Model Serving with KFServing and Tensorflow - MNIST Classification

INPUT –> MODEL –> PREDICTION

This notebook requires KFServing to be installed

NOTE: It is assumed that a model called mnist is already available in Hopsworks. An example of training a model for the MNIST handwritten digit classification problem is available in Jupyter/experiment/Tensorflow/mnist.ipynb

Model Serving on Hopsworks

The hops python library

hops is a helper library for Hops that facilitates development by hiding the complexity of running applications and iteracting with services.

Have a feature request or encountered an issue? Please let us know on github.

Serve the MNIST classifier

Check Model Repository for best model based on accuracy

Query Model Repository for best mnist Model

import hsml

conn = hsml.connection()
mr = conn.get_model_registry()

MODEL_NAME="mnist_e2e"
EVALUATION_METRIC="accuracy"

Connected. Call `.close()` to terminate connection gracefully.

best_model = mr.get_best_model(MODEL_NAME, EVALUATION_METRIC, "max")

print('Model name: ' + best_model.name)
print('Model version: ' + str(best_model.version))
print(best_model.training_metrics)

Model name: mnist_e2e
Model version: 1
{'accuracy': '0.625'}

Create Model Serving of Exported Model

from hops import serving

# Create serving instance
SERVING_NAME = "mniste2ekf"

response = serving.create_or_update(SERVING_NAME, # define a name for the serving instance
                                    model_path=best_model.model_path, # set the path of the model to be deployed
                                    model_server="TENSORFLOW_SERVING", # set the model server to run the model
                                    kfserving=True, # whether to serve the model using KFServing or the default tool in the current Hopsworks version
                                    # optional arguments
                                    model_version=best_model.version, # set the version of the model to be deployed
                                    topic_name="CREATE", # (optional) set the topic name or CREATE to create a new topic for inference logging
                                    inference_logging="ALL", # with KFServing, select the type of inference data to log into Kafka, e.g MODEL_INPUTS, PREDICTIONS or ALL
                                    instances=1, # with KFServing, set 0 instances to leverage scale-to-zero capabilities
                                    )

2022-01-28 11:52:48,685 INFO: Serving mniste2ekf successfully created

# List all available servings in the project
for s in serving.get_all():
    print(s.name)

mniste2ekf

# Get serving status
serving.get_status(SERVING_NAME)

'Stopped'

Classify digits with the MNIST classifier

Start Model Serving Server

if serving.get_status(SERVING_NAME) == 'Stopped':
    serving.start(SERVING_NAME)

2022-01-28 11:52:49,378 INFO: Serving with name: mniste2ekf successfully started

import time
while serving.get_status(SERVING_NAME) != "Running":
    time.sleep(5) # Let the serving startup correctly
time.sleep(10)

Check Model Serving for active servings

Send Prediction Requests to the Served Model using Hopsworks REST API

import json
import numpy as np

NUM_FEATURES=784

for i in range(10):
    data = {
                "signature_name": "serving_default", "instances": [np.random.rand(NUM_FEATURES).tolist()]
            }
    response = serving.make_inference_request(SERVING_NAME, data)
    print(response)

{'predictions': [[0.00228863931, 0.039052885, 0.299541056, 0.0292527322, 0.210720837, 0.00816287566, 0.0126433428, 0.0109003522, 0.344060391, 0.0433768257]]}
{'predictions': [[0.00586040691, 0.0307841301, 0.33769837, 0.0337129459, 0.254304558, 0.0237869881, 0.0141318291, 0.0193788894, 0.238279134, 0.0420627445]]}
{'predictions': [[0.00538377091, 0.061361514, 0.238303721, 0.0285069272, 0.501042247, 0.0104929302, 0.0204542708, 0.0179179348, 0.0851745605, 0.0313620083]]}
{'predictions': [[0.00573793799, 0.029261997, 0.369869381, 0.0178442206, 0.450838357, 0.00611973461, 0.00844837539, 0.0112320511, 0.079323791, 0.0213241577]]}
{'predictions': [[0.0119106388, 0.0629702583, 0.153047368, 0.0336705893, 0.3950378, 0.00748524442, 0.0267426614, 0.0312577672, 0.249878481, 0.0279991794]]}
{'predictions': [[0.0111142406, 0.0491779149, 0.204911619, 0.0337088816, 0.241256326, 0.0222033393, 0.0126153091, 0.0414450057, 0.302034378, 0.0815329328]]}
{'predictions': [[0.00563503848, 0.0279151835, 0.28025192, 0.0385764167, 0.363024831, 0.0126688043, 0.0207744408, 0.0177768413, 0.193134412, 0.0402421467]]}
{'predictions': [[0.00543945516, 0.126275226, 0.163412824, 0.04192359, 0.265003413, 0.0151998475, 0.0343197025, 0.0221705418, 0.241572261, 0.0846831948]]}
{'predictions': [[0.00477044843, 0.0468725786, 0.208328143, 0.0335769281, 0.347728401, 0.0162629858, 0.0180092286, 0.01045975, 0.283724517, 0.0302670337]]}
{'predictions': [[0.00746631622, 0.0523283072, 0.304017246, 0.027453661, 0.280292451, 0.0179669447, 0.0197735652, 0.0237298273, 0.174194977, 0.0927767605]]}

