Access to European HPC Infrastructure
The GenAI-HEI project focuses on developing a fine-tuned Large Language Model (LLM) for education, tailored to generate and assess exam content for universities. Utilizing HPC resources, this multimodal AI model will process extensive educational datasets to automate exam creation and grading. This solution aims to address challenges in traditional exam design, improving efficiency, fairness, and scalability in academic assessment. We were granted access for HPC resources at LUMI-G partition with GPU nodes.
In collaboration with University of Donja Gorica (UDG), and with the support from HPC NCC Montenegro, we applied for the Open Call for Access to European HPC Infrastructure. EuroHPC offers up to two years of access to HPC resources to develop, test, or run code for free.
These open calls are availale for machine learning, high performance data analytics, compute-intensive simulations in engineering, life science or any other domain. The applications for access to EuroHPC JU supercomputers include calls for benchmark, development, regular access mode, extreme scale and AI and Data-intensive applications. We won a grant for development access which covers covers for node hours access to HPC infrastructure.
