Guys i need help i'm working for a company who sell products and we need a chatbot to answer user's questions is it good to use llama2 as LLM and llamaindex with gptindex as embedding model ?
If you're making an outwards facing chatbot I assume data security isn't your main concern. In that case I'd probably use the Llamaindex OpenAI API abstractions for LLM + embeddings because it's a bit easier and performs better. But if you want, you can definitely use llama2 and one of the open source embedding models (like from Huggingface) with Llamaindex https://gpt-index.readthedocs.io/en/stable/examples/embeddings/huggingface.html
Oh thanks for your response i'll check the price for both LLM before choosing but the main problem i have actually with all the LLM is the scope of the answer because i want the chatbot to only answer the questions that concerns the company I work for, but the chatbot also answer questions like "what is the capital of England" and i'm reading the docs but i dont know how to fix that
