ASL-ml-immersion provides Google Cloud–based notebooks and code samples for ML/GenAI, organized into core, MLOps, and GenAI modules. It includes setup scripts, environment provisioning, and instructions to run notebooks via ASL kernels.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
A repository containing AI/ML learning materials and reference implementations intended to run on Google Cloud, organized into three modules: asl_core, asl_mlops, and asl_genai. The materials are structured with a contents folder containing topic-specific labs and solutions.
How it works
The project provides environment setup and instructions to clone the repo, create Python virtual environments, and build the environment using make. It supports running notebooks via module-specific kernels (ASL Core, ASL MLOps, ASL Agent) and includes environment-specific guidance for Agent Platform Workbench and Cloud Workstations.
Getting started
Install steps from the Environment Setup section:
- Clone the repository and run the setup script:
git clone https://github.com/GoogleCloudPlatform/asl-ml-immersion.git
cd asl-ml-immersion
bash scripts/setup_env.sh
- Run the environment build and kernel setup:
git clone https://github.com/GoogleCloudPlatform/asl-ml-immersion.git
cd asl-ml-immersion
make
- To run notebooks, select the appropriate kernel (ASL Core, ASL MLOps, or ASL Agent) in the Jupyter interface.
Recent releases
Latest releases include:
- keras3 ASL with Keras 3 (2026-01-14)
- keras2 Repo with Keras2 notebooks (2026-01-14): Archiving the repo with Keras2 and Tensorflow centric notebooks
- um_notebook ASL on Vertex AI User-Managed Notebook (2024-06-13)
- TF2.8 ASL on TF 2.8 (2023-08-21)
- TF2.6 ASL on TF 2.6 (2023-01-11): ASL contents based on TF 2.6 (before module update #278)
Traction
Stars: 2591 Forks: 1515 Open issues: 16
Licensing
Licensed under Apache-2.0. The license text is provided in the repository.
Environment notes
All code is intended to run on Google Cloud with an emphasis on Agent Platform Workbench or Cloud Workstations. Accelerators (GPU/TPU) are optional and not required for most notebooks, though some may benefit from them.
Caveats
Disclaimer notes state this is not an officially supported Google product and usage incurs Google Cloud charges. Licensing is Apache-2.0.