Erzmann, DavidDavidErzmannDittmer, SörenSörenDittmerHarms, HenrikHenrikHarmsMaaß, PeterPeterMaaß2025-07-112025-07-112023-08-01https://media.suub.uni-bremen.de/handle/elib/22192https://doi.org/10.26092/elib/4171We present and publish the DL4TO software library – a Python library for three-dimensional topology optimization. The framework is based on PyTorch and allows easy integration with neural networks. The library fills a critical void in the current research toolkit on the intersection of deep learning and topology optimization. We present the structure of the library’s main components and how it enabled the incorporation of physics concepts into deep learning models.enAlle Rechte vorbehaltenTopology optimizationDeep learningSoftware library600 Technik, Medizin, angewandte Wissenschaften::600 TechnikDL4TO : A Deep Learning Library for Sample-Efficient Topology OptimizationText::Konferenzveröffentlichung::Tagungsband::Konferenzbeitrag10.26092/elib/4171urn:nbn:de:gbv:46-elib221922