Thanks for replying!
I am on v5.11.0 of PyMC, so it could either be a bug or the Python gods screwing with me again 
I have ruled out (at least on my machine that it isn’t the duplicate sampling of the same variable doing it since the following code still throws the error):
Running the default step method or the same one you mentioned leads to no issues, which thus leads me to believe then for whatever reason having multiple steps in the array like stepmethod=[step1,step2] is what leads to the issue but then if it isn’t causing issues on v5.10.0 then perhaps it is a bug of sorts, since I can get the trace_plot out fine but the arviz diagnostic plots don’t like it, since the plot_pair function function throws its own error out:
The conda environment I’m working on should be clean so I will try rolling back to v5.10.0 and that should illustrate if it is a v5.11.0 bug or a bug with my own conda environment.