IlyaGusev/saiga_nemo_12b_gguf

• IlyaGusev/saigascored
• IlyaGusev/saigapreferences

• ru

Llama.cpp compatible versions of an original 12B model.

Download one of the versions, for example saiganemo12b.Q4KM.gguf.

wget https://huggingface.co/IlyaGusev/saiganemo12bgguf/resolve/main/saiganemo12b.Q4K_M.gguf

Download interactllama3llamacpp.py

wget https://raw.githubusercontent.com/IlyaGusev/rulm/master/selfinstruct/src/interactllama3_llamacpp.py

How to run:

pip install llama-cpp-python fire


python3 interactllama3llamacpp.py saiganemo12b.Q4KM.gguf

System requirements:

• 15GB RAM for q8_0 and less for smaller quantizations