Mozilla-Ocho / llamafile

Distribute and run LLMs with a single file.

Home Page:https://llamafile.ai

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Update llama 3 prompt template

tudoanh opened this issue · comments

Hi, I'm trying to run Llama 3 8B Q4 model, but seems like the prompt template has been changed.

Then I saw this new release from llama.cpp: https://github.com/ggerganov/llama.cpp/releases/tag/b2707

llama : add llama-3 chat template (#6751)

* Added llama-3 chat template

* Update llama.cpp

Can you update so we can run Llama 3 normally? Thanks a lot!

@tudoanh how are you running Llama 3? I'd love to know steps to run for when this issue is addressed.

I'm just using the Llamafile latest ver 0.7.3 with command like this

sh -c './llamafile-0.7.3 -m Meta-Llama-3-8B-Instruct.Q4_0.gguf -ngl 15 --port 8080 --host 0.0.0.0'

You can find the gguf files on huggingface. Example: https://huggingface.co/bartowski/Meta-Llama-3-8B-Instruct-GGUF