Find answers from the community

Updated last year

LLMs-at-DoD/tutorials/Chatting with your...

Hey Everyone! I work for department of defense, and I'm happy to announce we were able to publish a tutorial using llama-index on the official DoD Github πŸ™‚ https://github.com/deptofdefense/LLMs-at-DoD/blob/main/tutorials/Chatting%20with%20your%20Docs.ipynb open source ftw!
L
G
j
3 comments
Very cool! Thanks for sharing!

btw, I noticed that you are using mistral-openorca, but you are using the llama2 messages_to_prompt functions -- this might be decreasing your accuracy

The mistral-openorca model seems to be fine-tuned with a different scheme, so you'll need a custom messages_to_prompt function πŸ‘€

The format is something like
Plain Text
<|im_start|>system
You are MistralOrca, a large language model trained by Alignment Lab AI. Write out your reasoning step-by-step to be sure you get the right answers!
<|im_end|>
<|im_start|>user
How are you?<|im_end|>
<|im_start|>assistant
I am doing well!<|im_end|>
<|im_start|>user
Please tell me about how mistral winds have attracted super-orcas.<|im_end|>
<|im_start|>assistant

https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca
ah great catch! thank you for point this out @Logan M
this is awesome!
Add a reply
Sign up and join the conversation on Discord