LIMIT: Less Is More for Instruction Tuning Across Evaluation Paradigms

Aditi Jha, Sam Havens, Jeremy Dohmann, Alexander Trott, Jacob Portes

Keywords: large language models, instruction tuning, dataset size, small dataset, style alignment

TL;DR: We find that a carefully curated small dataset (1-6k) is sufficient for finetuning LLMs, to achieve good performance on both (1) traditional NLP benchmarks and (2) model-based evaluation.

Abstract: Large Language Models are traditionally finetuned on large instruction datasets. However recent studies suggest that small, high-quality datasets can suffice for general purpose instruction following. This lack of consensus surrounding finetuning best practices is in part due to rapidly diverging approaches to LLM evaluation. In this study, we ask whether a small amount of diverse finetuning samples can improve performance on both traditional perplexity-based NLP benchmarks, and on open-ended, model-based evaluation. We finetune open-source MPT-7B and MPT-30B models on finetuning datasets of various sizes ranging from 1k to 60k samples. We find that subsets of 1k-6k instruction finetuning samples are sufficient to achieve good performance on both (1) traditional NLP benchmarks and (2) model-based evaluation. Finally, we show that mixing textbook-style and open-ended QA finetuning datasets optimizes performance on both evaluation paradigms.

Submission Number: 47