EpochLex entry

Inference Backend

IN-fer-ens BACK-end

The software implementation within an inference-serving system that actually executes a model, often by wrapping a framework or runtime such as PyTorch, TensorRT, ONNX Runtime, or a custom execution layer.