Lipinski Drug Set (max_seq = 113, vocab=30)

Lipinski filtered, no charge, no radical, atom list limited (Molecules can only have 'C H O N S P N F Cl Br I' atoms.), and semantically filtered (SELFIES cannot have '+' or '-' in token), onehot and labeled with stoi/itos respectively pickle dictionary.<br>

本数据集经Lipinski规则筛选，无电荷、无自由基，原子类型受限（分子仅可包含C、H、O、N、S、P、F、Cl、Br、I原子），同时完成语义筛选（SELFIES的Token中不得包含'+'或'-'符号）；数据已完成独热编码，并分别通过stoi（string-to-index，字符转索引）与itos（index-to-string，索引转字符）进行标记，最终存储为pickle序列化字典。