Raw and aggregated evaluation results for the study Capability-Specific Degradation Patterns
in Quantized Small Language Models: seven open instruction-tuned small LLMs (1–4B parameters,
five architecture families) evaluated at FP16 and 4-bit (bitsandbytes NF4) across
six capabilities, for 84 controlled model × precision × benchmark evaluations.
This repository contains evaluation outputs… See the full description on the dataset page:
https://huggingface.co/datasets/Emil-7/llm-quant-degradation.
