# Lightweight fine - tuning

Orca 2 13B is a large language model fine - tuned by Microsoft based on LLAMA - 2. It provides quantized files in GGUF format and supports efficient inference in various hardware environments.