", "Threshold cannot be set for regression SVM. Set the threshold to NaN or modify the response column to an enum.");
+                }
+
+                if (!_train.vec(_parms._response_column).isNumeric()) {
+                    error("_threshold", "Threshold cannot be set for regression SVM. Set the threshold to NaN or modify the response column to an enum.");
+                    error("_response_column", "Regression SVM requires the response column type to be numeric.");
+                }
+            } else {
+                if (Double.isNaN(_parms._threshold)) {
+                    error("_threshold", "Threshold has to be set for binomial SVM. Set the threshold to a numeric value or change the response column type.");
+                }
+
+                if (responseDomains.length != 2) {
+                    error("_response_column", "Binomial SVM requires the response column's domain to be \"0\" and \"1\".");
+                }
+            }
+        }
+    }
+
+    private String[] responseDomains() {
+        int idx = _parms.train().find(_parms._response_column);
+        if (idx == -1) {
+            return null;
+        }
+        return _parms.train().domains()[idx];
+    }
+
+    @Override
+    public int numSpecialCols() {
+        return (hasOffsetCol() ? 1 : 0) +
+                (hasWeightCol() ? 1 : 0) +
+                (hasFoldCol() ? 1 : 0) + 1;
+    }
+
+    private final class SVMDriver extends Driver {
+
+        transient private H2OContext h2oContext = hc;
+        transient private SparkContext sc = hc.sparkContext();
+        transient private SQLContext sqlContext = SQLContext.getOrCreate(sc);
+
+        @Override
+        public void computeImpl() {
+            init(true);
+
+            // The model to be built
+            SVMModel model = new SVMModel(dest(), _parms, new SVMModel.SVMOutput(SVM.this));
+            model.delete_and_lock(_job);
+
+            RDD<LabeledPoint> training = getTrainingData(
+                    _train,
+                    _parms._response_column,
+                    model._output.nfeatures()
+            );
+            training.cache();
+
+            SVMWithSGD svm = new SVMWithSGD();
+            svm.setIntercept(_parms._add_intercept);
+
+            svm.optimizer().setNumIterations(_parms._max_iterations);
+
+            svm.optimizer().setStepSize(_parms._step_size);
+            svm.optimizer().setRegParam(_parms._reg_param);
+            svm.optimizer().setMiniBatchFraction(_parms._mini_batch_fraction);
+            svm.optimizer().setConvergenceTol(_parms._convergence_tol);
+            svm.optimizer().setGradient(_parms._gradient.get());
+            svm.optimizer().setUpdater(_parms._updater.get());
+
+            final org.apache.spark.mllib.classification.SVMModel trainedModel =
+                    (null == _parms._initial_weights) ?
+                            svm.run(training) :
+                            svm.run(training, vec2vec(_parms.initialWeights().vecs()));
+            training.unpersist(false);
+
+            model._output.weights_$eq(trainedModel.weights().toArray());
+            model._output.iterations_$eq(_parms._max_iterations);
+            model._output.interceptor_$eq(trainedModel.intercept());
+
+            model._output._training_metrics =
+                    SVMDriverUtil.computeMetrics(model, training, trainedModel, _parms.train(), responseDomains());
+            model.update(_job);
+
+            _job.update(model._parms._max_iterations);
+
+            if (_valid != null) {
+                model.score(_parms.valid()).delete();
+                model._output._validation_metrics = ModelMetrics.getFromDKV(model, _parms.valid());
+                model.update(_job);
+            }
+
+            model._output.interceptor_$eq(trainedModel.intercept());
+
+            Log.info(model._output._model_summary);
+
+        }
+
+        private Vector vec2vec(Vec[] vals) {
+            double[] dense = new double[vals.length];
+            for (int i = 0; i < vals.length; i++) {
+                dense[i] = vals[i].at(0);
+            }
+            return Vectors.dense(dense);
+        }
+
+        private RDD<LabeledPoint> getTrainingData(Frame parms, String _response_column, int nfeatures) {
