Comment by madisonmay

This is an excellent use case for LLM fine-tuning, purely because of the ease of generating a massive dataset of input / output pairs from public C code