Update docs and examples to reflect model_uri changes (mlflow#1254)

* Change 'model_path' param to 'model_uri'

* Flower classifier updates for UDF evaluation

* Fixes

docs/source/models.rst

@@ -535,7 +535,7 @@ accepts the following data formats as input:
                                        exist_okay=True)
 
     # Build an Azure ML container image for deployment
-    azure_image, azure_model = mlflow.azureml.build_image(model_path="<path-to-model>",
+    azure_image, azure_model = mlflow.azureml.build_image(model_uri="<path-to-model>",
                                                           workspace=azure_workspace,
                                                           description="Wine regression model 1",
                                                           synchronous=True)

examples/flower_classifier/README.rst

@@ -98,14 +98,14 @@ run_id ``101``.
   .. code-block:: bash
 
       # score the deployed model
-      python score_images_rest.py --port 54321 http://127.0.0.1 ./my_images_to_score
+      python score_images_rest.py --model-uri runs:/101/model --port 54321 http://127.0.0.1 --data-path /path/to/images/for/scoring
 
 
 - To test batch scoring in Spark, run score_images_spark.py to score the model in Spark like this:
 
 .. code-block:: bash
 
-   python score_images_spark.py ./my_images_to_score model --run-id 101
+   python score_images_spark.py --model-uri runs:/101/model --data-path /path/to/images/for/scoring
 
 
 

examples/flower_classifier/score_images_rest.py

@@ -11,6 +11,8 @@
 import click
 import pandas as pd
 
+from mlflow.utils import cli_args
+
 
 def score_model(path, uri, port):
     """
@@ -48,14 +50,14 @@ def read_image(x):
 
 @click.command(help="Score images.")
 @click.option("--port", type=click.INT, default=80, help="Port at which the model is deployed.")
-@click.argument("model_uri")
-@click.argument("input_data_path")
-def run(input_data_path, model_uri, port):
+@cli_args.MODEL_URI
+@click.argument("--data-path", "-d")
+def run(data_path, model_uri, port):
     """
     Score images with MLflow deployed deployed at given uri and port and print out the response
     to standard out.
     """
-    print(score_model(input_data_path, model_uri, port).text)
+    print(score_model(data_path, model_uri, port).text)
 
 
 if __name__ == '__main__':

examples/flower_classifier/score_images_spark.py

@@ -16,6 +16,7 @@
 
 import mlflow
 import mlflow.pyfunc
+from mlflow.utils import cli_args
 
 from pyspark.sql.types import *
 from pyspark.sql.types import Row
@@ -36,24 +37,23 @@ def read_images(spark, filenames):
                                            image=read_image_bytes_base64(x))).toDF(schema=schema)
 
 
-def score_model(spark, data_path, model_path, model_run_id=None):
+def score_model(spark, data_path, model_uri):
     if os.path.isdir(data_path):
         filenames = [os.path.abspath(os.path.join(data_path, x)) for x in os.listdir(data_path)
                      if os.path.isfile(os.path.join(data_path, x))]
     else:
         filenames = [data_path]
 
     image_classifier_udf = mlflow.pyfunc.spark_udf(spark=spark,
-                                                   path=model_path,
-                                                   run_id=model_run_id,
+                                                   model_uri=model_uri,
                                                    result_type=ArrayType(StringType()))
 
     image_df = read_images(spark, filenames)
 
     raw_preds = image_df.withColumn("prediction", image_classifier_udf("image")).select(
         ["filename", "prediction"]).toPandas()
     # load the pyfunc model to get our domain
-    pyfunc_model = mlflow.pyfunc.load_pyfunc(model_path, run_id=model_run_id)
+    pyfunc_model = mlflow.pyfunc.load_pyfunc(model_uri=model_uri)
     preds = pd.DataFrame(raw_preds["filename"], index=raw_preds.index)
     preds[pyfunc_model._column_names] = pd.DataFrame(raw_preds['prediction'].values.tolist(),
                                                      columns=pyfunc_model._column_names,
@@ -68,18 +68,17 @@ def score_model(spark, data_path, model_path, model_run_id=None):
 
 
 @click.command(help="Score images.")
-@click.option("--run-id", type=click.STRING, default=None, help="MLflow run id")
-@click.argument("data_path")
-@click.argument("model_path")
-def run(data_path, model_path, run_id):
+@cli_args.MODEL_URI
+@click.argument("--data-path", "-d")
+def run(model_uri, data_path):
     with pyspark.sql.SparkSession.builder \
             .config(key="spark.python.worker.reuse", value=True) \
             .config(key="spark.ui.enabled", value=False) \
             .master("local-cluster[2, 1, 1024]") \
             .getOrCreate() as spark:
         # ignore spark log output
         spark.sparkContext.setLogLevel("OFF")
-        print(score_model(spark, data_path, model_path, run_id))
+        print(score_model(spark, data_path, model_uri))
 
 
 if __name__ == '__main__':