+            return h2oContext.asDataFrame(new H2OFrame(parms), true, sqlContext)
+                    .javaRDD()
+                    .map(new RowToLabeledPoint(nfeatures, _response_column, parms.domains()[parms.find(_response_column)])).rdd();
+        }
+    }
+}
+
+class RowToLabeledPoint implements Function<Row, LabeledPoint> {
+    private final int nfeatures;
+    private final String _response_column;
+    private final String[] domains;
+
+    RowToLabeledPoint(int nfeatures, String response_column, String[] domains) {
+        this.nfeatures = nfeatures;
+        this._response_column = response_column;
+        this.domains = domains;
+    }
+
+    @Override
+    public LabeledPoint call(Row row) throws Exception {
+        StructField[] fields = row.schema().fields();
+        double[] features = new double[nfeatures];
+        for (int i = 0; i < nfeatures; i++) {
+            features[i] = toDouble(row.get(i), fields[i]);
+        }
+
+        return new LabeledPoint(
+                Arrays.binarySearch(domains, row.<String>getAs(_response_column)),
+                Vectors.dense(features));
+    }
+
+    private double toDouble(Object value, StructField fieldStruct) {
+        if (fieldStruct.dataType().sameType(DataTypes.ByteType)) {
+            return ((Byte) value).doubleValue();
+        }
+
+        if (fieldStruct.dataType().sameType(DataTypes.ShortType)) {
+            return ((Short) value).doubleValue();
+        }
+
+        if (fieldStruct.dataType().sameType(DataTypes.IntegerType)) {
+            return ((Integer) value).doubleValue();
+        }
+
+        if (fieldStruct.dataType().sameType(DataTypes.DoubleType)) {
+            return (Double) value;
+        }
+
+        if (fieldStruct.dataType().sameType(DataTypes.StringType)) {
+            return Double.parseDouble(value.toString());
+        }
+
+        throw new IllegalArgumentException("Target column has to be an enum or a number. " + fieldStruct.toString());
+    }
+}
+
diff --git a/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMModel.scala b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMModel.scala
new file mode 100644
index 000000000..bd50140d0
--- /dev/null
+++ b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMModel.scala
@@ -0,0 +1,104 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.ml.spark.models.svm
+
+import hex.ModelMetricsSupervised.MetricBuilderSupervised
+import hex._
+import water.codegen.CodeGeneratorPipeline
+import water.util.{JCodeGen, SBPrintStream}
+import water.{H2O, Key, Keyed}
+
+object SVMModel {
+
+  class SVMOutput(val b: SVM) extends Model.Output(b) {
+    var interceptor: Double = .0
+    var iterations: Int = 0
+    var weights: Array[Double] = null
+  }
+
+}
+
+class SVMModel private[svm](val selfKey: Key[_ <: Keyed[_ <: Keyed[_ <: AnyRef]]],
+                              val parms: SVMParameters,
+                              val output: SVMModel.SVMOutput)
+  extends Model[SVMModel, SVMParameters, SVMModel.SVMOutput](selfKey, parms, output) {
+
+  override def makeMetricBuilder(domain: Array[String]): MetricBuilderSupervised[Nothing] =
+    _output.getModelCategory match {
+      case ModelCategory.Binomial =>
+        new ModelMetricsBinomial.MetricBuilderBinomial(domain)
+      case ModelCategory.Regression =>
+        new ModelMetricsRegression.MetricBuilderRegression
+      case _ =>
+        throw H2O.unimpl
+    }
+
+  protected def score0(data: Array[Double], preds: Array[Double]): Array[Double] = {
+    java.util.Arrays.fill(preds, 0)
+    val pred =
+      data.zip(_output.weights).foldRight(_output.interceptor){ case ((d, w), acc) => d * w + acc}
+
+    if(_parms._threshold.isNaN) { // Regression
+      preds(0) = pred
+    } else { // Binomial
+      if(pred > _parms._threshold) {
+        preds(2) = 1
+        preds(1) = -1
+        preds(0) = 1
+      } else {
+        preds(2) = -1
+        preds(1) = 1
+        preds(0) = 0
+      }
+    }
+    preds
+  }
+
+  override protected def toJavaInit(sb: SBPrintStream, fileCtx: CodeGeneratorPipeline): SBPrintStream = {
+    val sbInitialized = super.toJavaInit(sb, fileCtx)
+    sbInitialized.ip("public boolean isSupervised() { return " + isSupervised + "; }").nl
+    JCodeGen.toStaticVar(sbInitialized, "WEIGHTS", _output.weights, "Weights.")
