1
A 125M model beat a 14B LLM at de-identifying medical text 40 faster, on CPU
小模型逆袭！125M参数在医疗文本去识别上碾压14B大模型，CPU跑还快40倍
Your data never leaves the machine - and you can check my math Building localscrub, a local-first PHI de-identification cascade, and benchmarking it h…