It feels like several things are happening which will support a resurgence of interest in LLM fine tuning for local model use.