Published March 8, 2023
| Version v3
Software
Open
Register Tiling for Unstructured Sparsity in Neural Network
- 1. University of Toronto
- 2. McMaster University
Description
This artifact generates figures of the submitted draft of “Register Tiling for Unstructured Sparsity in Neural Network Inference”. The artifact generates graphs of the paper for an intel processor with 20 cores. The tested dataset is a collection of matrices from deep learning matrix collection (DLMC). The artifact is packaged as a singularity container and is set up on a supercomputer, Niagara in the Digital Research Alliance of Canada. For the latest version of the artifact, please check https://github.com/SpRegTiling/sparse-register-tiling
Files
LICENSE.txt
Files
(2.7 GB)
Name | Size | Download all |
---|---|---|
md5:b7c6cd20bc330f0a0820fcef6a3852fc
|
2.6 GB | Download |
md5:cb4eb0f370d854acc308d093a80b0e9c
|
80.2 MB | Download |
md5:ffdf47e2d52f9398bae493722868e360
|
1.1 kB | Preview Download |
md5:bb2f435a7660e6fcf535d2da82fec148
|
3.0 kB | Preview Download |
md5:32bbf2ddd917810b80819da304ec5185
|
54.1 kB | Preview Download |