Customers want to use Elastic Inference (https://aws.amazon.com/machine-learning/elastic-inference/) to save cost while running inference. An example showcasing this in a notebook would be great.
Also, would be good to highlight some limitations customers may run into while using Elastic Inference
This is kind of hack and I am still waiting for more native kubernetes integrations. We can design a device plugin for this. For short term, we can consider to add an example for this in serving section. I suggest to check with internal team first.
Customers want to use Elastic Inference (https://aws.amazon.com/machine-learning/elastic-inference/) to save cost while running inference. An example showcasing this in a notebook would be great. Also, would be good to highlight some limitations customers may run into while using Elastic Inference