Hands-on MLOps projects to explore and learn the practical aspects of machine learning engineering for production.
| # | Project | Description |
|---|---|---|
| 1 | Deploying a Deep Learning model | Rapid guide to deploying a computer vision model trained to identify common objects in images using YOLOv3 model and FastAPI. |
| 2 | Birds, Cats, and Dogs | Simple Convolutional Neural Network (CNN) addressing class imbalance and overfitting issues in a data-centric approach. |
| 3 | YouTube Spam | Simple classifier exploring how different labeling strategies affect machine learning model performance. |
| 4 | Earnings Predictor | Introduction to generating and visualizing statistics, detecting and fixing anomalies in an evaluation dataset with TensorFlow Data Validation (TFDV). |
| 5 | Predicting Patient Readmission | In-depth look at TFDV to generate and visualize statistics from a dataframe, infer a dataset schema, calculate, visualize, and fix anomalies. |
| 6 | Hello World | Simple feature engineering task to collect data, define metadata, create a preprocessing function, and generate a graph using TensorFlow Transform. |
| 7 | Earnings Predictor Pipeline | Machine learning pipeline using TensorFlow Extended (TFX) components: ExampleGen, StatisticsGen, SchemaGen, ExampleValidator, and Transform. |
| 8 | Traffic Volume Pipeline | Advanced feature engineering pipeline utilizing TFX components: data splitting, dataset analysis, validation, and transforms. |
| 9 | Breast Cancer Features Analysis | Exploration of various feature selection techniques: correlation-based filtering, recursive feature elimination (RFE), and feature importance embedding. |
| 10 | Chicago Taxi Metadata | Walkthrough of ML Metadata: recording and retrieving project metadata from a MetadataStore. |
| 11 | Income Dataset Schemas | Deep dive into updating an inferred schema and storing the updated version in the TFX metadata store. |
| 12 | Forest Cover Pipeline Components | Production ML data pipeline: data ingestion, data validation, and data transformation. |
| 13 | Weather Data Transformation | Feature engineering process to prepare seasonal data: handling time series, periodicity, and producing batches of dataset windows. |
| 14 | Accelerometer Data Transformation | Preparation of raw sensor time-series data by grouping it into windows and extracting useful features. |
| 15 | Image Data Transformation | Image data preprocessing: parsing images and converting them into float arrays for feature engineering. |
| 16 | Hyperparameter Tuning | Introduction to hyperparameter tuning with Keras Tuner: automate the search for optimal settings to enhance model performances. |
| 17 | TFX Tuner and Model Training | Introduction to hyperparameter tuning and model training with TFX. |
| 18 | Manual Feature Engineering | Exploration of the impact of manually transforming raw features to improve predictions, using TensorFlow and Keras. |
| 19 | Algorithmic Dimensionality Reduction | Explore dimensionality reduction algorithms using PCA, SVD, and NMF for improved model performance. |
| 20 | Quantization and Pruning | Lab optimizing models for mobile and IoT devices using TF Lite format, post-training quantization, quantization aware training, and weight pruning. |
| 21 | Distributed Training | Explore distributed training in TensorFlow and Keras using a multi-worker training strategy. |
| 22 | Knowledge Distillation | Discover model compression through knowledge distillation, where a smaller student model learns from a more complex teacher model. |
| 23 | TensorFlow Model Analysis | Dive into TensorFlow Model Analysis (TFMA) to analyze and visualize model performance across various data slices, set thresholds, and compare models. |
| 24 | TFX Evaluator | Explore model analysis in the TFX pipeline using TFX Evaluator to ensure trained models meet metrics requirements and compare with previous models. |
| 25 | Fairness Indicators | Explore fairness metrics in a face image dataset using Fairness Indicators, analyzing model predictions across age groups. |
| 26 | Shapley Values | Investigate SHAP (SHapley Additive exPlanations) by training a CNN and computing Shapley values for each class. |
| 27 | Permutation Feature Importance | Train a Random Forest classifier, compute feature importance, re-train with top features, and compare with other classifiers. |
| 28 | Intro to Docker | Dive into the essential foundations for deploying Machine Learning models through Docker installation and curl setup. |
| 29 | TensorFlow Serving with Docker | Get a first look at using TensorFlow Serving with Docker, a straightforward introduction to this serving system for machine learning models. |
| 30 | Serve a Model with TensorFlow Serving | Explore TensorFlow Serving without Docker, gaining insights into installation, loading models, REST API interaction, and model versioning. |
| 31 | Deploy a ML model with fastAPI and Docker | Deploy a web server hosting a Random Forest classifier using FastAPI and Docker, exploring batching and single prediction endpoints. |
| 32 | Practice Kubernetes in your Local Environment | Expand your Kubernetes skills: set up Minikube locally, create objects with YAML, deploy Tensorflow model servers, and explore autoscaling. |
| 33 | Load testing servers with Docker Compose and Locust | Conduct a load test on servers with Docker Compose and Locust, comparing latency across different batching configurations. |
| 34 | ETL Pipelines and Batch Predictions | Explore the ETL (Extract, Transform, Load) process with Apache Beam and TensorFlow. |
| 35 | Building ML Pipelines with Kubeflow | Hands-on experience with Kubeflow Pipelines for automating and orchestrating machine learning workflows. |
