Explore projects

Week 2 Assignment 2

Week 2 Assignment 2

Course in Deep Learning at HSLU

Serial Crystallography (SX) involves the processing of thousands of diffraction patterns in random orientations. To compile a complete data set, these patterns must be indexed, integrated, and merged. Herein, we introduce TORO (Torch-Powered Robust Optimization) Indexer, a robust, adaptable, and performant indexing algorithm. Developed using the PyTorch framework, TORO Indexer can operate on GPUs, CPUs, TPUs, and other accelerators supported by PyTorch, ensuring it adapts to a broad range of computational setups. On a modern NVIDIA A100 GPU, TORO achieves an indexing speed of 250 images/s.