asl-ml-immersion
Notebooks, code samples and reference for machine learning and generative ai on Google Cloud for the Advanced Solutions Lab (ASL) bootcamps.
What is asl-ml-immersion?
Notebooks, code samples and reference for machine learning and generative ai on Google Cloud for the Advanced Solutions Lab (ASL) bootcamps.
Overview of asl-ml-immersion
What it is
asl-ml-immersion is a repository maintained by Google Cloud’s Advanced Solutions Lab (ASL) team. It contains notebooks, code samples, and reference materials for machine learning and generative AI on Google Cloud. The repository is designed for use in ASL bootcamps and covers a wide range of AI/ML topics through three main modules: asl_core, asl_mlops, and asl_genai. The materials are organized by topic, with each topic containing a labs folder for hands-on exercises and a solutions folder for reference code. The repository is licensed under Apache-2.0 and is open source.
Key features
- Three main modules: asl_core (model architectures for various data modalities using TensorFlow and Keras), asl_mlops (operationalizing models on Google Cloud’s Agent Platform with tools like Kubeflow pipelines), and asl_genai (generative AI and agent systems using Gemini and Google ADK).
- Learning materials are organized by topic, each with labs and solutions folders for self-paced practice.
- Environment setup scripts for Agent Platform Workbench and Cloud Workstations, with optional GPU (Nvidia T4) attachment.
- Supports multiple frameworks including TensorFlow, Keras, Scikit-learn, and PyTorch.
- Includes kernel scripts and sample code scaffolds to accelerate AI/ML projects.
Use cases
- Learning and practicing machine learning and generative AI concepts on Google Cloud.
- Building and deploying production ML systems using TensorFlow, Scikit-learn, or PyTorch models at scale.
- Exploring generative AI and agent systems with Gemini and agentic frameworks.
- Hands-on training for participants of Google Cloud’s Advanced Solutions Lab bootcamps.
Getting started
Clone the repository and run the setup script in Cloud Shell to configure project infrastructure (APIs, IAM, Buckets). Then, inside the environment (Agent Platform Workbench or Cloud Workstations), clone the repository again and run make to build the environment (venvs and Jupyter kernels). bash git clone --depth 1 https://github.com/GoogleCloudPlatform/asl-ml-immersion.git cd asl-ml-immersion bash scripts/setup_env.sh After setup, open a Jupyter notebook and select the appropriate kernel (ASL Core, ASL MLOps, or ASL Agent).
License and pricing
The repository is licensed under Apache-2.0. Pricing is not stated; the repository is open source.
Based on the project's public README and repository metadata.