aws-samples / amazon-kendra-langchain-extensions

Samples to build Generative AI applications with LangChain and Amazon Kendra
https://aws.amazon.com/blogs/machine-learning/quickly-build-high-accuracy-generative-ai-applications-on-enterprise-data-using-amazon-kendra-langchain-and-large-language-models/
MIT No Attribution
158 stars 104 forks source link

Is the prompt format for llama-2 correct? #60

Closed utility-aagrawal closed 9 months ago

utility-aagrawal commented 9 months ago

Hi,

Thank you for sharing your work here! It's very helpful. I am working on a RAG pipeline using Kendra, langchain and llama-2. I created a Kendra index and have been using llama-2 python script for Q&A. I see the prompts in the py file but they don't seem to be following the correct llama-2 format. Is that handled somewhere else in the code. I couldn't find where it is. Can you point me to it?

As per my understanding, this is the correct format to prompt Llama-2: https://huggingface.co/blog/llama2#how-to-prompt-llama-2

Let me know in case of any questions. Thanks!

MithilShah commented 9 months ago

I havent seen a need for special prompt. Tested with the one that we have here and it seems to work well. Are you seeing any issues?

MithilShah commented 9 months ago

Fixed