0

WebInf: Accelerating WebGPU-based In-browser DNN Inference via Adaptive Model Partitioning