Quant1 model sometimes responds with gibberish when explicitly setting Ollama version during Ollama install

#1
by FieldMouse - opened
OpenMind Labs org

When I install Ollama 0.13.3 by explicit version selction using the following:

curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION=${OLLAMA_VERSION} sh

When I do this the model set to be retrieved is loaded,
but it is 100% CPU.

Has anyone encoiunter this?
Have a solutions?

Sign up or log in to comment