研究深度学习训练中超参数对病理图像模型可复现性的影响
Hyperparameter Optimization and Reproducibility in Deep Learning Model Training
- 系统测试不同超参数与数据增强对模型性能影响
- 0.7-0.8的随机裁剪率表现最佳，学习率低于5.0e-5会降低性能
- 提供可复现数字病理模型的实用配置建议
可复现性仍是数字病理学基础模型训练中的关键挑战，常受软件随机性、硬件非确定性和超参数报告不一致的影响。我们基于QUILT-1M数据集训练了一个CLIP模型，并在三个下游病理数据集（PatchCamelyon、LC25000-Lung和LC25000-Colon）上系统评估了不同超参数设置和增强策略的影响。尽管运行间存在差异，但发现清晰趋势：随机裁剪比例0.7–0.8优于更激进（0.6）或保守（0.9）的设置；无本地损失的分布式训练提升了稳定性；学习率低于5.0e-5在所有数据集上均导致性能下降。其中，LC25000（结肠）数据集提供了最可复现的基准。结果表明，可复现性不仅依赖透明记录，更取决于精心设计的实验配置，本文提出实用规则以指导未来可复现数字病理基础模型的开发。
原文摘要 · Abstract (English)
Reproducibility remains a critical challenge in foundation model training for histopathology, often hindered by software randomness, hardware non-determinism, and inconsistent hyperparameter reporting. To investigate these issues, we trained a CLIP model on the QUILT-1M dataset and systematically evaluated the impact of different hyperparameter settings and augmentation strategies across three downstream histopathology datasets (PatchCamelyon, LC25000-Lung, and LC25000-Colon). Despite variability across runs, we identified clear trends: RandomResizedCrop values of 0.7-0.8 outperformed more aggressive (0.6) or conservative (0.9) settings, distributed training without local loss improved stability, and learning rates below 5.0e-5 consistently degraded performance across all datasets. The LC25000 (Colon) dataset consistently provided the most reproducible benchmark. These findings highlight that reproducibility in computational pathology depends not only on transparent documentation but also on carefully chosen experimental configurations, and we provide practical rules to guide future efforts in developing reproducible foundation models for digital pathology.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。