Skip to yearly menu bar Skip to main content


PandaLM: An Automatic Evaluation Benchmark for LLM Instruction Tuning Optimization

Yidong Wang · Zhuohao Yu · Wenjin Yao · Zhengran Zeng · Linyi Yang · Cunxiang Wang · Hao Chen · Chaoya Jiang · Rui Xie · Jindong Wang · Xing Xie · Wei Ye · Shikun Zhang · Yue Zhang

Halle B #272
[ ]
Tue 7 May 1:45 a.m. PDT — 3:45 a.m. PDT


Instruction tuning large language models (LLMs) remains a challenging task, owing to the complexity of hyperparameter selection and the difficulty involved in evaluating the tuned models. To determine the optimal hyperparameters, an automatic, robust, and reliable evaluation benchmark is essential. However, establishing such a benchmark is not a trivial task due to the challenges associated with evaluation accuracy and privacy protection. In response to these challenges, we introduce a judge large language model, named PandaLM, which is trained to distinguish the superior model given several LLMs. PandaLM's focus extends beyond just the objective correctness of responses, which is the main focus of traditional evaluation datasets. It addresses vital subjective factors such as relative conciseness, clarity, adherence to instructions, comprehensiveness, and formality. To ensure the reliability of PandaLM, we collect a diverse human-annotated test dataset, where all contexts are generated by humans and labels are aligned with human preferences. Our findings reveal that PandaLM-7B offers a performance comparable to both GPT-3.5 and GPT-4. Impressively, PandaLM-70B surpasses their performance. PandaLM enables the evaluation of LLM to be fairer but with less cost, evidenced by significant improvements achieved by models tuned through PandaLM compared to their counterparts trained with default Alpaca's hyperparameters. In addition, PandaLM does not depend on API-based evaluations, thus avoiding potential data leakage.

Chat is not available.