GPU instances for inference
We run small ML models next to our API. A few GPU instance types (even older cards) in one region would be enough to start.
We run small ML models next to our API. A few GPU instance types (even older cards) in one region would be enough to start.