During this presentation, after walking through a few ways to use MLflow on Azure directly, we'll cover how upcoming solutions from our group leverage MLflow for core functionality. BenchML is a new repository that aims to provide consumers of prebuilt ML endpoints visibility into the performance of each public offering for a given dataset as well as comparing results across multiple offerings. Using MLflow, BenchML is able to remain cloud-agnostic and offer a delightful local experience while leveraging the aforementioned integration to provide Azure users with a fully managed experience.
Speaker Bio: Akshaya is an engineer in the AI Platform at Microsoft, having released both GA versions of Azure Machine Learning over the years and the OSS repo MMLSpark. As the recent version of Azure ML pivoted to become more of an open platform rather than a managed product, his focus has shifted outward for open-source platform definitions for cloud-scale implementations and focused on MLflow for the Azure ML managed tracking store.
This talk was presented at the Bay Area MLflow Meetup at Databricks HQs in San Francisco: https://www.meetup.com/Bay-Area-MLflow/events/266614106/
Unveiling Design Patterns: A Visual Guide with UML Diagrams
MLflow on and inside Azure
1.
2.
3. Cloud
Azure ML
Azure Databricks
SQL Server Big Data
Clusters
Edge
Azure Cognitive Services
Containers
SQL Database Edge
Frameworks
ONNX
MLflow
4.
5. Machine Learning Accelerate the end-to-end machine
learning lifecycle
Databricks Best destination for big data analytics and
AI with Apache Spark
SQL Server
Deploy scalable clusters of SQL Server,
Spark, and HDFS containers running on
Kubernetes
6. Built with your needs in mind
Support for open source frameworks
Managed compute
MLOps (DevOps for machine learning)
Simple deployment
Automated machine learning
Seamlessly integrated with the Azure Portfolio
Boost your data science
productivity
Increase your rate of
experimentation
Deploy and manage your
models everywhere Tool agnostic Python SDK
7. Experiments
Local machine
Virtual machine
Azure ML Compute
Azure Databricks
ML Tracking and Model
Deployment
Azure Machine
Learning
Experiments and
Metrics Tracking
Metrics Artifacts
Logging API
Tracking URI
8.
9. • v1.0.0 released
ONNX
Runtime
• ONNX flavor support, discussed
at the previous meetup
MLflow +
ONNX
• Preview, Native PREDICT with
ONNX models
SQL Database
Edge
• Azure ML + Cognitive Services
• Custom Vision as ONNX
Exportable
Containers
10.
11.
12.
13.
14.
15. Measure Quality
Platform for measuring quality of
cognitive services solving the same
problem across vendors
Support major evaluation metrics for
each service
Offer Reproducibility
Track and reproduce these results
over time
Adopt industry-standard, open
source implementations
Answer Common Questions
How well does each Cognitive
Service work against my dataset?
What is the best choice of
parameters, thresholds and
normalizations for my scenario?
Explain differences between our
results and those of our competitors
Azure Machine Learning Services empowers you to bring AI to everyone with an end-to-end, scalable, trusted platform.
Boost your data science productivity
Python pip-installable extensions for Azure Machine Learning that enable data scientists to build and deploy machine learning and deep learning models
Now available for Computer Vision, Text Analytics and Time-Series Forecasting.
Increase your rate of experimentation
Rapidly prototype on your desktop, then easily scale up on virtual machines or scale out using Spark clusters
Proactively manage model performance, identify the best model, and promote it using data-driven insights
Collaborate and share solutions using popular Git repositories.
Deploy and manage your models everywhere
Use Docker containers to deploy models into production faster in the cloud, on-premises, or at the edge
Promote your best performing models into production and retrain them when their performance degrades
Azure Machine Learning Services are built with your needs in mind, providing:
GPU-enabled virtual machines
Low-latency predictions at scale
Integration with popular Python IDEs
Role-based access controls
Model versioning
Automated model retraining
(Optional: other services)
Azure Machine Learning Workbench integrates with ONNX models
Work with your ONNX models from Visual Studio Code Tools for AI.
Build deep learning models and call services straight from your favorite IDE easier with Azure Machine Learning services built right in.
Create a seamless developer experience across desktop, cloud, or at the edge.
AI Toolkit for Azure IoT Edge
MMLSpark is an open-source Spark package that enables you to quickly create powerful, highly-scalable predictive and analytical models for large image and text datasets by using deep learning and data science tools for Apache Spark.
Azure Machine Learning Services seamlessly integrates with the rest of the Azure portfolio.
<Transition>: Azure Machine Learning Services allows you to deploy models to many different production environments.