Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Vertex AI
Unified ML Platform for the entire
AI workflow on Google Cloud
Advancing AI - Google For Startups - August, 2021...
● Among the Top3 romanians on Stackoverflow 192k reputation
● Google Developer Expert on Cloud technologies
● Crafting Web...
1. What is Vertex AI
2. Gather, Import & label datasets
3. Build, train and deploy ML solutions
4. Upload, deploy, serve e...
“VertexAI is a managed ML platform for practitioners
to accelerate experiments and deploy AI models.
Vertex AI - Unified M...
Where does VertexAI fit in?
Application Servers
Vertex AI
Desktop client
Mobile client
End-to-end platform for
ML model dev...
VertexAI is a unified MLOps platform
Vertex AI - Unified ML Platform @martonkodok
Operational
Model
Programming
Model
No In...
What’s included in VertexAI?
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL Environment (DL VM + DL ...
VertexAI supports...
Vertex AI - Unified ML Platform @martonkodok
UI based model development
# Define job
job = aiplatform...
Using Vertex AI throughout your ML workflow
Vertex AI - Unified ML Platform @martonkodok
Gather data Train model
Scalably
d...
@martonkodok
Gather,
Import & label
datasets at
scale
Part #1
1. Gathering data, datasets, AutoML models
Vertex AI - Unified ML Platform @martonkodok
Gather data Train model
Scalably
d...
VertexAI: Gather, Import & label datasets at scale
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL En...
“ Datasets
Vertex AI - Unified ML Platform @martonkodok
Datasets
Vertex AI - Unified ML Platform @martonkodok
Vertex AI datasets (fully managed)
• Fully serverless
• Region based...
“ VertexAI Managed Datasets + Models
(AutoML*)
Vertex AI - Unified ML Platform @martonkodok
* legacy name, previous genera...
- Regression/classification
- Forecasting
- Single-label classification
- Multi-label classification
- Text entity extraction...
Vertex AI: Managed dataset + model
Vertex AI - Unified ML Platform @martonkodok
Image
Vertex AI: Managed dataset + model
Vertex AI - Unified ML Platform @martonkodok
Tabular
Vertex AI: Managed dataset + model
Vertex AI - Unified ML Platform @martonkodok
Text
Vertex AI: Managed dataset + model
Vertex AI - Unified ML Platform @martonkodok
Video
Data labeling + Feature Store
Vertex AI - Unified ML Platform @martonkodok
Data labeling (fully managed)
• Create data lab...
@martonkodok
Build,
train & deploy
models at
scale
Part #2
2. Train models
Vertex AI - Unified ML Platform @martonkodok
Gather data Train model
Scalably
deploy
model
Evaluate,
monit...
VertexAI: Build, train & deploy models at scale
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL Envir...
1. AutoML out-of-the box training integration
No-code solution. You must target one of the AutoML’s predefined objectives.
...
Pre-built containers for custom training
https://cloud.google.com/vertex-ai/docs/training/pre-built-containers @martonkodo...
3. Deploying models
Vertex AI - Unified ML Platform @martonkodok
Gather data Train model
Scalably
deploy
model
Evaluate,
m...
“ Let’s have a moment…. about … hosting
Vertex AI - Unified ML Platform @martonkodok
Rewind: Application hosting on serverless CloudRun
Vertex AI - Unified ML Platform @martonkodok
Serving traffic
Source Cod...
“ You can deploy models on VertexAI
and get an Endpointto serve predictions
rapidly and reliably.
Vertex AI - Unified ML P...
1. You can use models whetherornotthemodelwastrained on Vertex AI.
2. Deploy a model and get aREST endpointto serve predic...
AutoML
Entity Extraction
on Vertex AI
Demo
Demo costs: $12
Training $3.00 per hour
Deployment $0.05 per hour
Prediction $5.00 per 1,000 text records
Vertex AI: Text ...
Manage your
models with
confidence
Part #4
VertexAI: Manage your models with confidence
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL Environme...
VertexAI: Manage your models with confidence
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL Environme...
Explainable AI
Vertex AI - Unified ML Platform @martonkodok
Explainable AI
• Interpret predictions made by ML models
• Rec...
Explainable AI: What-if Tool (pair-code.github.io/what-if-tool)
Vertex AI - Unified ML Platform @martonkodok
What-if Tool
...
VertexAI: Pipelines - Orchestrate your model
Vertex AI - Unified ML Platform @martonkodok
Data Labeling
AutoML
DL Environm...
Vertex AI: Pipelines
Vertex AI - Unified ML Platform
Source: Piero Esposito
https://github.com/piEsposito/vertex-ai-tutori...
Conclusion
Vertex AI - Unified ML Platform @martonkodok
1. Build with the groundbreaking ML tools that power Google
2. Approachable from the non-ML developer perspective (AutoML,...
Takeaways
Vertex AI - Unified ML Platform @martonkodok
1. Vertex AI: Building a fraud detection model with AutoML (using BQ public dataset)
https://codelabs.developers.google.co...
Who, what to follow, engage in Q&A
Vertex AI - Unified ML Platform @martonkodok
1. https://amygdala.github.io/
2. https://...
Google Cloud some other managed API offerings
Vertex AI - Unified ML Platform @martonkodok
Document AI
• Ready to use out-...
Google Cloud some other managed API offerings
Vertex AI - Unified ML Platform @martonkodok
AI and Machine Learning
Cloud V...
Thank you.
Slides available on:
slideshare.net/martonkodok
Reea.net - Integrated web solutions driven by creativity
to del...
You’ve finished this document.
Download and read it offline.
Upcoming SlideShare
What to Upload to SlideShare
Next
Upcoming SlideShare
What to Upload to SlideShare
Next
Download to read offline and view in fullscreen.

