AI Frameworks · GoogleCloudPlatform

asl-ml-immersion

This is Google Cloud's official educational repository containing machine learning and generative AI notebooks, code samples, and reference architectures for running AI workloads on Google Cloud. It covers three main areas: core ML model architectures (DNNs, CNNs, RNNs, transformers), MLOps operationalization on Vertex AI, and generative AI with Gemini and agentic frameworks.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/GoogleCloudPlatform/asl-ml-immersion

2.6k

GitHub stars

1.5k

Forks

Jupyter Notebook

Primary language

Apache-2.0

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	GoogleCloudPlatform/asl-ml-immersion
Owner	GoogleCloudPlatform
Primary language	Jupyter Notebook
License	Apache-2.0 — OSI-approved
Stars	2.6k
Forks	1.5k
Open issues	18
Latest release	keras3 (2026-01-14)
Last updated	2026-07-08
Source	https://github.com/GoogleCloudPlatform/asl-ml-immersion

What asl-ml-immersion is

Jupyter Notebook–based learning materials implementing TensorFlow/Keras models for tabular, image, text, and time-series data, plus Vertex AI training/serving and Kubeflow MLOps pipelines. Organized into three modules (asl_core, asl_mlops, asl_genai) with separate Python virtual environments and kernels, designed for Vertex AI Workbench and Cloud Workstations.

Quickstart

Get the asl-ml-immersion source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/GoogleCloudPlatform/asl-ml-immersion.git2cd asl-ml-immersion3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Google Cloud ML Bootcamp Training

Structured learning path for teams adopting Vertex AI, with labs and solution notebooks spanning beginner to advanced production ML systems. Ideal for ASL bootcamp participants and organizations standardizing on GCP for ML workflows.

Generative AI & Agent Development

Pre-built examples and scaffolds for building agentic systems with Google's ADK framework and Gemini API. Accelerates prototyping of multi-turn agents and RAG applications on Google Cloud infrastructure.

MLOps Pipeline Implementation

Reference implementations for operationalizing TensorFlow, scikit-learn, and PyTorch models at scale using Vertex AI training, hyperparameter tuning, and Kubeflow-based orchestration. Demonstrates best practices for production handoff.

Implementation considerations

Requires active GCP project with APIs enabled (Vertex AI, Cloud Storage, Workbench/Workstations) and associated costs; setup script automates some IAM and infrastructure provisioning.
Each module (asl_core, asl_mlops, asl_genai) has isolated venv and dependencies; ensure adequate disk and memory for multiple kernels if running notebooks in parallel.
Some notebooks recommend GPU (T4) acceleration; cost/benefit depends on batch size and model architecture; CPU-only execution is viable for many labs.
Contributing requires Google employee status; external users can fork and adapt but cannot upstream PRs, limiting customization feedback loop.
Notebooks assume familiarity with Jupyter, Python, and basic ML concepts; not a zero-to-one introduction for non-technical stakeholders.

When to avoid it — and what to weigh

Non-Google Cloud Deployment — Repository is tightly coupled to Google Cloud services (Vertex AI, Cloud Storage, Cloud Workstations, Gemini API). Limited portability to AWS, Azure, or on-premises without significant refactoring.
Proprietary ML Framework Requirement — Focus is TensorFlow/Keras, scikit-learn, and PyTorch. If your org standardizes on JAX, Hugging Face transformers exclusively, or proprietary frameworks, material may require translation.
Self-Contained Offline Environment — Setup script and execution assume active GCP project with internet access for API provisioning, IAM configuration, and cloud resource allocation. Not suitable for air-gapped or disconnected deployments.
Production Code Library — Repository is educational and labeled 'not an officially supported Google product.' Code is reference-grade, not hardened for direct production use without review and customization.

License & commercial use

Apache License 2.0 (Apache-2.0). Permissive open-source license allowing modification, distribution, and private use with liability and trademark disclaimers. No attribution requirement for derivative use.

Apache 2.0 permits commercial use of the code itself. However, the repository is explicitly not an officially supported Google product, and executing these notebooks on Google Cloud will incur GCP service charges (Vertex AI, compute, storage). Legal review recommended before embedding in proprietary products; warranty disclaimers apply.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Strong
License clarity	Clear
Deployment complexity	Moderate
DEV.co fit	Strong
Assessment confidence	High

Security considerations

Repository contains educational code samples with no explicit security audit claim. Notebooks interact with GCP services using credentials in the runtime environment; ensure IAM roles follow least-privilege. Code samples are not vetted for injection vulnerabilities or cryptographic best practices; treat as reference, not production-hardened. Sensitive data (API keys, credentials) should never be hardcoded in notebooks; use GCP Secret Manager or workbench environment variables.

Alternatives to consider

DeepLearning.AI (Coursera / YouTube)

Free, framework-agnostic ML courses with Jupyter notebooks. Less GCP-specific; covers TensorFlow, PyTorch, and LLM fundamentals without vendor lock-in.

Hugging Face Course & Hub

Community-driven transformer and NLP focus. Better for LLM fine-tuning and open-source model deployment; less MLOps/operationalization depth than asl-ml-immersion.

Databricks Academy / MLflow + MLOps Examples

Unified data+ML platform with notebooks and production pipeline templates. Cloud-agnostic (AWS, Azure, GCP); no tight GCP coupling required.

Software development agency

Build on asl-ml-immersion with DEV.co software developers

Start with the setup script in Cloud Shell, clone the repository, and work through the labs. Ideal for teams adopting Vertex AI and agentic AI frameworks.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

generative-ai ml-engineering generative-ai

Related on DEV.co

Explore the category and the services that help you build with it.

AI Frameworks AI Application Development Cloud Deployment Custom Software & MLOps Talk to DEV.co

asl-ml-immersion FAQ

Can I run these notebooks outside Google Cloud?

Partially. Core ML model code (asl_core) can be ported to any TensorFlow/Keras environment. MLOps and GenAI modules depend on Vertex AI, Gemini API, and ADK; refactoring required for AWS SageMaker, Azure ML, or on-prem Kubernetes.

Do I need a Google Cloud account?

Yes. Setup script provisions Vertex AI, Cloud Storage, IAM, and optionally Workbench/Workstations. GCP billing will apply. Free trial credits may cover initial exploration.

Is this suitable for production ML systems?

No. Repository is educational reference material, not officially supported by Google. Use as architectural guidance and PoC accelerator; harden, test, and audit before production. Consider Vertex AI managed services for production operationalization.

Can I contribute improvements or fixes?

Currently Googlers only. External contributors can fork, adapt, and use locally but cannot submit upstream PRs. See CONTRIBUTING.md for details.

Software development & web development with DEV.co

Need help beyond evaluating asl-ml-immersion? DEV.co is a software development agency offering software development services and web development for teams of every size. Our software developers and web developers build custom software, web applications, APIs, and ai frameworks integrations — and maintain them long-term.

Ready to accelerate your ML journey on Google Cloud?

Start with the setup script in Cloud Shell, clone the repository, and work through the labs. Ideal for teams adopting Vertex AI and agentic AI frameworks.

Talk to DEV.co Schedule a Consultation