Bug report: Inference endpoints with CPU does not wake up when they scale to zero

Site Feedback
1

As the title said, if you have a model running on CPU with AWS then it won’t wake up if it is scaled to zero. No such issue with using a GPU instance.

Might be the new instances?