LLaMA

github.com ↗

- LLaMA is intended as a minimal, hackable and readable example to load LLaMA (arXiv) models and run inference.