赋能较小模型：利用思维链微调 LLaMA 和 Gemma 以应对乌克兰语考试任务
Empowering Smaller Models: Tuning LLaMA and Gemma with Chain-of-Thought for Ukrainian Exam Tasks
AI总结：
本文基于 ZNO-Eval，研究用2,032个逐步解答对 LLaMA 与 Gemma 进行参数高效思维链微调，以提升乌克兰语考试推理表现并超过多个较大模型。
摘要：
领先的大型语言模型已在标准化教育测试等推理密集型任务中展现出令人瞩目的能力。然而，在低资源环境中，它们往往需要大量训练，而相关基础设施又难以获得。小型或紧凑模型虽然效率更高，但通常对代表性不足的语言缺乏充分支持，从而在关键领域留下性能差距。本研究基于 ZNO-Eval 基准测试的发现，探索对紧凑的开放权重语言模型进行参数高效微调，以处理代表性不足的乌克兰语中的推理密集型任务的潜力。对 LLaMA 3.1（80亿参数）、LLaMA 3.2（30亿参数）和 Gemma 2（90亿参数）模型进行基于思维链解答的参数高效微调，在复杂匹配任务上带来了最高17.4%的适度测试分数提升，并且与仅基于答案字母进行微调相比，总体提升1.6%，同时提供了更强的可解释性和稳健性。此外，所提出的联合生成任务主题和逐步解答的微调方法，在匹配任务上优于标准思维链微调；由于引导模型回忆并应用与领域相关的信息，该方法相比最佳 LLaMA 3.2 模型带来5.4%的增益。将所得结果与 Qwen、DeepSeek R1、OpenAI o1 和 o3、Gemini 以及 Claude 等领先开放权重和专有模型的零样本评估进行对比，结果突出表明：在单张 A100 GPU 上，使用2,032个逐步解答以及2,000万至5,000万个可训练参数对 LLaMA 和 Gemma 模型进行微调，可使其表现超过 GPT-4o mini、Mistral Large 以及更大的开放权重模型。本研究还评估了将量化适配器与基础模型合并会如何影响生成质量。源代码和经过微调的模型可在 https://github.com/NLPForUA/ZNO 获取。
英文摘要：
Leading large language models have demonstrated impressive capabilities in reasoning-intensive tasks, such as standardized educational testing. However, they often require extensive training in low-resource settings with inaccessible infrastructure. Small or compact models, though more efficient, frequently lack sufficient support for underrepresented languages, leaving a performance gap in critical domains. This work explores the potential of parameter-efficient fine-tuning of compact open-weight language models to handle reasoning-intensive tasks in the underrepresented Ukrainian language, building on the findings of the ZNO-Eval benchmark. Parameter-efficient fine-tuning of LLaMA 3.1 (8 billion parameters), LLaMA 3.2 (3 billion parameters), and Gemma 2 (9 billion parameters) models on chain-of-thought solutions resulted in a modest test score improvement of up to 17.4% on complex matching tasks and 1.6% overall compared to tuning on answer letters alone, offering enhanced interpretability and robustness. In addition, the proposed tuning method with joint task topic and step-by-step solution generation outperforms standard chain-of-thought tuning in matching tasks and provides a 5.4% gain over the best LLaMA 3.2 model due to guiding the model to recall and apply domain-relevant information. Contrasting obtained results with zero-shot evaluations of leading open-weight and proprietary models such as Qwen, DeepSeek R1, OpenAI o1 and o3, Gemini, and Claude, highlight that fine-tuning LLaMA and Gemma models with 2,032 step-by-step solutions and 20 to 50 million trainable parameters on a single A100 GPU lets them outperform GPT-4o mini, Mistral Large, and larger open-weight models. This research also evaluates how merging the quantized adapter with the base model influences the generation quality. Source code and tuned models are available at https://github.com/NLPForUA/ZNO.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
