Overview

The FAIR project aims to build “Minerva”, an Italian Large Language Model. This initiative focuses on creating robust evaluation frameworks for language models with specific attention to Italian language capabilities.

Role

  • WP6 - Evaluation: Responsible for Work Package 6 focusing on model evaluation
  • Zero/Few-shot Evaluation (T6.2): Developing and implementing evaluation methodologies for zero-shot and few-shot learning scenarios

Key Responsibilities

  • Design evaluation protocols for Italian language models
  • Implement zero-shot and few-shot testing frameworks
  • Benchmark performance across various NLP tasks
  • Ensure rigorous evaluation standards

Project Details

Year: 2024
Project: FAIR - Foundazione per l’Intelligenza Artificiale Responsabile
Focus: Italian LLM development, Model evaluation, Benchmark design