Using an MLflow Model Artifact in a Model REST API

You can use MLflow to create, deploy, and manage models as REST APIs to serve predictions

1. To create an MLflow model add the following information when you run an experiment:

   mlflow.log_artifacts ("output")
   mlflow.sklearn.log_model(lr, "model")

   For example:

   # The data set used in this example is from http://archive.ics.uci.edu/ml/datasets/Wine+Quality
   # 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
   from mlflow.models import infer_signature
   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__":
       warnings.filterwarnings("ignore")
       np.random.seed(40)
   
       # Read the wine-quality csv file from the URL
       csv_url = (
           "https://raw.githubusercontent.com/mlflow/mlflow/master/tests/datasets/winequality-red.csv"
       )
       try:
           data = pd.read_csv(csv_url, sep=";")
       except Exception as e:
           logger.exception(
               "Unable to download training & 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}):".format(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)
   
           predictions = lr.predict(train_x)
           signature = infer_signature(train_x, predictions)
           mlflow.sklearn.log_model(lr, "model", signature=signature, registered_model_name="testmodel")

   In this example we are training a machine learning model using linear regression to predict wine quality. This script creates the MLflow model artifact and logs it to the model directory: /home/cdsw/.experiments/<experiment_id>/<run_id>/artifacts/models

2. To view the model, navigate to the Experiments page and select your experiment name. Cloudera AI displays the Runs page and lists all of your current runs.
3. Click the run from step 1 that created the MLflow model. Cloudera AI displays the Runs detail page
4. Click Artifacts to display a list of all the logged artifacts for the run.

   

5. Click model. Cloudera AI displays the MLflow information you use to create predictions for your experiment.