Description
The artifact contains a self contained docker image with the dataset and the software package that was used to generate the graphs in the paper. The artifact also contains the source files without the docker image. The artifact contains two main directories containing the auto-scheduler (written in Python) described in Section 06 in the paper, and the language extension to SparseLNR/TACO described in the Section 05 in the paper.
Data Documentation
The matrices and tensors used in the paper are contained within the artifact. These matrices and tensors obtained from the SuiteSparse Collection, Network Repository, Formidable Repository of Open Sparse Tensors and Tools, and 1998 DARPA Intrusion Detection Evaluation Dataset.
Provenance
https://github.com/adhithadias/tensor-schedules https://github.com/adhithadias/SparseLNR/blob/fix/sparsesched-multi-temp-multi-dim-temp/LICENSE
License
Index Terms
- Reproduction package of article "SparseAuto: An Auto-Scheduler for Sparse Tensor Computations Using Recursive Loop Nest Restructuring"