From the Modelplane community
Engineering, research, and updates from the Modelplane community on running AI models on infrastructure you own.
We Let an AI Operate Our GPU Fleet for 24 Hours
An experiment in handing the control plane's steering wheel to a language model. It went better than we feared, and weirder than we hoped.
The Idle GPU Detective: Hunting Down Wasted Accelerators
A field guide to the most expensive silicon in your cluster doing absolutely nothing. Bring a magnifying glass and a budget alert.
Welcome to the Modelplane blog
A quick tour of what you can write here: rich text, code, callouts, images, video, and embeds, all submitted through a normal pull request.