saiga_mistral_7b_gguf

Llama.cpp compatible versions of an original 7B model.

Download one of the versions, for example model-q4_K.gguf.

wget https://huggingface.co/IlyaGusev/saiga_mistral_7b_gguf/resolve/main/model-q4_K.gguf

Download interact_mistral_llamacpp.py

wget https://raw.githubusercontent.com/IlyaGusev/rulm/master/self_instruct/src/interact_mistral_llamacpp.py

How to run:

pip install llama-cpp-python fire

python3 interact_mistral_llamacpp.py model-q4_K.gguf

System requirements:

• 10GB RAM for q8_0 and less for smaller quantizations