What the others said about dataset size, but there's no reason to fully "reinvent the wheel".
Train a Lora on your posts and bolt it on to your favourite LLM. It won't quite be you, but you'll be horrified and entertained by how close it can get :D
I have no idea how to do this
Please Login to reply.
Its non-trivial, but:
https://zohaib.me/a-beginners-guide-to-fine-tuning-llm-using-lora/