+    sbInitialized
+  }
+
+  override protected def toJavaPredictBody(bodySb: SBPrintStream,
+                                           classCtx: CodeGeneratorPipeline,
+                                           fileCtx: CodeGeneratorPipeline,
+                                           verboseCode: Boolean) {
+    bodySb.i.p("java.util.Arrays.fill(preds,0);").nl
+    bodySb.i.p(s"double prediction = ${_output.interceptor};").nl
+    bodySb.i.p("for(int i = 0; i < data.length; i++) {").nl
+    bodySb.i(1).p("prediction += (data[i] * WEIGHTS[i]);").nl
+    bodySb.i.p("}").nl
+
+    if (_output.nclasses == 1) {
+      bodySb.i.p("preds[0] = prediction;").nl
+    } else {
+      bodySb.i.p(s"if(prediction > ${_parms._threshold}) {").nl
+      bodySb.i(1).p("preds[2] = 1;").nl
+      bodySb.i(1).p("preds[1] = -1;").nl
+      bodySb.i(1).p("preds[0] = 1;").nl
+      bodySb.i.p(s"} else {").nl
+      bodySb.i(1).p("preds[2] = -1;").nl
+      bodySb.i(1).p("preds[1] = 1;").nl
+      bodySb.i(1).p("preds[0] = 0;").nl
+      bodySb.i.p(s"}").nl
+    }
+
+  }
+
+}
diff --git a/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMParameters.java b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMParameters.java
new file mode 100644
index 000000000..66b6ce7ce
--- /dev/null
+++ b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/SVMParameters.java
@@ -0,0 +1,63 @@
+package org.apache.spark.ml.spark.models.svm;
+
+import hex.Model;
+import water.Key;
+import water.fvec.Frame;
+
+public class SVMParameters extends Model.Parameters {
+    @Override
+    public String algoName() { return "SVM"; }
+
+    @Override
+    public String fullName() { return "Support Vector Machine (*Spark*)"; }
+
+    @Override
+    public String javaName() { return SVMModel.class.getName(); }
+
+    @Override
+    public long progressUnits() { return _max_iterations; }
+
+    public final Frame initialWeights() {
+        if (null == _initial_weights) {
+            return null;
+        } else {
+            return _initial_weights.get();
+        }
+    }
+
+    public int _max_iterations = 1000;
+    public boolean _add_intercept = false;
+    public double _step_size = 1.0;
+    public double _reg_param = 0.01;
+    public double _convergence_tol = 0.001;
+    public double _mini_batch_fraction = 1.0;
+    public double _threshold = 0.0;
+    public Updater _updater = Updater.L2;
+    public Gradient _gradient = Gradient.Hinge;
+    public Key<Frame> _initial_weights = null;
+
+    public void validate(SVM svm) {
+        if (_max_iterations < 0 || _max_iterations > 1e6) {
+            svm.error("_max_iterations", " max_iterations must be between 0 and 1e6");
+        }
+
+        if(_step_size <= 0) {
+            svm.error("_step_size", "The step size has to be positive.");
+        }
+
+        if(_reg_param <= 0) {
+            svm.error("_reg_param", "The regularization parameter has to be positive.");
+        }
+
+        if(_convergence_tol <= 0) {
+            svm.error("_convergence_tol", "The convergence tolerance has to be positive.");
+        }
+
+        if(_mini_batch_fraction <= 0) {
+            svm.error("_mini_batch_fraction", "The minimum batch fraction has to be positive.");
+        }
+
+    }
+
+
+}
diff --git a/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Updater.java b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Updater.java
new file mode 100644
