Case Study: Canonical builds its first GenAI model with Ubuntu Data Science Stack

Canonical builds its first GenAI model in 20 minutes with Ubuntu DSS

The customer, Canonical, sought to make generative AI and large language model development more accessible for both beginners and experienced practitioners, eliminating the complex setup typically involved. They utilized their own product, Ubuntu's Data Science Stack (DSS), to provide a simplified foundation for this technical challenge.

The solution implemented by Ubuntu involved using DSS to quickly provision a GPU-ready environment, allowing a user to seamlessly run Hugging Face's LLM training course. This removed hours of configuration headaches, enabling focused experimentation and the successful fine-tuning of a model within approximately 20 minutes. The result demonstrated how Ubuntu's tool lowers barriers to entry, making advanced AI development straightforward and accessible.

Canonical