May I use Python Node in a DAQmx Style? If I keep Python Node Ref alive does that means the Python Node will keep python code's state at background?

[jiangliang]

If I use call python node to load model every time for each prediction, it would be awefully slow, load the model itself takes about 10 senonds, but prediction takes about 1 second. It's doesn't seem right to call the whole excution all at once.

But the thing is, if I keep the reference alive, does that means I can call Python Node later with inner state value keepped in the Phython ENV?

Thanks!