4 ms·Alpaca-LoRA (LLaMA instruct fine-tune) – open-source training code and weights6 points by Tiberium 4y ago