Data Flywheel Foundational Blueprint Helm Chart

The NVIDIA Data Flywheel Foundational Blueprint Helm Chart simplifies the deployment process by providing a comprehensive, production-ready Kubernetes deployment that:

• Automatically configures all required infrastructure services (Elasticsearch, MongoDB, Redis)
• Integrates with NeMo microservices
• Manages secrets and configurations
• Orchestrates the complete Data Flywheel workflow stack with scalable worker deployments and monitoring capabilities.

Built on the NVIDIA NeMo Microservices platform, this blueprint enables developers to quickly bootstrap a control-plane service that continuously improves fine-tuned AI models by automating end-to-end workflows with NeMo Customizer and NeMo Evaluator to discover and promote more efficient model variants using real logging and tool-calling data.

What the Container Does

• API Service: Hosts a FastAPI-based REST API (on /api) for managing flywheel runs, datasets, and model evaluation workflows.
• Workflow Orchestration: Runs Celery workers to execute distributed, multi-stage workflows for data collection, dataset creation, model deployment, evaluation, and promotion.
• Integration: Connects to MongoDB (for metadata), Elasticsearch (for logging and data), and Redis (for task queueing).
• Model Management: Automates the process of spinning up, evaluating, and managing NeMo Inference Microservices (NIMs) and related model artifacts.
• Data Handling: Handles ingestion, deduplication, splitting, and uploading of datasets for training, validation, and evaluation.
• Monitoring: Optionally runs Flower for Celery task monitoring and can be extended with Kibana for log/data visualization.

Resources

Documentation

Readme | Helm Installation Guide | Quickstart

Code

Source Code | Notebooks

Ethical Considerations

NVIDIA believes Trustworthy AI is a shared responsibility, and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure the models meet requirements for the relevant industry and use case and address unforeseen product misuse. For more detailed information on ethical considerations for the models, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards. Please report security vulnerabilities or NVIDIA AI concerns here.

License

Use of the models in this blueprint is governed by the NVIDIA AI Foundation Models Community License

Governing Terms

Software and Blueprint

• Software: NVIDIA Software License Agreement
• AI Products: Product-Specific Terms for NVIDIA AI Products
• Dataset: NVIDIA Asset License Agreement

Models

• AI Foundation Models: AI Foundation Models Community License Agreement
• Meta Llama 3.1 70B Instruct: Llama 3.1 Community License Agreement
• NVIDIA Llama 3.2 Embedding Models: Llama 3.2 Community License Agreement

Built with Llama.