quick-start Project that Domino creates inside your own deployment, and a set of public GitHub repositories you can open as Projects of your own. Start with quick-start for the platform basics, then pick a repository when you need a specific capability such as an agent, distributed training, or a flow.
The quick-start Project
Domino creates a quick-start Project for every new user at signup. It has example files that show how to take advantage of Domino’s features, and a detailed README. Because Domino provisions it for you, it’s the quickest example to open. Some features ship their example code there: the /custom metrics example folder holds end-to-end code for custom Model Monitoring metrics.
Administrators can start new users on Projects their own team maintains instead, by changing the default Project configuration record.
Example repositories on GitHub
Open any of these as a Git-based Project. Each group links to the documentation for the capability; the repositories hold the code.Agents
Agents explains how you build an agentic system in Domino, evaluate it against real data, and ship it.simple_domino_agent is a minimal agent with tool calls, tracing, evaluation, and deployment, plus a production app with a chat UI in chat_app.py.
dominodatalab/simple_domino_agent
Loading repository data...
rag-agent-demo is a retrieval-augmented generation (RAG) agent built on ChromaDB, using Domino’s tracing instrumentation.
dominodatalab/rag-agent-demo
Loading repository data...
simple_agent_api_only is a REST API and agent-to-agent (A2A) example, with instructions for pairing it with a Domino-hosted agent registry and orchestrator.
dominodatalab/simple_agent_api_only
Loading repository data...
Distributed computing
Scale distributed workloads covers the on-demand Spark, Dask, Ray, and MPI clusters these examples run on.domino-quickstart-ray covers Ray environment setup and multiple types of training algorithms.
dominodatalab/domino-quickstart-ray
Loading repository data...
domino-quickstart-dask covers Dask environment setup and multiple types of training algorithms.
dominodatalab/domino-quickstart-dask
Loading repository data...
reference-project-dask is a Dask pipeline covering environment setup, data pipelines, and model training. Start with dask_ml.ipynb.
dominodatalab/reference-project-dask
Loading repository data...
spark-quickstart-winequality covers Spark environment setup, Project creation, and model training.
dominodatalab/spark-quickstart-winequality
Loading repository data...
Model Registry
Model Registry covers how Domino tracks and manages the models these notebooks register.reference-project-domino-mlflow-supported-models holds notebooks that register XGBoost, Sklearn, and Python-function MLflow models, two of them with integrated model monitoring.
dominodatalab/reference-project-domino-mlflow-supported-models
Loading repository data...
Flows
Get started with Flows builds a basic flow and then a training flow.domino-ai-flows is the template repository you fork to build a training flow.
dominodatalab/domino-ai-flows
Loading repository data...
Run an example in Domino
1
Fork the repository if you plan to change it
Fork it to your own Git provider so you have somewhere to push commits. Skip this if you only want to read and run the code as published.
2
Add Git credentials if you'll write commits back
Domino reads a public repository without credentials. Writing commits back to the remote, using a private repository, or connecting over SSH each need Git credentials stored in your Domino account.
3
Create a Git-based Project that points at it
Follow Create a Project and give Domino the repository URL.
4
Open a Workspace or run a Job
Work through the code interactively in a Workspace, or run one of its scripts as a Job.
Related
- Getting started with Domino: a guided tutorial that builds and deploys a transaction review App, from a Git-based Project through an authenticated model endpoint.
- The python-domino library: the Python bindings for the Domino APIs, whose repository carries its own example code.
- Blueprints