diff --git a/evaluation.Jenkinsfile b/evaluation.Jenkinsfile
new file mode 100644
index 0000000..722f2dc
--- /dev/null
+++ b/evaluation.Jenkinsfile
@@ -0,0 +1,23 @@
+pipeline {
+    agent { 
+        dockerfile true 
+    }
+
+    stages {
+        stage('Build') {
+            steps {
+                git 'https://git.wmi.amu.edu.pl/s434749/ium_434749.git'
+                copyArtifacts fingerprintArtifacts: true, projectName: 's434749-training', selector: lastSuccessful()
+                sh 'python3 train_model.py eval'
+            }
+
+            post {
+                success {
+                    def results = readFile "${env.WORKSPACE}/results.txt"
+                    emailext body: 'Evaluation of CNN for english phonetic embeddings has finished successfully!\n'+results, subject: 's434749 evaluation finished', to: '26ab8f35.uam.onmicrosoft.com@emea.teams.ms'
+                    archiveArtifacts 'results.txt'
+                }
+            }
+        }
+    }
+}
diff --git a/train-model.Jenkinsfile b/train-model.Jenkinsfile
index cdbb94e..c52601f 100644
--- a/train-model.Jenkinsfile
+++ b/train-model.Jenkinsfile
@@ -8,12 +8,13 @@ pipeline {
             steps {
                 git 'https://git.wmi.amu.edu.pl/s434749/ium_434749.git'
                 copyArtifacts fingerprintArtifacts: true, projectName: 's434749-create-dataset', selector: lastSuccessful()
-                sh 'python3 train_model.py'
+                sh 'python3 train_model.py train'
             }
 
             post {
                 success {
-                    archiveArtifacts 'cnn.pth,results.txt'
+                    emailext body: 'Training of CNN for english phonetic embeddings has finished successfully', subject: 's434749 training finished', to: '26ab8f35.uam.onmicrosoft.com@emea.teams.ms'
+                    archiveArtifacts 'cnn.pth'
                 }
             }
         }
diff --git a/train_model.py b/train_model.py
index 5b6a642..52e847d 100644
--- a/train_model.py
+++ b/train_model.py
@@ -13,6 +13,7 @@ from torch.utils.data import Dataset, DataLoader
 import re
 import random
 import os
+import sys
 from tqdm import tqdm
 from Levenshtein import distance as levenshtein_distance
 
@@ -167,16 +168,20 @@ def evaluate_monte_carlo(model, repeats):
             outer_bar.update(1)
         with open('results.txt', 'w+') as r:
             print("Average estimation error " + str(diff.item() / i))
-            r.write("Average estimation error " + str(diff.item() / i)+"\n")
+            r.write("Average estimation error " + str(diff.item() / i) + "\n")
 
 
 cnn = CNN(kernel_size=3, hidden_layers=14, channels=MAX_LEN, embedding_size=MAX_LEN).to(DEVICE)
 if os.path.isfile('cnn.pth'):
     cnn.load_state_dict(torch.load('cnn.pth', map_location=torch.device('cpu')))
 else:
-    train_model(cnn)
-    torch.save(cnn.state_dict(), 'cnn.pth')
+    if len(sys.argv) > 1 and sys.argv[1] == 'train':
+        train_model(cnn)
+        torch.save(cnn.state_dict(), 'cnn.pth')
+    else:
+        print("cnn.pth missing!")
+        exit(2)
 
-cnn.eval()
-print("Training finished! Starting evaluation!")
-evaluate_monte_carlo(cnn, 1)
+if len(sys.argv) > 1 and sys.argv[1] == 'eval':
+    cnn.eval()
+    evaluate_monte_carlo(cnn, 1)