Currently the performance tests are only semi-automated and only a subset of algorithms are included, but largely the process is manual. An automatic benchmark suite can be used to detect regressions or improvements in critical components of the system. This GSoC would begin by building a benchmark infrastructure to conduct experiments that compare different choices in critical parts like Sparsity thresholds for Dense/Sparse datasets, Dataset Size, Memory Limit and Hyper parameters decisions.


Sai Krishna Kalyan


  • Nakul Jindal