Share

Vertex AI - Unified ML Platform for the entire AI workflow on Google Cloud

Download to read offline

Vertex AI is a managed ML platform for practitioners to accelerate experiments and deploy AI models.

Enhanced developer experience
- Build with the groundbreaking ML tools that power Google
- Approachable from the non-ML developer perspective (AutoML, managed models, training)
- Ease the life of a data scientist/ML (has feature store, managed datasets, endpoints, notebooks)
- Infrastructure management overhead have been almost completely eliminated
- Unified UI for the entire ML workflow
- End-to-end integration for data and AI with build pipelines that outperform and solve complex ML tasks
- Explainable AI and TensorBoard to visualize and track ML experiments

Related Books

Free with a 30 day trial from Scribd

See all

Related Audiobooks

Free with a 30 day trial from Scribd

See all
  • Be the first to like this

Vertex AI - Unified ML Platform for the entire AI workflow on Google Cloud

  1. 1. Vertex AI Unified ML Platform for the entire AI workflow on Google Cloud Advancing AI - Google For Startups - August, 2021 Márton Kodok Google Developer Expert at REEA.net
  2. 2. ● Among the Top3 romanians on Stackoverflow 192k reputation ● Google Developer Expert on Cloud technologies ● Crafting Web/Mobile backends at REEA.net ● BigQuery + Redis database engine expert Slideshare: martonkodok Articles: martonkodok.medium.com Twitter: @martonkodok StackOverflow: pentium10 GitHub: pentium10 Vertex AI - Unified ML Platform @martonkodok About me
  3. 3. 1. What is Vertex AI 2. Gather, Import & label datasets 3. Build, train and deploy ML solutions 4. Upload, deploy, serve endpoints for your ML model like an API 5. Demo 6. Tools to manage your models with confidence 7. Conclusions Agenda Vertex AI - Unified ML Platform @martonkodok
  4. 4. “VertexAI is a managed ML platform for practitioners to accelerate experiments and deploy AI models. Vertex AI - Unified ML Platform @martonkodok
  5. 5. Where does VertexAI fit in? Application Servers Vertex AI Desktop client Mobile client End-to-end platform for ML model development and deployment Backend Vertex AI - Unified ML Platform @martonkodok Application Logic
  6. 6. VertexAI is a unified MLOps platform Vertex AI - Unified ML Platform @martonkodok Operational Model Programming Model No Infra Management Managed Security Pay only for usage Model-as-a-service oriented Streamlined model development Open SDKs, integrates with ML frameworks
  7. 7. What’s included in VertexAI? Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  8. 8. VertexAI supports... Vertex AI - Unified ML Platform @martonkodok UI based model development # Define job job = aiplatform.AutoMLTabularTrainingJob( display_name='price-predict-training', optimization_prediction_type='regression' ) # Run job model = job.run( dataset=ds, target_column='median_house_value', model_display_name='house-value-prediction', ) Code-based model development
  9. 9. Using Vertex AI throughout your ML workflow Vertex AI - Unified ML Platform @martonkodok Gather data Train model Scalably deploy model Evaluate, monitor, retrain
  10. 10. @martonkodok Gather, Import & label datasets at scale Part #1
  11. 11. 1. Gathering data, datasets, AutoML models Vertex AI - Unified ML Platform @martonkodok Gather data Train model Scalably deploy model Evaluate, monitor, retrain
  12. 12. VertexAI: Gather, Import & label datasets at scale Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  13. 13. “ Datasets Vertex AI - Unified ML Platform @martonkodok
