Source models: (.pt) https://huggingface.co/karpathy/tinyllamas

Converted to GGML latest model file format gguf.

https://github.com/ggerganov/llama.cpp