index 000000000..be5ff38e5
--- /dev/null
+++ b/ml/src/main/scala/org/apache/spark/ml/spark/models/svm/Updater.java
@@ -0,0 +1,37 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.ml.spark.models.svm;
+
+import org.apache.spark.mllib.optimization.L1Updater;
+import org.apache.spark.mllib.optimization.SimpleUpdater;
+import org.apache.spark.mllib.optimization.SquaredL2Updater;
+
+public enum Updater {
+    L2(new SquaredL2Updater()),
+    L1(new L1Updater()),
+    Simple(new SimpleUpdater());
+
+    private org.apache.spark.mllib.optimization.Updater sparkUpdater;
+
+    Updater(org.apache.spark.mllib.optimization.Updater sparkUpdater) {
+        this.sparkUpdater = sparkUpdater;
+    }
+
+    public org.apache.spark.mllib.optimization.Updater get() {
+        return sparkUpdater;
+    }
+}
diff --git a/ml/src/test/scala/org/apache/spark/ml/spark/models/svm/SVMModelTest.scala b/ml/src/test/scala/org/apache/spark/ml/spark/models/svm/SVMModelTest.scala
new file mode 100644
index 000000000..c412cb1fd
--- /dev/null
+++ b/ml/src/test/scala/org/apache/spark/ml/spark/models/svm/SVMModelTest.scala
@@ -0,0 +1,81 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.ml.spark.models.svm
+
+import org.apache.spark.SparkContext
+import org.apache.spark.h2o.utils.SharedSparkTestContext
+import org.apache.spark.mllib.classification
+import org.apache.spark.mllib.linalg.Vectors
+import org.junit.runner.RunWith
+import org.scalatest.FunSuite
+import org.scalatest.junit.JUnitRunner
+
+import scala.util.Random
+
+@RunWith(classOf[JUnitRunner])
+class SVMModelTest extends FunSuite with SharedSparkTestContext {
+
+  override def createSparkContext: SparkContext = new SparkContext("local[*]", "test-local", conf = defaultSparkConf)
+
+  test("Should score the same regression value.") {
+    val sqlContext = sqlc
+    import sqlContext.implicits._
+    val h2oContext = hc
+    import h2oContext.implicits._
+
+    // Generate random training data
+    val trainRDD = sc.parallelize(1 to 50, 1).map(v => {
+      val values = Array.fill(5){0}.map(x => Random.nextDouble())
+      //val label = Math.round(Random.nextDouble())
+      val label = if (Math.round(Random.nextDouble()) > 0.5) "1" else "0"
+      (label, Vectors.dense(values))
+    }).cache()
+    val trainDF = trainRDD.toDF("Label", "Vector")
+
+    val trainFrame = hc.asH2OFrame(trainDF, "bubbles")
+    trainFrame.replace(0, trainFrame.vec(0).toCategoricalVec).remove()
+
+    val initialWeights = Vectors.dense(1, 1, 1, 1, 1)
+    val weightsDF = sc.parallelize(Array(Tuple1(initialWeights))).toDF("Vector")
+    val weightsFrame = hc.asH2OFrame(weightsDF, "weights")
+
+    // Learning parameters
+    val parms = new SVMParameters
+    parms._train = trainFrame
+    parms._response_column = 'Label
+    parms._initial_weights = weightsFrame
+
+    val svm = new SVM(parms, h2oContext)
+
+    // Train model
+    val h2oSVMModel: SVMModel = svm.trainModel.get
+
+    val sparkSVMModel = new classification.SVMModel(
+      Vectors.dense(h2oSVMModel.output.weights),
+      h2oSVMModel.output.interceptor
+    )
+
+    // Make sure both scoring methods return the same results
+    val h2oPredsVec = h2oSVMModel.score(trainFrame).vec(0)
+    val h2oPreds = (0 until 50).map(h2oPredsVec.at(_)).toArray
+    val sparkPreds = sparkSVMModel.predict(trainRDD.map(_._2)).collect()
+
+    assert( h2oPreds.sameElements(sparkPreds) )
+
+  }
+
+}