  14. 14. Datasets Vertex AI - Unified ML Platform @martonkodok Vertex AI datasets (fully managed) • Fully serverless • Region based • Free to store • In tandem with AutoML managed models Custom Datasets Cloud Storage, BigQuery or on Internet Accessing managed dataset from your app: - JSONL (default) - CSV or BigQuery stream
  15. 15. “ VertexAI Managed Datasets + Models (AutoML*) Vertex AI - Unified ML Platform @martonkodok * legacy name, previous generation naming from AI Platform
  16. 16. - Regression/classification - Forecasting - Single-label classification - Multi-label classification - Text entity extraction - Text sentiment analysis - Video action recognition - Video classifications for entire video, shots, frames - Video object tracking Vertex AI: Managed Datasets + Models Vertex AI - Unified ML Platform @martonkodok Image Tabular Text Video - Single-label classification - Multi-label classification - Image object detection - Image segmentation
  17. 17. Vertex AI: Managed dataset + model Vertex AI - Unified ML Platform @martonkodok Image
  18. 18. Vertex AI: Managed dataset + model Vertex AI - Unified ML Platform @martonkodok Tabular
  19. 19. Vertex AI: Managed dataset + model Vertex AI - Unified ML Platform @martonkodok Text
  20. 20. Vertex AI: Managed dataset + model Vertex AI - Unified ML Platform @martonkodok Video
  21. 21. Data labeling + Feature Store Vertex AI - Unified ML Platform @martonkodok Data labeling (fully managed) • Create data labeling, annotation tasks • Use human labelers • Use Google’s labeler workforce • Use your own workforce Feature Store (fully managed) • Centralized repository for organizing, storing, and serving ML features • Organization can efficiently share, discover, re-use features • Data Model: Entity Type -> Feature • Ingest data from BigQuery or Cloud Storage Pro: point-in-time lookup from time series
  22. 22. @martonkodok Build, train & deploy models at scale Part #2
  23. 23. 2. Train models Vertex AI - Unified ML Platform @martonkodok Gather data Train model Scalably deploy model Evaluate, monitor, retrain
  24. 24. VertexAI: Build, train & deploy models at scale Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  25. 25. 1. AutoML out-of-the box training integration No-code solution. You must target one of the AutoML’s predefined objectives. 2. Custom Training - run your own training applications in the cloud Train with one of the Google’s pre-builtcontainers or useyourown. 3. Hyperparameter tuning jobs - searchesforbestcombination of hyperparameter values by optimizing values across a series of trials. Available for custom training. Your training app must adhere to accepting Vertex AI parameters. You need to report metrics to Vertex AI. Training https://cloud.google.com/vertex-ai/docs/training/using-hyperparameter-tuning @martonkodok
  26. 26. Pre-built containers for custom training https://cloud.google.com/vertex-ai/docs/training/pre-built-containers @martonkodok Tensorflow ML Framework version 1.15, 2.1-2.4 use with Cuda 11.x GPU scikit-learn ML Framework version 0.23 No GPUs PyTorch ML Framework version 1.4 - 1.7 use with Cuda 11.x GPU XGBoost ML Framework version 1.1 No GPUs
  27. 27. 3. Deploying models Vertex AI - Unified ML Platform @martonkodok Gather data Train model Scalably deploy model Evaluate, monitor, retrain
  28. 28. “ Let’s have a moment…. about … hosting Vertex AI - Unified ML Platform @martonkodok
