Duration: 0:00

455 views

12

Developing with Unreal Engine is complex. Information is found across tutorials, user manuals, API documentation, as well as within the source code itself. Discover the groundbreaking capabilities of Retrieval Augmented Generation (RAG), which simplifies data retrieval through a user-friendly chatbot interface.

This demo is powered by an NVIDIA GPU-accelerated architecture with an AI foundational Large Language Model, CodeLlama 34B, and a vector datastore of information on Unreal Engine 5.

Experience the future of information retrieval with RAG, adapting seamlessly to evolving data without the expense and effort of retraining or fine-tuning models.

Start building a chatbot: https://www.nvidia.com/en-us/launchpa...
Learn more about deploying generative AI with NVIDIA NIM: https://ai.nvidia.com/
Explore enterprise solutions for game development: https://www.nvidia.com/game-dev

0:00 - LLMs and RAG for game development
0:49 - Building an Unreal Engine 5 RAG Chatbot
1:05 - Chatbot architecture
1:30 - Fine-tuning with RAG
2:09 - Conclusion