research
∙
10/13/2020
Revisiting BFloat16 Training
State-of-the-art generic low-precision training algorithms use a mix of ...
share
research
∙
02/29/2020
Understanding the Downstream Instability of Word Embeddings
Many industrial machine learning (ML) systems require frequent retrainin...
research
∙
10/09/2019
PipeMare: Asynchronous Pipeline Parallel DNN Training
Recently there has been a flurry of interest around using pipeline paral...
research
∙
03/09/2018
High-Accuracy Low-Precision Training
Low-precision computation is often used to lower the time and energy cos...
research
∙
08/25/2017