  29. 29. Rewind: Application hosting on serverless CloudRun Vertex AI - Unified ML Platform @martonkodok Serving traffic Source Code Cloud Run (fully managed) • Build source code into container • Deploy Container to production in seconds • Fully serverless • No cluster to manage • Pay for what you use
  30. 30. “ You can deploy models on VertexAI and get an Endpointto serve predictions rapidly and reliably. Vertex AI - Unified ML Platform @martonkodok
  31. 31. 1. You can use models whetherornotthemodelwastrained on Vertex AI. 2. Deploy a model and get aREST endpointto serve predictions realtime or batched 3. Specify a prediction traffic split in your endpoint. 4. VPC Private Network option for custom-trained models/tabular models 5. Can use a customer-managed encryption key(CMEK) 6. Out of the box Cloud Logging integration Vertex AI: Endpoints Vertex AI - Unified ML Platform @martonkodok Vertex AI Endpoints Backend Prediction deploy REST
  32. 32. AutoML Entity Extraction on Vertex AI Demo
  33. 33. Demo costs: $12 Training $3.00 per hour Deployment $0.05 per hour Prediction $5.00 per 1,000 text records Vertex AI: Text Entity Extraction (training set 75K records) Vertex AI - Unified ML Platform @martonkodok
  34. 34. Manage your models with confidence Part #4
  35. 35. VertexAI: Manage your models with confidence Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  36. 36. VertexAI: Manage your models with confidence Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  37. 37. Explainable AI Vertex AI - Unified ML Platform @martonkodok Explainable AI • Interpret predictions made by ML models • Receive a score explaining how much each factor contributed to the model predictions • What-If Tool lets you investigate model behavior at a glance AI Explanations samples Github: GoogleCloudPlatform/ai-platform-samples - Training, deploying, and explaining a tabular data model - Training, deploying, and explaining an image model Limitations: doesn’t work well on low-contrast, X-rays, one shade, panoramas, very tall, very wide images.
  38. 38. Explainable AI: What-if Tool (pair-code.github.io/what-if-tool) Vertex AI - Unified ML Platform @martonkodok What-if Tool • Model probing, from within any workflow • test performance in hypothetical situations • analyze the importance of different data features • visualize model behavior across multiple models and subsets of data Tutorials, demos onpair-code.github.io/what-if-tool • Available on many platforms (TensorBoard, Jupyter, Colaboratory, Vertex AI) • Supports what-if Analyses (explore counterfactuals, fairness measures, partial dependence plots) • Visualizes Model Performances (threshold simulation, up to 2 model comparison, dataset summary statistics)
  39. 39. VertexAI: Pipelines - Orchestrate your model Vertex AI - Unified ML Platform @martonkodok Data Labeling AutoML DL Environment (DL VM + DL Container) Prediction Feature Store Training Experiments Data Readiness Feature Engineering Training/ HP-Tuning Model Monitoring Model serving Understanding/ Tuning Edge Model Management Notebooks Pipelines (Orchestration) Explainable AI Hybrid AI Model Monitoring Metadata Vision Translation Tables Language Video AI Accelerators Models Datasets Custom Models Containers Python Endpoints Vizier Optimization
  40. 40. Vertex AI: Pipelines Vertex AI - Unified ML Platform Source: Piero Esposito https://github.com/piEsposito/vertex-ai-tutorials
  41. 41. Conclusion Vertex AI - Unified ML Platform @martonkodok
