Hi, @zoghiboy
First, I will appreciate your effort.
And RAG and LoRA are both good methods to chat with your document.
If you don’t satisfy the result, I recommend you to use RAG.
In RAG you can use vector database. This is very effective way to chat with your document.
But you should consider the response time. Which db service you choose, the time limit is different.
And you can also use some APIs like deepseek or openai free apis.
You can also view my examples.
If you want to help from me, I will be here anytime for you.