JULAIN-Talk-Thijs-Vogels-9-June-2022.md

@@ -31,11 +31,11 @@ PowerSGD can yield communication savings of > 100x and was used successfully to
 Thijs is a PhD student at EPFL’s Machine Learning & Optimization Laboratory under Martin Jaggi.
 He works on developing and understanding practical optimization algorithms for large-scale distributed training of deep learning models.
  
-https://arxiv.org/pdf/1905.13727.pdf
+*Readings:*
+* [RelaySum for Decentralized Deep Learning on Heterogeneous Data](https://arxiv.org/pdf/2110.04175.pdf), NeurIPS 2020 
+* [Practical Low-Rank Communication Compression in Decentralized Deep Learning](https://arxiv.org/pdf/2008.01425.pdf  ), NeurIPS 2020 
+* [PowerSGD: Practical Low-Rank Gradient Compression for Distributed Optimization](https://arxiv.org/pdf/1905.13727.pdf), NeurIPS 2019 
 
-https://arxiv.org/pdf/2008.01425.pdf  
-
-https://arxiv.org/pdf/2110.04175.pdf
 
 
 ---