diff --git a/Jenkinsfile_train b/Jenkinsfile_train
new file mode 100644
index 0000000..5d95677
--- /dev/null
+++ b/Jenkinsfile_train
@@ -0,0 +1,35 @@
+pipeline {
+    agent { 
+        dockerfile  true
+    }
+    parameters {
+        epochs(
+            defaultValue: '200',
+            description: 'number of epochs',
+            name: 'EPOCH'
+             )
+    }
+
+    stages {
+        stage('Stage 1') {
+            steps {
+                echo 'Hello world!' 
+            }
+        }
+
+        stage('Copy from different Pipeline') {
+			steps {
+                copyArtifacts fingerprintArtifacts: true, projectName: 's444517-create-dataset', selector: lastSuccessful()
+			}
+		}
+
+        stage('Get data save artifacts') {
+            steps {
+                    sh 'python3 ./nn_train.py $EPOCH'
+                    archiveArtifacts artifacts: 'my_model/saved_model.pb'
+                
+            }
+        }
+    }
+}
+
diff --git a/nn_train.py b/nn_train.py
index 6f58717..8f4331b 100644
--- a/nn_train.py
+++ b/nn_train.py
@@ -1,6 +1,7 @@
 
 import pandas as pd
 import numpy as np
+import sys
 
 from tensorflow.keras.models import Sequential
 from tensorflow.keras.layers import Dense
@@ -52,13 +53,13 @@ model = Sequential()
 model.add(Dense(number_of_classes, activation='relu'))
 model.add(Dense(number_of_classes, activation='softmax',input_dim=number_of_features))
 model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy', 'categorical_accuracy'])
-model.fit(x_train_set, dummy_y, epochs=200, validation_data=(x_validate_set, dummy_yv))
-#model.save("my_model/")
+model.fit(x_train_set, dummy_y, epochs=int(sys.argv[1]), validation_data=(x_validate_set, dummy_yv))
+model.save("my_model/")
 
 #model predictions
 #model = keras.models.load_model('my_model')
 yhat = model.predict(x_test_set)
 f = open("results.txt", "w")
 for numerator, single_pred in enumerate(yhat):
-    f.write(f"PREDICTED: {sorted(y_class_names)[np.argmax(single_pred)]}, ACTUAL: {y_test_set[numerator]} {sorted(y_class_names)[np.argmax(single_pred)] == y_test_set[numerator]}\n")
+    f.write(f"{sorted(y_class_names)[np.argmax(single_pred)]},{y_test_set[numerator]}\n")
 f.close()