3 ms·Using Small Models to Compare Language Learning and Tokenizer Performance1 points by PaulHoule 2y agodeleted 2y ago[deleted]