Neural Architecture Search (NAS) automates network design, but conventional methods demand substantial computational resources. We propose a closed-loop pipeline leveraging large language models (LLMs) to iteratively generate, evaluate, and refine convolutional neural network architectures for image classification on a single consumer-grade GPU without LLM fine-tuning. Central to our approach is a historical feedback memory inspired by Markov chains: a sliding window of $K{=}5$ recent improvement attempts keeps context size constant while providing sufficient signal for iterative learning. Unlike prior LLM optimizers that discard failure trajectories, each history entry is a structured diagnostic triple -- recording the identified problem, suggested modification, and resulting outcome -- treating code execution failures as first-class learning signals. A dual-LLM specialization reduces per-call cognitive load: a Code Generator produces executable PyTorch architectures while a Prompt Improver handles diagnostic reasoning. Since both the LLM and architecture training share limited VRAM, the search implicitly favors compact, hardware-efficient models suited to edge deployment. We evaluate three frozen instruction-tuned LLMs (${\leq}7$B parameters) across up to 2000 iterations in an unconstrained open code space, using one-epoch proxy accuracy on CIFAR-10, CIFAR-100, and ImageNette as a fast ranking signal. On CIFAR-10, DeepSeek-Coder-6.7B improves from 28.2% to 69.2%, Qwen2.5-7B from 50.0% to 71.5%, and GLM-5 from 43.2% to 62.0%. A full 2000-iteration search completes in ${\approx}18$ GPU hours on a single RTX~4090, establishing a low-budget, reproducible, and hardware-aware paradigm for LLM-driven NAS without cloud infrastructure.

翻译：神经架构搜索（NAS）实现了网络设计的自动化，但传统方法需要大量的计算资源。我们提出了一种闭环流程，利用大型语言模型（LLM）在单个消费级GPU上，无需对LLM进行微调，即可迭代地生成、评估和优化用于图像分类的卷积神经网络架构。我们方法的核心是受马尔可夫链启发的历史反馈记忆：一个包含最近$K{=}5$次改进尝试的滑动窗口，在保持上下文大小恒定的同时，为迭代学习提供足够的信号。与先前丢弃失败轨迹的LLM优化器不同，每个历史条目都是一个结构化的诊断三元组——记录识别出的问题、建议的修改以及产生的结果——将代码执行失败视为首要的学习信号。双LLM专业化降低了每次调用的认知负荷：代码生成器负责生成可执行的PyTorch架构，而提示改进器则处理诊断推理。由于LLM和架构训练共享有限的VRAM，搜索过程会隐式地偏向于适合边缘部署的紧凑、硬件高效的模型。我们在无约束的开放代码空间中，使用CIFAR-10、CIFAR-100和ImageNette数据集上的单轮代理准确率作为快速排序信号，评估了三个冻结的指令微调LLM（参数量${\leq}7$B）在多达2000次迭代中的表现。在CIFAR-10上，DeepSeek-Coder-6.7B的准确率从28.2%提升至69.2%，Qwen2.5-7B从50.0%提升至71.5%，GLM-5从43.2%提升至62.0%。一次完整的2000次迭代搜索在单个RTX~4090上约需${\approx}18$个GPU小时完成，这为无需云基础设施的LLM驱动NAS建立了一种低预算、可复现且硬件感知的新范式。