  42. 42. 1. Build with the groundbreaking ML tools that power Google 2. Approachable from the non-ML developer perspective (AutoML, managed models, training) 3. Ease the life of a data scientist/ML (has feature store, managed datasets, endpoints, notebooks) 4. Infrastructure management overhead have been almost completely eliminated 5. Unified UI for the entire ML workflow 6. End-to-end integration for data and AI with build pipelines that outperform and solve complex ML tasks 7. Explainable AI and TensorBoard to visualize and track ML experiments Vertex AI: Enhanced developer experience Vertex AI - Unified ML Platform @martonkodok
  43. 43. Takeaways Vertex AI - Unified ML Platform @martonkodok
  44. 44. 1. Vertex AI: Building a fraud detection model with AutoML (using BQ public dataset) https://codelabs.developers.google.com/vertex-automl-tabular 2. Building a financial ML model with the What-If Tool and Vertex AI https://codelabs.developers.google.com/vertex-xgb-wit 3. How to Use Pipeline on Google Cloud’s Vertex AI https://betterprogramming.pub/how-to-use-pipeline-on-google-clouds-vertex-ai-863b429c811f 4. Use Document AI to Intelligently Process your Handwritten Forms https://codelabs.developers.google.com/codelabs/docai-form-parser-v3-python Takeaways: Codelabs Vertex AI - Unified ML Platform @martonkodok
  45. 45. Who, what to follow, engage in Q&A Vertex AI - Unified ML Platform @martonkodok 1. https://amygdala.github.io/ 2. https://twitter.com/SRobTweets 3. https://twitter.com/lak_gcp 4. https://www.youtube.com/watch?v=gT4qqHMiEpA&list=PLIivdWyY5sqJ1YuMdGjRwJ3fFYZ_vWQ62 5. https://stackoverflow.com/questions/tagged/google-cloud-vertex-ai
  46. 46. Google Cloud some other managed API offerings Vertex AI - Unified ML Platform @martonkodok Document AI • Ready to use out-of-the-box processors for general document goals • Parse the contents of a form, table, or invoice. • Convert images to text • Classify documents, extract entities • 21 different processors and counting Natural Language AI • Reveal the structure and meaning of text • Extract information about people, places and events • Better understand sentiments and customer conversation
  47. 47. Google Cloud some other managed API offerings Vertex AI - Unified ML Platform @martonkodok AI and Machine Learning Cloud Vision API Speech-to-Text Cloud Natural Language API Cloud Translation API Video Intelligence API Advanced Solutions Lab AutoML Cloud TPU AutoML Video Intelligence AutoML Natural Language AutoML Translation AutoML Vision Recommendations AI AutoML Tables Cloud Jobs API Cloud Inference API Data Labeling Vertex AI Dialog Flow Enterprise Edition Text-to-Speech BigQuery ML
  48. 48. Thank you. Slides available on: slideshare.net/martonkodok Reea.net - Integrated web solutions driven by creativity to deliver projects.

Vertex AI is a managed ML platform for practitioners to accelerate experiments and deploy AI models. Enhanced developer experience - Build with the groundbreaking ML tools that power Google - Approachable from the non-ML developer perspective (AutoML, managed models, training) - Ease the life of a data scientist/ML (has feature store, managed datasets, endpoints, notebooks) - Infrastructure management overhead have been almost completely eliminated - Unified UI for the entire ML workflow - End-to-end integration for data and AI with build pipelines that outperform and solve complex ML tasks - Explainable AI and TensorBoard to visualize and track ML experiments

Views

Total views

127

On Slideshare

0

From embeds

0

Number of embeds

0

Actions

Downloads

14

Shares

0

Comments

0

Likes

0

×