| 36 | Developing Custom TFX Components | Learn to build custom components in TensorFlow Extended (TFX) for more flexible machine learning pipelines. |
| 37 | Model versioning with TensorFlow Serving and Docker | Learn to version machine learning models using TensorFlow Serving and Docker for controlled deployments and A/B testing. |
| 38 | Intro to CI/CD pipelines with GitHub Actions | Explore GitHub Actions for automating ML workflows, including unit testing with pytest for code quality assurance. |
| Type | Title | Description |
|---|---|---|
| Lab | A Tour of Google Cloud Hands-on Labs | Walkthrough of Google Cloud and the Qwiklabs platform. |
| Lab | Classify Images of Clouds in the Cloud with AutoML Images | Hands-on lab introducing the AutoML UI for classifying images of clouds in the Google Cloud environment. |
| Course | Integrate with Machine Learning APIs | Hands-on course exploring basic features of the Cloud Vision API, Cloud Translation API, and Cloud Natural Language API. |
| Lab | Running Distributed TensorFlow using Vertex AI | Learn to deploy a distributed TensorFlow training pipeline using Vertex AI's MirrorStrategy and set up an endpoint for cloud-based online predictions. |
| Lab | Machine Learning with TensorFlow in Vertex AI | Learn to develop a TensorFlow model in Vertex AI Workbench: train, create input data pipeline, deploy to an endpoint, and get predictions. |
| Lab | Introduction to Docker | Explore Docker basics, including container building, running, debugging, and image management with Google Artifact Registry. |
| Lab | Vertex AI Workbench Notebook: Qwik Start | Hands-on introduction to TensorFlow model training, deployment on Vertex AI, and predicting income categories. |
| Lab | Deploy a BigQuery ML Customer Churn Classifier to Vertex AI | Deploy a BigQuery ML XGBoost model to Vertex AI for online predictions, predicting user churn in a real mobile application dataset. |
| Course | Deploy to Kubernetes in Google Cloud | Learn about Google Kubernetes Engine, configure and build Docker containers, create clusters, and deploy applications. |
| Course | Advanced ML: ML Infrastructure | Advanced-level course covering machine learning at scale and leveraging advanced ML infrastructure tools on Google Cloud Platform. |
| Lab | Data Loss Prevention: Qwik Start - JSON | Set up a JSON file, send it to the Data Loss Prevention API, inspect and hide sensitive information. |
| Course | Automate Interactions with Contact Center AI | Learn to build and design conversation flows, add phone gateways, use Dialogflow for troubleshooting, and review logs to debug virtual agents. |
| Course | Generative AI Fundamentals | Introduction to generative AI, covering fundamental concepts in generative AI, large language models, and responsible AI. |
| Course | Analyze Images with the Cloud Vision API | Learn to analyze images with the Cloud Vision API, including practical use cases such as reading text embedded in an image. |
| Course | Analyze Speech and Language with Google APIs | Explore real-world applications of Natural Language and Speech APIs for speech and language analysis. |
| Course | Detect Manufacturing Defects using Visual Inspection AI | Master Visual Inspection AI to detect manufacturing defects by deploying solutions and testing their defect identification capabilities. |
| Course | Get Started with Looker | Learn how to analyze, visualize, and curate data using Looker Studio and Looker. |
| Course | Exploring Data with Looker | Explore data using Looker's interface, covering dimensions, measures, filtering, pivoting, and merging explores. |
| Course | Build LookML Objects in Looker | Learn to design and build LookML objects such as dimensions, measures, and views in Looker. |
| Course | Predict Soccer Match Outcomes with BigQuery ML | Explore sports data science fundamentals using BigQuery, including creating and analyzing a soccer dataset. |
| Course | Manage Data Models in Looker | Learn LookML best practices, optimize queries, and implement caching policies for effective data model management. |
| Course | Build and Deploy ML Solutions on Vertex AI | Explore Google Cloud's Vertex AI platform for training, evaluating, tuning, explaining, and deploying machine learning solutions. |
| Course | Insights from Data with BigQuery | Explore basic features of BigQuery, covering SQL queries, database table management, query troubleshooting, and creating reports in Looker. |
| Course | Create ML Models with BigQuery ML | Learn how to use BigQuery ML to create machine learning models, including classification, forecasting, and implementation of a chatbot. |
| Course | Build and Optimize Data Warehouses with BigQuery | Transform your data warehouse with BigQuery, covering joins, unions, table partitioning, and working with JSON, arrays, and structs. |
| Course | Foundational Data, ML, and AI Tasks in Google Cloud | Explore basic features of Google Cloud's machine learning and AI technologies, covering BigQuery, Cloud Speech AI, Dataflow, Dataproc, and more. |
| Course | Vector Search and Embeddings | Learn about Vertex AI Vector Search and how to build search applications with large language model (LLM) APIs for embeddings. |
Check out our collection of articles for those beginning their MLOps journey. Find tips, tricks, and motivational content to keep you engaged and motivated throughout your learning process.
Based on Machine Learning Engineering for Production (MLOps) Specialization by Andrew Ng, Laurence Moroney, and Robert Crowe (2023).
If you find these projects helpful or interesting, please consider starring the repository. It's a simple gesture that helps to boost the visibility of the project and show appreciation for the effort put into creating it. Thank you!