We are developing new features based on your feedback - please stay tuned!
5 results (0.010 seconds)
1) AGGREGATHOR: Byzantine Machine Learning via Robust Gradient Aggregation
Georgios Damaskinos, El Mahdi El Mhamdi, Rachid Guerraoui, Arsany Guirguis, Sebastien Rouault
Where published: SysML'19
Article: PDF
Artifact DOI: 10.5281/zenodo.2548779
Unified artifact appendix: Link
Artifact before standardization: GitHub
Standardized CK workflow: Link (ReproIndex)
Standardized CK pipelines (programs):
Reproducible results: Open review via GitHub issues
Some results replicated:
Reproducible methodology: ACM and cTuning
2) Beyond Data and Model Parallelism for Deep Neural Networks
Zhihao Jia, Matei Zaharia, Alex Aiken
Where published: SysML'19
Article: PDF
Artifact DOI: 10.5281/zenodo.2549847
Unified artifact appendix: Link
Artifact before standardization: GitHub
Some results replicated:
Reproducible methodology: ACM and cTuning
3) Kernel machines that adapt to GPUs for effective large batch training
Siyuan Ma, Mikhail Belkin
Where published: SysML'19
Article: PDF
Artifact DOI: 10.5281/zenodo.2574996
Unified artifact appendix: Link
Artifact before standardization: GitHub
Some results replicated:
Reproducible methodology: ACM and cTuning
4) Optimizing DNN Computation with Relaxed Graph Substitutions
Zhihao Jia, James Thomas, Todd Warszawski, Mingyu Gao, Matei Zaharia, Alex Aiken
Where published: SysML'19
Article: PDF
Artifact DOI: 10.5281/zenodo.2549853
Unified artifact appendix: Link
Artifact before standardization: GitHub
Some results replicated:
Reproducible methodology: ACM and cTuning
5) Priority-based Parameter Propagation for Distributed DNN Training
Anand Jayarajan, Jinliang Wei, Garth Gibson, Alexandra Fedorova, Gennady Pekhimenko
Where published: SysML'19
Article: PDF
Artifact DOI: 10.5281/zenodo.2549852
Unified artifact appendix: Link
Artifact before standardization: GitHub
Standardized CK workflow: Link (ReproIndex)
Standardized CK pipelines (programs):
Reproducible results: https://github.com/ctuning/reproduce-sysml19-paper-p3/issues
Some results replicated:
Reproducible methodology: ACM and cTuning