a curated repository for benchmarking and evaluating small, open-source language models (LLMs) optimized for edge computing environments (CPU-only or low-VRAM devices). The project focuses on reproducible, interpretable, and standardized evaluation of LLM capabilities across multiple domains .
raspberry-pi benchmarking reproducible-research vram textsummarization edgeai ollama ollama-api logicalreasoning knowledgecomparison deductivereasoning timestamplogging modelparameters systemprep reproduciblebenchmark ollama-models
-
Updated
Feb 3, 2026 - Python