8 accepted papers
MARS has recently emerged as a state-of-the-art optimizer, consistently outperforming AdamW in large language model (LLM) training. It modifies the momentum-based variance reduction (MVR) update by introducing a multiplicative coefficient $\gamma$, which scales the momentum correction term. However,…
We provide the first proof of convergence for normalized error feedback algorithms across a wide range of machine learning problems. Despite their popularity and efficiency in training deep neural networks, traditional analyses of error feedback algorithms rely on the smoothness assumption that doe…
Federated learning is an emerging framework for collaborative machine-learning on devices which do not want to share local data. State-of-the art methods in federated learning reduce the communication frequency, but are not guaranteed to converge to the optimal model parameters. These methods also e…
Zeroth-order methods have become important tools for solving problems where we have access only to function evaluations. However, the zeroth-order methods only using gradient approximations are n times slower than classical first-order methods for solving n-dimensional problems. To accelerate the co…
With the increasing scale of machine learning tasks, it has become essential to reduce the communication between computing nodes. Early work on gradient compression focused on the bottleneck between CPUs and GPUs, but communication-efficiency is now needed in a variety of different system architec…
Noise is inherited in many optimization methods such as stochastic gradient methods, zeroth-order methods and compressed gradient methods. For such methods to converge toward a global optimum, it is intuitive to use large step-sizes in the initial iterations when the noise is typically small compare…
The veritable scale of modern data necessitates information compression in parallel/distributed big-data optimization. Compression schemes using memory-based error compensation have displayed superior performance in practice, however, to date there are no theoretical explanations for these observed…
Distributed training of massive machine learning models, in particular deep neural networks, via Stochastic Gradient Descent (SGD) is becoming commonplace. Several families of communication-reduction methods, such as quantization, large-batch methods, and gradient sparsification, have been proposed.…