tapestry

Project Tapestry aims to give every nation and participant frontier AI they can call their own — uniting a global consortium to train a shared frontier model fa

What is tapestry?

Project Tapestry aims to give every nation and participant frontier AI they can call their own — uniting a global consortium to train a shared frontier model family from which partners build and own sovereign models aligned to their national, socio-cultural, and industrial needs.

Overview of tapestry

What it is

Project Tapestry is a global consortium of partner organizations and individuals who bring expertise, data sets, and compute resources together to build a new foundation model family, one that is trained on a larger and more diverse corpus than ever before. Its aim is to enable truly sovereign AI by ensuring that ownership of data and compute remains with partners, and that they can build sovereign derivative models that they own, based on the consortium-trained base models and built using Tapestry's open-source training platform.

Key features

  • Global consortium uniting partners to train a shared frontier model family.
  • Sovereign AI: partners own derivative models built from consortium base models.
  • Open-source training platform.
  • Cultural alignment techniques (e.g., Inglehart–Welzel Cultural Map).
  • Consortium training with federated learning and LLM pre/post-training methods.
  • Data responsibly used with permissive use and privacy requirements.
  • Contribution mechanism for new ideas and technologies.

Use cases

  • Nations and organizations building sovereign AI models aligned to their national, socio-cultural, and industrial needs.
  • Contributing diverse datasets to improve model diversity and representation.
  • Developing cultural alignment and re-alignment techniques.
  • Participating in consortium training as a partner with data or compute resources.
  • Experimenting with federated learning and distributed training approaches.

Getting started

The recommended quick start command for setting up the development environment is: shell make one-time-setup This command installs uv and fswatch (if HomeBrew is available) and then installs the project's Python dependencies. If you cannot use GNU make, see the repository README for alternative installation steps using uv directly.

License and pricing

The license is not stated in the README or metadata. The project's metadata lists the pricing type as open_source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/the-ai-alliancetapestry.svg)](https://yunodb.com/tools/the-ai-alliancetapestry?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/the-ai-alliancetapestry?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/the-ai-alliancetapestry.svg" alt="Featured on YunoDB" height="20"></a>