I wish I could fine tune a model to my coding style, effectively creating a coding agent clone of myself.
can't you use all of the code you've written as the training data somehow? (am tarded)
pretty much using your github repo as a RAG
pretty much using your github repo as a RAG