sacred

.env

@@ -0,0 +1,5 @@
+MONGO_INITDB_ROOT_USERNAME=admin
+MONGO_INITDB_ROOT_PASSWORD=IUM_2021
+ME_CONFIG_BASICAUTH_USERNAME=mongo_express_user
+ME_CONFIG_BASICAUTH_PASSWORD=mongo_express_pw
+MONGO_DATABASE=sacred

requirements.txt

@@ -5,3 +5,5 @@ numpy
 matplotlib
 mlflow
 dvc
+sacred
+pymongo

sacredboard/Dockerfile

@@ -0,0 +1,5 @@
+FROM python:3.6-jessie
+
+RUN pip install https://github.com/chovanecm/sacredboard/archive/develop.zip
+
+ENTRYPOINT sacredboard -mu mongodb://$MONGO_INITDB_ROOT_USERNAME:$MONGO_INITDB_ROOT_PASSWORD@mongo:27017/?authMechanism=SCRAM-SHA-1 $MONGO_DATABASE

train_sacred.py

@@ -0,0 +1,73 @@
+import pandas as pd
+from tensorflow import keras
+from tensorflow.keras import layers
+import argparse
+from sacred import Experiment
+from sacred.observers import FileStorageObserver, MongoObserver
+
+ex = Experiment("464980", interactive=True, save_git_info=False)
+ex.observers.append(FileStorageObserver('experiments'))
+ex.observers.append(MongoObserver(url='mongodb://admin:IUM_2021@tzietkiewicz.vm.wmi.amu.edu.pl:27017',
+                                  db_name='sacred'))
+
+
+@ex.capture
+def capture_params(epochs):
+    print(f"epochs: {epochs}")
+
+
+class RegressionModel:
+    def __init__(self, optimizer="adam", loss="mean_squared_error"):
+        self.model = keras.Sequential([
+            layers.Input(shape=(5,)),  # Input layer
+            layers.Dense(32, activation='relu'),  # Hidden layer with 32 neurons and ReLU activation
+            layers.Dense(1)  # Output layer with a single neuron (for regression)
+        ])
+        self.optimizer = optimizer
+        self.loss = loss
+        self.X_train = None
+        self.X_test = None
+        self.y_train = None
+        self.y_test = None
+
+    def load_data(self, train_path, test_path):
+        data_train = pd.read_csv(train_path)
+        data_test = pd.read_csv(test_path)
+        self.X_train = data_train.drop("Performance Index", axis=1)
+        self.y_train = data_train["Performance Index"]
+        self.X_test = data_test.drop("Performance Index", axis=1)
+        self.y_test = data_test["Performance Index"]
+
+    def train(self, epochs=30):
+        self.model.compile(optimizer=self.optimizer, loss=self.loss)
+        self.model.fit(self.X_train, self.y_train, epochs=epochs, batch_size=32, validation_data=(self.X_test, self.y_test))
+        capture_params(epochs)
+
+    def predict(self, data):
+        prediction = self.model.predict(data)
+        return prediction
+
+    def evaluate(self):
+        test_loss = self.model.evaluate(self.X_test, self.y_test)
+        print(f"Test Loss: {test_loss:.4f}")
+        return test_loss
+
+    def save_model(self):
+        self.model.save("model.keras")
+        ex.add_artifact("model.keras")
+
+
+@ex.main
+def main(_run):
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--epochs')
+
+    args = parser.parse_args()
+    model = RegressionModel()
+    model.load_data("df_train.csv", "df_test.csv")
+    model.train(epochs=int(args.epochs))
+    _run.log_scalar("testing.mean_square_error", model.evaluate())
+    model.save_model()
+
+
+ex.run()

training/Jenkinsfile

@@ -37,6 +37,19 @@ pipeline {
                 archiveArtifacts artifacts: 'model.keras', onlyIfSuccessful: true
             }
         }
+        stage('Experiments') {
+            agent {
+                dockerfile {
+                    filename 'Dockerfile'
+                    reuseNode true
+                }
+            }
+            steps {
+                sh "chmod +x ./train_sacred.py"
+                sh "python ./train_sacred.py --epochs ${params.EPOCHS}"
+                archiveArtifacts artifacts: 'experiments', onlyIfSuccessful: true
+            }
+        }
         stage('Run training'){
             steps{
                 script {