nextrelease benchmark

[kawa-kokosowa]

Run various benchmarks which can lead to quantifiable data established every release, in this case, establish and record various benchmarks for the v0.14.0 "jack-o'-lantern" release (https://github.com/lily-seabreeze/sappho/milestone/1). The desired effect is having a base to compare optimizations against, or any refactoring at all.

Open to discussion; hoping to hear approaches. A few I can think of:

• Use something like pycallgraph for a single frame, or many frames averaged out (I think it may do this automagically)
• Benchmarks through demo.py
• Use tests to determine benchmarks

[kawa-kokosowa]

@txanatan is this something that'd interest you?