Add MLFlow model generator && serving it

Jenkinsfile_train

@@ -20,8 +20,14 @@ pipeline {
                 sh "python3 sacred_exp.py"
             }
         }
+        stage("Run MLFlow training"){
+            steps {
+                sh "python3 ml_model.py ${verbose} ${epochs}"
+            }
+        }
         stage('Save trained model files') {
 			steps{
+                archiveArtifacts 'movies_on_streaming_platforms_model/**'
                 archiveArtifacts 'sacred_file/**'
 				archiveArtifacts 'linear_regression.h5'
 			}

ml_model.py

@@ -0,0 +1,54 @@
+import pandas as pd
+import numpy as np
+import tensorflow as tf
+import os.path
+
+import mlflow
+import sys
+
+from tensorflow import keras
+from tensorflow.keras import layers
+from tensorflow.keras.layers.experimental import preprocessing
+
+arguments = sys.argv[1:]
+
+verbose = int(arguments[0])
+epochs = int(arguments[1])
+
+ # Wczytanie danych
+train_data = pd.read_csv("./MoviesOnStreamingPlatforms_updated.train")
+test_data = pd.read_csv("./MoviesOnStreamingPlatforms_updated.test")
+
+# Stworzenie modelu 
+columns_to_use = ['Year', 'Runtime', 'Netflix']
+train_X = tf.convert_to_tensor(train_data[columns_to_use])
+train_Y = tf.convert_to_tensor(train_data[["IMDb"]])
+
+test_X = tf.convert_to_tensor(test_data[columns_to_use])
+test_Y = tf.convert_to_tensor(test_data[["IMDb"]])
+
+normalizer = preprocessing.Normalization(input_shape=[3,])
+normalizer.adapt(train_X)
+
+model = keras.Sequential([
+    keras.Input(shape=(len(columns_to_use),)),
+    normalizer,
+    layers.Dense(30, activation='relu'),
+    layers.Dense(10, activation='relu'),
+    layers.Dense(25, activation='relu'),
+    layers.Dense(1)
+])
+
+model.compile(loss='mean_absolute_error',
+                optimizer=tf.keras.optimizers.Adam(0.001),
+                metrics=[tf.keras.metrics.RootMeanSquaredError()])
+
+model.fit(train_X, train_Y, verbose=verbose, epochs=epochs)
+
+signature = mlflow.models.signature.infer_signature(train_X.numpy(), model.predict(train_X.numpy()))
+input_data = test_X
+
+with mlflow.start_run():
+    mlflow.keras.save_model(model, "movies_on_streaming_platforms_model", input_example=input_data.numpy(), signature=signature)
+
+

mlruns/0/724dc0d672664057b760fe5f18801036/meta.yaml

@@ -0,0 +1,15 @@
+artifact_uri: file:///Volumes/seagate/ium_434704/mlruns/0/724dc0d672664057b760fe5f18801036/artifacts
+end_time: 1621772380826
+entry_point_name: ''
+experiment_id: '0'
+lifecycle_stage: active
+name: ''
+run_id: 724dc0d672664057b760fe5f18801036
+run_uuid: 724dc0d672664057b760fe5f18801036
+source_name: ''
+source_type: 4
+source_version: ''
+start_time: 1621772379167
+status: 3
+tags: []
+user_id: wj

mlruns/0/724dc0d672664057b760fe5f18801036/tags/mlflow.source.git.commit

@@ -0,0 +1 @@
+b82917a5d90e071de9b67f2b6648be0353c54b62

mlruns/0/724dc0d672664057b760fe5f18801036/tags/mlflow.source.name

@@ -0,0 +1 @@
+ml_model.py

mlruns/0/724dc0d672664057b760fe5f18801036/tags/mlflow.source.type

@@ -0,0 +1 @@
+LOCAL

mlruns/0/724dc0d672664057b760fe5f18801036/tags/mlflow.user

@@ -0,0 +1 @@
+wj

movies_on_streaming_platforms_model/MLmodel

@@ -0,0 +1,20 @@
+flavors:
+  keras:
+    data: data
+    keras_module: tensorflow.keras
+    keras_version: 2.5.0
+    save_format: tf
+  python_function:
+    data: data
+    env: conda.yaml
+    loader_module: mlflow.keras
+    python_version: 3.9.1
+saved_input_example_info:
+  artifact_path: input_example.json
+  format: tf-serving
+  type: ndarray
+signature:
+  inputs: '[{"type": "tensor", "tensor-spec": {"dtype": "float64", "shape": [-1, 3]}}]'
+  outputs: '[{"type": "tensor", "tensor-spec": {"dtype": "float32", "shape": [-1,
+    1]}}]'
+utc_time_created: '2021-05-23 12:19:39.663838'

movies_on_streaming_platforms_model/conda.yaml

@@ -0,0 +1,10 @@
+channels:
+- defaults
+- conda-forge
+dependencies:
+- python=3.9.1
+- pip
+- pip:
+  - mlflow
+  - tensorflow==2.5.0-rc1
+name: mlflow-env

movies_on_streaming_platforms_model/data/keras_module.txt

@@ -0,0 +1 @@
+tensorflow.keras

movies_on_streaming_platforms_model/data/save_format.txt

@@ -0,0 +1 @@
+tf