Optimizing Training Settings for Explainable Chemprop Model using Layer-Wise Relevance Propagation

This repository contains data for the thesis titled "Optimizing Training Settings for Explainable Chemprop Model Using Layer-Wise Relevance Propagation." Each zip file is associated with 1 target and includes the following: Datasets: All datasets associated with the specified target. Metrics: Evaluations of the model's performance, faithfulness, and correctness. Models: Checkpoints for all models related to the given target. For more details about the study, check this GitHub repo: https://github.com/DinhLongHuynh/lrp_chemprop本仓库收录了题为《基于逐层相关性传播（Layer-Wise Relevance Propagation）优化可解释性Chemprop模型训练配置》的学术论文配套数据。每个压缩文件对应一个目标任务，并包含以下内容： 数据集：与指定目标任务相关的全部数据集。 评估指标：涵盖模型性能、忠实性与正确性的模型评估结果。 模型文件：与给定目标任务相关的全部模型的检查点文件。 如需了解该研究的更多细节，请访问此GitHub仓库：https://github.com/DinhLongHuynh/lrp_chemprop