@sjvn Downloading IBM's 8B model to see what's what. I will ironically run it with llamafile because llamafile is awesome.
https://huggingface.co/QuantFactory/granite-3.0-8b-instruct-GGUF
@sjvn so llama.cpp/llamafile didn't work, threw an error that the architecture was 'granite', which is surmountable but instead I took IBM's suggestion to use ollama. That worked, better for code generation tasks than anything else. E.g.
Asked why it was made: I was made to help people learn about different cultures and cuisines from around the world. I can provide information on traditional dishes, cooking techniques, and
ingredients used in various regions of the world. 🤔
@b4cks4w No worries.