Figure out a rule of thumb for tensor arena size

[eloquentarduino]

[maarten-pennings]

Thanks for the lib. Works great.

I was also puzzled when figuring out the arena size. I wrote my process down for my project. See the section starting with One of the configuration parameters for the interpreter is the TENSOR_ARENA_SIZE. in my project https://github.com/maarten-pennings/MachineLearning/blob/main/rock-paper-scissors/rock-paper-scissors.ipynb

Maybe this helps.

[eloquentarduino]

I will take time to review your notebook, thank you for the hint.