Tootfinder

My contribution to #footpathFriday , just fresh from the press ;-)
I went to the local hill / mountain today to escape the fog / clouds. Well and also to see whether the fog would create some nice mood to capture along the way that I'm walking rather often.
And indeed it was a very pleasant experience to walk into the fog and breaking through it to see some blue sky.

Zugebaute #Batteriespeicherkapazität pro Monat in #Deutschland mit Stand vom 22.12.2025.
378 Datensätze ohne #Batteriespeichertechnologie herausgefil…

So... awkward. 😜
#betweentwoferns #humor

Firing up my Fat Mac so that I can install these babies and get out on the internets and see the e-worlds.... soon I'll be cruising at A 62-baud ...
#Apple #eWorld

Scaling Vision Transformers: Evaluating DeepSpeed for Image-Centric Workloads
Huy Trinh, Rebecca Ma, Zeqi Yu, Tahsin Reza
https://arxiv.org/abs/2602.21081 https://arxiv.org/pdf/2602.21081 https://arxiv.org/html/2602.21081
arXiv:2602.21081v1 Announce Type: new
Abstract: Vision Transformers (ViTs) have demonstrated remarkable potential in image processing tasks by utilizing self-attention mechanisms to capture global relationships within data. However, their scalability is hindered by significant computational and memory demands, especially for large-scale models with many parameters. This study aims to leverage DeepSpeed, a highly efficient distributed training framework that is commonly used for language models, to enhance the scalability and performance of ViTs. We evaluate intra- and inter-node training efficiency across multiple GPU configurations on various datasets like CIFAR-10 and CIFAR-100, exploring the impact of distributed data parallelism on training speed, communication overhead, and overall scalability (strong and weak scaling). By systematically varying software parameters, such as batch size and gradient accumulation, we identify key factors influencing performance of distributed training. The experiments in this study provide a foundational basis for applying DeepSpeed to image-related tasks. Future work will extend these investigations to deepen our understanding of DeepSpeed's limitations and explore strategies for optimizing distributed training pipelines for Vision Transformers.
toXiv_bot_toot

Folks, our very own @… has his new book, 99th Day: A Warning About Technology, coming out next Monday.
https://gerrymcgovern.com/books/99th-day/<…

99th Day: A Warning About Technology

"Desktop PCs roasting on an open fire..."
Sony CDMavica MVC-CD500 (5MP CCD, 2003)
#ShittyCameraChallenge

Häufigkeitsverteilung der Bruttoleistung von #Balkonkraftwerk​en in #Deutschland mit Stand vom 22.12.2025.
Unplausible Inbetriebnahmedaten ersetzt, Sonderregelungen nach #Solarpaket…

And Congress & the Courts have done NOTHING.
This is why I have so little faith in American democracy. The entire system of check & balances have failed.
The only solution is to remove ambivalent politicians from Congress during midterms this year.
Original post: https://bs…