Need advice for building a conversational chatbot by using LoRa on open source LLM

[sonic sail]

After watching the anime Steins;Gate, I'm inspired to replicate the AI character Amadeus by using a LoRa on an existing opensource LLM. I've already gathered a dataset of all of the conversations in the entire series. I just need some advice regarding which LLM to chose and what format must be used.

I'm still a beginner in training 🙂

[distant halo]

Instead of finetuning / using LoRa, how about giving a shot to existing models out there right now. You can give character description, dialogue example, personality traits and it copies it upto some extent.

[nocturne trench]

Ayo? @distant halo level 5 !!!

[distant halo]

Not sure about others but with RAG, you can probably throw in a knowledge document and it can find information about the character and respond to questions related to it.