Edge Inference

Executing model forward passes on local devices near data sources rather than cloud servers, reducing latency and enhancing priv