I’d be curious to use them on larger data sets. Log files, for instance. High volume, might be low value per line but not much cost per token if you already hav

I’d be curious to use them on larger data sets. Log files, for instance. High volume, might be low value per line but not much cost per token if you already have a gpu to interpret them using a LLM.