JAWWAD Edge ML · Quantization · Systems Optimization
> init edge_runtime > load yolov8n.onnx [ONNX Runtime] > quantize fp32 -> int8 > deploy: no cloud, no excuses $ status: READY QUANTIZATION SCOPEFP32
INT8
~90% smaller · <40ms latency
UNIDA Gontor, Indonesia Erasmus+ KA171 · Univ. of Seville


