Federated learning is an emerging distributed machine learning framework for privacy
preservation. However, models trained in federated learning usually have worse …

Machine learning (ML) models are widely used in many important domains. For efficiently
processing these computational-and memory-intensive applications, tensors of these …

Gradient compression is a widely-established remedy to tackle the communication
bottleneck in distributed training of large deep neural networks (DNNs). Under the error …

Federated learning (FL) enables multiple clients to collaboratively train a shared model, with
the help of a parameter server (PS), without disclosing their local datasets. However, due to …

While training can mostly be accelerated by reducing the time needed to propagate neural
gradients back throughout the model, most previous works focus on the quantization/pruning …

Gradient compression is an effective technique for improving the efficiency of distributed
training. However, introducing gradient compression can reduce model accuracy and …

In classical federated learning, the clients contribute to the overall training by communicating
local updates for the underlying model on their private data to a coordinating server …

Artificial intelligence (AI) research and market have grown rapidly in the last few years and
this trend is expected to continue with many potential advancements and innovations in this …

Communication overhead severely hinders the scalability of distributed machine learning
systems. Recently, there has been a growing interest in using gradient compression to …

Deep learning recommendation systems at scale have provided remarkable gains through
increasing model capacity (ie wider and deeper neural networks), but it comes at significant …