![Edision argus mini combo receiver user manual](https://kumkoniak.com/82.jpg)
![garfield pdf download garfield pdf download](https://cdn.slidesharecdn.com/ss_thumbnails/garfield-fat-cat-3pack-17-211023231704-thumbnail-4.jpg)
Following the standard parameter server model, we assume that a minority of worker machines can be controlled by an adversary and behave arbitrarily. We present AGGREGATHOR, a framework that implements state-of-the-art robust (Byzantine-resilient) distributed stochastic gradient descent. In particular, (a) Byzantine resilience, unlike crash resilience, induces an accuracy loss, and (b) the throughput overhead comes much more from communication (70%) than from aggregation. Our evaluation highlights several interesting facts about the cost of Byzantine resilience. We report on our evaluation of Garfield with different (a) baselines, (b) ML models (e.g., ResNet-50 and VGG), and (c) hardware infrastructures (CPUs and GPUs).
![garfield pdf download garfield pdf download](https://tunovelaligera.net/en/wp-content/uploads/sites/3/2021/07/Garfield-What-Leftovers-by-Jim-Davis.jpg)
Our implementation supports full-stack computations on both CPUs and GPUs. We integrate Garfield with two widely-used ML frameworks, TensorFlow and PyTorch, while achieving transparency: applications developed with either framework do not need to change their interfaces to be made Byzantine resilient. On the other hand, Garfield uses statistically-robust gradient aggregation rules (GARs) to achieve resilience against Byzantine workers. Following the classical server/worker architecture, Garfield replicates the parameter server while relying on the statistical properties of stochastic gradient descent to keep the models on the correct servers close to each other. Garfield leverages ML specificities to make progress despite consensus being impossible in such an asynchronous, Byzantine environment. We present Garfield, a system that provably achieves Byzantine resilience in ML applications without assuming any trusted component nor any bound on communication or computation delays.
![garfield pdf download garfield pdf download](https://image.slidesharecdn.com/garfield-fat-cat-3pack-17-211023231704/85/downloadpdf-garfield-fat-cat-3pack-17-2-320.jpg)
Byzantine Machine Learning (ML) systems are nowadays vulnerable for they require trusted machines and/or a synchronous network.
![Edision argus mini combo receiver user manual](https://kumkoniak.com/82.jpg)