笔记本 iGPU 上用 SFT 与 QLoRA 微调 Gemma 4 Edge 模型生成 MATLAB 信号处理代码
Fine-Tuning Gemma 4 Edge Models for MATLAB Signal-Processing Code with SFT and QLoRA on a Laptop iGPU
AI 导读
研究者在笔记本集成 GPU 上对 Gemma 4 Edge 模型 E2B 和 E4B 进行监督微调(SFT),使用 66 个经检查的样本覆盖 11 个 MATLAB 信号处理子任务,采用 LoRA 和 QLoRA,并通过 OpenVINO 部署 int8 模型。
正文 · 原文
Abstract:Small language models that run on a laptop are attractive for writing signal-processing code offline, but they make systematic errors on digital signal processing (DSP) conventions. We apply supervised fine-tuning (SFT) to the Gemma 4 edge models E2B and E4B with 66 checked examples covering 11 signal-processing MATLAB subtasks, using low-rank adaptation (LoRA) and quantized LoRA (QLoRA) on the integrated GPU (iGPU) of a laptop, and deploy the 8-bit integer (int8) models with the adapters through OpenVINO. On held-out variants scored by deterministic checkers, the pass rate rises from 18/66 to 37/66 (LoRA) and 37/66 (QLoRA) for E2B and from 32/66 to 47/66 for E4B (QLoRA). Reading all 159 failing programs shows that the adapters remove most errors of convention, such as the sign of a lag, units and array shapes, but few errors in building signals and noise. Applied at run time, the adapter costs 17-19% of the generation speed.
| Comments: | 18 pages, 3 figures, 4 tables |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Software Engineering (cs.SE); Signal Processing (eess.SP) |
| Cite as: | arXiv:2611.23205 [cs.LG] |
| (or arXiv:2611.23205v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.23205 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mehmet Altan Toksöz [view email]
[v1]
Tue, 17 Nov 2026 18:47:37 UTC (58 KB)
来源:arXiv cs.AI · arxiv.org