Quantization support for ONNX using Intel Neural Compressor (formerly named Intel Low Precision Optimization Tool)