wait i thought llama 2 was not open source. How do you use it??
Discussion
Compile the latest llama.cpp then download the latest LLaMA 2 uncensored 7B weights here (search TheBloke's profile if you need larger weights):
https://huggingface.co/TheBloke/llama2_7b_chat_uncensored-GGML
How small is this? I know Llama is good I just didn't expect a 7B model to run on a phone... wow!

