diff --git a/dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/tensorflow/Learn.scala b/dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/tensorflow/Learn.scala
index 0b8ee1045..b9c3862d1 100644
--- a/dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/tensorflow/Learn.scala
+++ b/dynaml-core/src/main/scala-2.11/io/github/mandar2812/dynaml/tensorflow/Learn.scala
@@ -21,7 +21,7 @@ package io.github.mandar2812.dynaml.tensorflow
 import _root_.io.github.mandar2812.dynaml.pipes.DataPipe
 import _root_.io.github.mandar2812.dynaml.models.TFModel
 import io.github.mandar2812.dynaml.tensorflow.layers.{DynamicTimeStepCTRNN, FiniteHorizonCTRNN, FiniteHorizonLinear}
-import org.platanios.tensorflow.api.learn.{Mode, StopCriteria}
+import org.platanios.tensorflow.api.learn.{ClipGradients, Mode, NoClipGradients, StopCriteria}
 import org.platanios.tensorflow.api.learn.layers.{Input, Layer}
 import org.platanios.tensorflow.api.ops.NN.SameConvPadding
 import org.platanios.tensorflow.api.ops.io.data.Dataset
@@ -492,7 +492,8 @@ private[tensorflow] object Learn {
     stopCriteria: StopCriteria,
     stepRateFreq: Int = 5000,
     summarySaveFreq: Int = 5000,
-    checkPointFreq: Int = 5000)(
+    checkPointFreq: Int = 5000,
+    clipGradients: ClipGradients = NoClipGradients)(
     training_data: Dataset[
       (IT, TT), (IO, TO),
       (ID, TD), (IS, TS)],
@@ -503,7 +504,8 @@ private[tensorflow] object Learn {
       val model = tf.learn.Model.supervised(
         input, architecture,
         target, processTarget,
-        loss, optimizer)
+        loss, optimizer,
+        clipGradients)
 
       println("\nTraining model.\n")
 
@@ -597,9 +599,10 @@ private[tensorflow] object Learn {
     optimizer: Optimizer,
     summariesDir: java.nio.file.Path,
     stopCriteria: StopCriteria,
-    stepRateFreq: Int,
-    summarySaveFreq: Int,
-    checkPointFreq: Int)(
+    stepRateFreq: Int = 5000,
+    summarySaveFreq: Int = 5000,
+    checkPointFreq: Int = 5000,
+    clipGradients: ClipGradients = NoClipGradients)(
     training_data: Dataset[IT, IO, ID, IS],
     inMemory: Boolean): UnsupModelPair[IT, IO, ID, IS, I] = {