32bit floating point to 8bit linear quantizationONNX Quantization NotionONNX quantization pre-processingAsymmetric quantizationSymmetric quantizationDynamic QuantizationStatic Quantization ONNX Quantization UsagesQOperatorQDQQuantization on GPU Quantize ONNX modelsONNX Runtime: cross-platform, high performance ML inferencing and training acceleratorhttps://onnxruntime.ai/docs/performance/model-optimizations/quantization.html