Monitor Prediction Requests and Responses using Kafka

from hops import kafka
from confluent_kafka import Producer, Consumer, KafkaError

Setup Kafka consumer and subscribe to the topic containing the prediction logs

TOPIC_NAME = serving.get_kafka_topic(SERVING_NAME)

config = kafka.get_kafka_default_config()
config['default.topic.config'] = {'auto.offset.reset': 'earliest'}
consumer = Consumer(config)
topics = [TOPIC_NAME]
consumer.subscribe(topics)

Read the Kafka Avro schema from Hopsworks and setup an Avro reader

json_schema = kafka.get_schema(TOPIC_NAME)
avro_schema = kafka.convert_json_schema_to_avro(json_schema)

Read messages from the Kafka topic, parse them with the Avro schema and print the results

PRINT_INSTANCES=False
PRINT_PREDICTIONS=True

for i in range(0, 10):
    msg = consumer.poll(timeout=5.0)
    if msg is not None:
        value = msg.value()
        try:
            event_dict = kafka.parse_avro_msg(value, avro_schema)  
            payload = json.loads(event_dict["payload"])
            
            if (event_dict['messageType'] == "request" and not PRINT_INSTANCES) or \
                (event_dict['messageType'] == "response" and not PRINT_PREDICTIONS):
                continue
            
            print("INFO -> servingId: {}, modelName: {}, modelVersion: {},"\
                  "requestTimestamp: {}, inferenceId:{}, messageType:{}".format(
                       event_dict["servingId"],
                       event_dict["modelName"],
                       event_dict["modelVersion"],
                       event_dict["requestTimestamp"],
                       event_dict["inferenceId"],
                       event_dict["messageType"]))

            if event_dict['messageType'] == "request":
                print("Instances -> {}\n".format(payload['instances']))
                
            if event_dict['messageType'] == "response":
                print("Predictions -> {}\n".format(payload['predictions']))

        except Exception as e:
            print("A message was read but there was an error parsing it")
            print(e)
    else:
        print("timeout.. no more messages to read from topic")

INFO -> servingId: 33, modelName: mnist_e2e, modelVersion: 1,requestTimestamp: 1643370799, inferenceId:8fe0ab82-a9fd-424e-b154-b88eaeb82312, messageType:response
Predictions -> [[0.00531423138, 0.0495751873, 0.215446427, 0.0254741, 0.456469536, 0.00679346081, 0.0227076765, 0.0156853292, 0.151415035, 0.0511190034]]

INFO -> servingId: 33, modelName: mnist_e2e, modelVersion: 1,requestTimestamp: 1643370799, inferenceId:2eb8f1b7-5fe5-4a15-921e-1077b3bc5548, messageType:response
Predictions -> [[0.015544543, 0.04791224, 0.231773868, 0.0455554053, 0.352704018, 0.0196769126, 0.023360543, 0.0201557297, 0.190379784, 0.0529369563]]

INFO -> servingId: 33, modelName: mnist_e2e, modelVersion: 1,requestTimestamp: 1643370799, inferenceId:2a85e0ed-70a4-43b3-8afd-43fdb5bb888c, messageType:response
Predictions -> [[0.0119756339, 0.108432077, 0.249693975, 0.0291424617, 0.297212154, 0.0278416723, 0.0237392467, 0.0305440109, 0.182935417, 0.038483344]]

INFO -> servingId: 33, modelName: mnist_e2e, modelVersion: 1,requestTimestamp: 1643370799, inferenceId:92f62e3a-f05e-442f-91c2-7f8de7d80de9, messageType:response
Predictions -> [[0.0061565917, 0.0573133416, 0.157412216, 0.0603140071, 0.319210559, 0.01605873, 0.018614579, 0.0371795483, 0.204262689, 0.123477831]]

INFO -> servingId: 33, modelName: mnist_e2e, modelVersion: 1,requestTimestamp: 1643370800, inferenceId:4c50013c-834d-4704-9808-6a6cb25dbb95, messageType:response
Predictions -> [[0.011226072, 0.0487205796, 0.279118955, 0.0241690688, 0.347856402, 0.0182868, 0.0276911054, 0.0257906299, 0.145419195, 0.0717211664]]