← Back to all news
arXiv AI By Avik Bhatnagar, Federico Nicolas Peccia, Oliver Bringmann

TASTE: Throughput-Aware Batch Size Tuning for On-Device Edge Learning

Read the original on arXiv AI →

The paper presents TASTE, a method that uses Bayesian optimization to tune batch size for on‑device edge learning, aiming to maximize hardware throughput while preserving accuracy. Experiments on devices like the Raspberry Pi 4 show that the tuned batch size, combined with gradient accumulation and linear learning‑rate scaling, can double training throughput compared to using the maximum batch size. In online continual learning, the optimal batch size also helps balance stability and plasticity, reducing catastrophic forgetting without sacrificing efficiency.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv Machine Learning
Jul 23

Leveraging ECRAM for Edge Continual Learning

arXiv:2607. 19661v1 Announce Type: cross Abstract: Several edge computing platforms, such as autonomous vehicles and smart sensing devices, need to adapt to dynamic environments in real time by learning from new data in the field.

By Nabila Tasnim, Haoran Liu, Qing Cao, Saugata Ghose
agents
More like this →
arXiv AI
Jul 21

SelectInfer: Selective Neuron Loading and Computation for On-Device LLMs

arXiv:2607. 18081v1 Announce Type: cross Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities across a range of Natural Language Processing (NLP) tasks, but their high computational and memory demands pose significant challenges for deployment on resource-constrained edge devices.

By Huzaifa Shaaban Kabakibo, Eric Schniedermeyer, Artem Burchanow, Lin Wang
llmsnlpfine-tuningefficiency
More like this →
Hugging Face Trending Papers
Aug 27

Parameter Efficient Continual Learning for Sparse Event-Based Transformers

The paper introduces sLoTh, a parameter‑efficient continual learning framework for sparse event‑based vision transformers. By freezing the backbone and limiting plasticity to low‑rank attention updates (seLoRA) and shared neuronal threshold modulation, sLoTh adapts to new tasks while updating less than 1% of the parameters and avoiding replay buffers. Experiments on CIFAR‑100, Tiny‑ImageNet, ImageNet‑100, and ImageNet‑R show competitive rehearsal‑free performance across up to 100 tasks and achieve roughly 6.5× lower energy consumption than dense vision transformers.

llmsroboticsfine-tuning
More like this →
arXiv Computer Vision
Aug 28

Parameter Efficient Continual Learning for Sparse Event-Based Transformers

The paper introduces sLoTh, a parameter‑efficient continual learning framework for sparse event‑based vision transformers. sLoTh freezes the backbone and limits plasticity to low‑rank attention updates (seLoRA) and shared neuronal threshold modulation, updating less than 1% of parameters without replay buffers. Experiments on CIFAR‑100, Tiny‑ImageNet, ImageNet‑100, and ImageNet‑R show competitive rehearsal‑free performance across up to 100 tasks while achieving roughly 6.5× lower energy consumption than dense vision transformers.

By Vaishnavi Nagabhushana, Kartikay Agrawal, Ayon Borthakur
llmsroboticsfine-tuning
More like this →