# MLFlow

### Adding an MLFlow app

To add MLflow to your working instance, open **Applications** and select **MLFlow 2.13 + Jupyter**. Start the application, and MLflow opens in a new **JupyterLab editor tab**. If you close that tab, reopen MLflow from the **JupyterLab Command Palette** with `Command/Ctrl+Shift+C`, then run the MLflow command shown there.

### MLFlow in Visual Studio Code

MLflow is also available in the VS Code-based application. To use it, open the VS Code app that includes MLflow, then open the Command Palette and run the MLflow command shown on the page to open MLflow in a new VS Code tab.

### Tracking model training

{% hint style="info" %}
MLFlow server runs on port <http://127.0.0.1:8080>, you will need to set this tracking server in your code with:

`import mlflow`\
`mlflow.set_tracking_uri("http://localhost:8080")`
{% endhint %}

The following tutorial, adapted from the MLFlow documentation, shows how to track model training and register the trained model with MLFlow on Nuvolos:

<pre class="language-python"><code class="lang-python"><strong># The data set used in this example is from http://archive.ics.uci.edu/ml/datasets/Wine+Quality
</strong># P. Cortez, A. Cerdeira, F. Almeida, T. Matos and J. Reis.
# Modeling wine preferences by data mining from physicochemical properties. In Decision Support Systems, Elsevier, 47(4):547-553, 2009.

import os
import warnings
import sys

import pandas as pd
import numpy as np
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score
from sklearn.model_selection import train_test_split
from sklearn.linear_model import ElasticNet
from urllib.parse import urlparse
import mlflow
import mlflow.sklearn

import logging

logging.basicConfig(level=logging.WARN)
logger = logging.getLogger(__name__)


def eval_metrics(actual, pred):
    rmse = np.sqrt(mean_squared_error(actual, pred))
    mae = mean_absolute_error(actual, pred)
    r2 = r2_score(actual, pred)
    return rmse, mae, r2


if __name__ == "__main__":
    mlflow.set_tracking_uri("http://localhost:8080")
    mlflow.set_experiment("Wine Quality")
    warnings.filterwarnings("ignore")
    np.random.seed(40)

    # Read the wine-quality csv file from the URL
    csv_url = (
        "http://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/winequality-red.csv"
    )
    try:
        data = pd.read_csv(csv_url, sep=";")
    except Exception as e:
        logger.exception(
            "Unable to download training &#x26; test CSV, check your internet connection. Error: %s", e
        )

    # Split the data into training and test sets. (0.75, 0.25) split.
    train, test = train_test_split(data)

    # The predicted column is "quality" which is a scalar from [3, 9]
    train_x = train.drop(["quality"], axis=1)
    test_x = test.drop(["quality"], axis=1)
    train_y = train[["quality"]]
    test_y = test[["quality"]]

    alpha = float(sys.argv[1]) if len(sys.argv) > 1 else 0.5
    l1_ratio = float(sys.argv[2]) if len(sys.argv) > 2 else 0.5

    with mlflow.start_run():
        lr = ElasticNet(alpha=alpha, l1_ratio=l1_ratio, random_state=42)
        lr.fit(train_x, train_y)

        predicted_qualities = lr.predict(test_x)

        (rmse, mae, r2) = eval_metrics(test_y, predicted_qualities)

        print("Elasticnet model (alpha=%f, l1_ratio=%f):" % (alpha, l1_ratio))
        print("  RMSE: %s" % rmse)
        print("  MAE: %s" % mae)
        print("  R2: %s" % r2)

        mlflow.log_param("alpha", alpha)
        mlflow.log_param("l1_ratio", l1_ratio)
        mlflow.log_metric("rmse", rmse)
        mlflow.log_metric("r2", r2)
        mlflow.log_metric("mae", mae)

        tracking_url_type_store = urlparse(mlflow.get_tracking_uri()).scheme

        # Model registry does not work with file store
        if tracking_url_type_store != "file":

            # Register the model
            # There are other ways to use the Model Registry, which depends on the use case,
            # please refer to the doc for more information:
            # https://mlflow.org/docs/latest/model-registry.html#api-workflow
            mlflow.sklearn.log_model(lr, "model", registered_model_name="ElasticnetWineModel")
        else:
            mlflow.sklearn.log_model(lr, "model")
</code></pre>