Fine-tuning T5 & LLAMA and using RAG

Job ID: 38704929

Budget: $30 – $250 USD

Hello!

I want to build any recommendation model for example medical one, so when I enter a prompt I got a response. For that I want to use two different ways. One is finetuning a large language model like LLAMA 3 or FLAN T5 with the dataset. The second is using RAG and FAISS function to retrieve the answer from the database with the use of victor database for bothe the prompt and the response. Finally, I will compare the responses that comes from a RAG code and the finetuning code.


I found online some codes like this one in this tuturial (https://www.youtube.com/watch?v=AjQPRomyd-k ), but it didn’t work when I try it with other dataset.

It’s a project for a class, and I want to be done with it as soon as possible because I already wasted almost a month trying but unfortunately I couldn’t do it.