How To Adapt AI for Low-Resource Languages with NVIDIA Nemotron

​Join NVIDIA and Dicta to explore how Sovereign AI initiatives can empower local innovation through advanced language model adaptation. This session highlights how Dicta built a Hebrew reasoning model using the Nemotron architecture and NVIDIA’s NeMo framework—showing how open models, transparent data, and efficient training tools can accelerate AI development for low‑resource languages.

​We’ll cover the main practices for customizing foundation models, managing data quality and safety, and deploying domain‑specific reasoning systems. The session will include a brief Q&A and provide references and tools to help you begin your own model adaptation journey.