🔬 LOOM-Scope: LOng-cOntext Model evaluation framework
🎯 Framework Overview
LOOM-Scope is a comprehensive and efficient framework for long-context model evaluation, featuring a robust and multifaceted evaluation framework meticulously crafted through 6 core cognitive dimensions and 15 specialized benchmark suites, aiming to deliver a comprehensive, nuanced assessment of large language models (LLMs) in long-context scenarios.
Core Cognitive Dimensions form the… See the full description on the dataset page: https://huggingface.co/datasets/AmamiSora/LOOM_Data.
