Skip to main content

Machine Learning Model Deployment

This course is designed to introduce three primary machine learning deployment strategies and illustrate the implementation of each strategy on Databricks. Following an exploration of the fundamentals of model deployment, the course delves into batch inference, offering hands-on demonstrations and labs for utilizing a model in batch inference scenarios, along with considerations for performance optimization. The second part of the course comprehensively covers pipeline deployment, while the final segment focuses on real-time deployment. Participants will engage in hands-on demonstrations and labs, deploying models with Model Serving and utilizing the serving endpoint for real-time inference.


Languages Available: English | 日本語 | Português BR | 한국어

Skill Level
Associate
Duration
4h
Prerequisites

At a minimum, you should be familiar with the following before attempting to take this content:

- Familiarity with the Databricks Data Intelligence Platform and basic workspace operations (create clusters, run code in notebooks, use basic notebook operations, import repos from git)

- Intermediate programming experience with Python, including data manipulation libraries (pandas, numpy) and working with APIs (databricks-sdk, REST endpoints)

- Basic knowledge of MLflow for experiment tracking, model logging, model registry operations, and model versioning

- Understanding of machine learning fundamentals, including model training, evaluation, batch inference, and real-time deployment concepts

- Intermediate experience with Unity Catalog for data governance and model registry management

- Basic familiarity with Feature Engineering concepts, including feature tables, feature lookups, and offline vs online feature stores

- Understanding of Delta Lake operations (create tables, perform updates, optimize files, and liquid clustering) and data storage optimization techniques

- Basic knowledge of Apache Spark and PySpark for distributed data processing and User Defined Functions (UDFs)

Outline

1. Model Deployment Fundamentals

• Model Deployment Strategies

• Model Deployment with MLflow


2. Batch Deployment 

• Introduction to Batch Deployment

• Demo: Batch Deployment

• Lab: Batch Deployment


3. Pipeline Deployment 

• Introduction to Pipeline Deployment

• Demo: Pipeline Deployment


4. Real-time Deployment and Online Stores

• Introduction to Real-time Deployment

• Databricks Model Serving

• Demo: Real-time Deployment with Model Serving

• Demo: Custom Model Deployment with Model Serving

• Lab: Real-time Deployment with Model Serving

Upcoming Public Classes

Date
Time
Your Local Time
Language
Price
Jan 07
09 AM - 01 PM (Asia/Singapore)
-
English
$750.00
Jan 07
01 PM - 05 PM (America/New_York)
-
English
$750.00
Jan 08
09 AM - 01 PM (Europe/Paris)
-
English
$750.00

Public Class Registration

If your company has purchased success credits or has a learning subscription, please fill out the Training Request form. Otherwise, you can register below.

Private Class Request

If your company is interested in private training, please submit a request.

See all our registration options

Registration options

Databricks has a delivery method for wherever you are on your learning journey

Runtime

Self-Paced

Custom-fit learning paths for data, analytics, and AI roles and career paths through on-demand videos

Register now

Instructors

Instructor-Led

Public and private courses taught by expert instructors across half-day to two-day courses

Register now

Learning

Blended Learning

Self-paced and weekly instructor-led sessions for every style of learner to optimize course completion and knowledge retention. Go to Subscriptions Catalog tab to purchase

Purchase now

Scale

Skills@Scale

Comprehensive training offering for large scale customers that includes learning elements for every style of learning. Inquire with your account executive for details

Upcoming Public Classes

Data Engineer

Databricks Performance Optimization - Mandarin Chinese

Databricks Performance Optimization 课程教导数据工程师和分析师如何在 Databricks Data Intelligence Platform 上诊断、衡量和修复性能瓶颈,以及如何将性能提升与成本关联起来。本课程遵循"先衡量、先平台"的工作流:让平台的自动优化功能承担繁重工作,通过 Query Profile 和系统表验证已应用的优化,再在必要时进行手动调优。

学员将以 Query Profile、Performance Insights 和查询历史记录系统表为基础,学习如何衡量性能。在此基础上,他们将探索关键优化技术,包括数据布局与自调优托管表、liquid clustering、缓存与中间结果、shuffle、数据倾斜、溢出、行爆炸、驱动程序性能、Python UDF、serverless compute、Photon 以及成本归因。

在整个课程中,学员将针对合成零售数据中刻意设计的慢查询进行探索,观察不同优化技术对性能的影响。他们将利用文件裁剪、任务执行时间、Photon 覆盖率和成本等依据来评估优化效果。两个基于场景的实验室将强化从诊断到验证的完整优化工作流。

注意:Databricks Academy 正在将 Databricks 环境中的课堂教学转为基于 notebook 的形式,不再使用幻灯片进行授课。您可以在 Vocareum 实验环境中访问课程 notebook。

Languages Available: English | 日本語 | Português BR | 한국어 | Français | Español | 普通话

Paid
4h
Lab
instructor-led
Professional

Databricks Data Privacy - Mandarin Chinese

Databricks 数据隐私课程面向数据工程师,讲解如何在 Databricks Data Intelligence Platform 上发现、保护、删除个人数据并对其负责。课程将隐私视为建立在 Unity Catalog 和 Delta Lake 之上的一组技术决策与控制。课程把隐私与安全、治理和合规区分开来,然后围绕识别、保护、管理、验证的生命周期,以及学习者可应用于每项技术的简短决策模型来组织工作。

学习者首先使用 Data Classification 查找敏感列,并用治理标签记录每列应如何处理。他们用基于标签的属性访问控制(ABAC)策略保护这些列,并在中央策略不适用时使用手动列掩码和行过滤器。去标识化进一步涵盖加盐哈希、通过单独治理的保险库进行令牌化,以及列加密,盐值和密钥保存在 Databricks Secrets 中。对于删除权,学习者跟踪请求,借助分类和血缘查找主体数据,用 DELETE 和 MERGE 删除并传播变更,再用 REORG TABLE ... APPLY (PURGE) 和考虑保留期的 VACUUM 清除底层文件。课程最后将血缘、审计日志、查询历史和 Delta 表历史组合成证据链,证明各项控制是否生效。

所有数据均为合成数据。演示沿着一家跨国零售商的数据资产走完生命周期,并以已掩码的查询结果、空的删除检查或表历史记录等证据结束每项控制。两个基于场景的实验强化该工作流。第一个实验中,学习者修复子公司客户表上损坏的保护。第二个实验中,他们完成一份只做了一半的删除请求,并汇总完成证据。本课程为技术课程,不提供法律建议。

注意: Databricks Academy 正在将 Databricks 环境中的课堂教学过渡为基于 Notebook 的形式,不再使用幻灯片进行讲座。您可以在 Vocareum 实验室环境中访问讲座 Notebook。

Paid
4h
Lab
instructor-led
Professional

Questions?

If you have any questions, please refer to our Frequently Asked Questions page.