SaeedLab/MolDeBERTa
收藏Hugging Face2026-04-28 更新2026-03-29 收录
下载链接：
https://hf-mirror.com/datasets/SaeedLab/MolDeBERTa
下载链接
链接失效反馈官方服务：
资源简介：
该存储库提供了两个来自PubChem的大规模SMILES字符串数据集，分别包含10百万和123百万个分子。10M子集对应于ChemBERTa研究中使用的数据集，用于直接基准测试。123M子集则代表完整的PubChem数据库，通过处理完整的PubChem数据集构建而成。这些数据集用于MolDeBERTa预训练，旨在加速材料和药物发现。
This repository provides two large-scale datasets of SMILES strings derived from PubChem, available in two configurations: 10M and 123M. The 10M subset comprises 10 million molecules sampled from PubChem, corresponding exactly to the dataset used in the ChemBERTa study. The 123M dataset comprises 123 million molecules, representing the PubChem database, constructed by processing the full PubChem dataset using the SMILES strings. These datasets are used for MolDeBERTa pretraining to accelerate material and drug discovery.
提供机构：
SaeedLab
