6G Master Thesis: Resource Efficient Large Language Models at the Edge
| Reference number | |
| Coordinator | RISE Research Institutes of Sweden AB - RISE AB - Digitala System |
| Funding from Vinnova | SEK 100 000 |
| Project duration | January 2025 - June 2025 |
| Status | Completed |
| Venture | 6G - Competence supply |
| Call | 6G - Supervision of degree work |
Important results from the project
The project has largely met its objective, even though the final report has been delayed. It has contributed to increased knowledge about how LLMs can be optimized for edge environments and strengthened the collaboration between RISE and LTU within TinyML.
Expected long term effects
The project is expected to contribute to the development of energy-efficient AI solutions for 6G. It opens up new research avenues in quantized training and strengthens the infrastructure for future collaborative projects.
Approach and implementation
The student studied available research in the area, used code from a previous project and trained models on RISE´s GPU cluster. Supervision was provided weekly with RISE and biweekly with LTU. The project largely followed plan, but the final report is delayed until September 2025.