---
base_model:
- OnomaAIResearch/Illustrious-XL-v2.0
tags:
- quantization
quantized_by: btaskel
pipeline_tag: text-to-image
---
From civitai/fannon:
https://huggingface.co/OnomaAIResearch/Illustrious-XL-v2.0

Based on my experience, Q4_K_S and Q4_K_M are usually the balance points between model size, quantization, and speed.

In some benchmarks, selecting a large-parameter high-quantization LLM tends to perform better than a small-parameter low-quantization LLM.

根据我的经验，通常Q4_K_S、Q4_K_M是模型尺寸/量化/速度的平衡点

在某些基准测试中，选择大参数高量化模型往往比选择小参数低量化模型表现更好。

----------

You have amazing hardware?!

I'm using 16GB DDR RAM and an R5 5600 for interest-based quantization work, along with a 50Mbps bandwidth speed. It might not be able to quantize models with higher parameters.

您有惊人的硬件？？！

我正在使用16G DDR内存和R5 5600进行基于兴趣的量化工作，以及50Mbps的带宽速度，可能会无法为更高参数的模型进行量化。