restart/ crash/ tracker/ loading with minor fixes..

data, fraction removed..

slightly better optimization with data, fraction & batch parameters

ints8_quantization, as executorch is failing

ExecuTorch optimization, UI improvements..

path corrected in model.py

int8=True for speeding up inference

original sources url added

openvino adapted and adjustments

onnx --> openvino, intel hardware

png xet enabled