yesterday:
- three new "greenland°hard word" benchmarks°hard word: letter count, unit conversion°hard word, part-of-speech°hard word detection°hard word. 💡 it should not°hard word be a surprise to the contemporary°hard word reader that the models struggle°hard word the most with "letter count" - how many "r"s are in strawberry.
still to do:
- code°hard word cleanup°hard word 💡 the "run" method°hard word is written°hard word in slightly different form seven times
- more benchmarks°hard word
- dashboard°hard word UI°hard word improvements°hard word