This project aims to provide beginners with a more accessible implementation of the LLAMA (Train/Inference/SFT/RLHF) through concise code and detailed comments. 从0到1实现LLAMA的训练、微调、部署。
------------------2023/08/04--------- create this repo,If you find it interesting, please give it a star. 点点关注 点点关注
------------------2023/08/19--------- uploaded tokenize.py json_process.py 上传了训练tokenzier的代码
------------------2023/08/27--------- uploaded llama model code 上传了llama model的代码.其中generate.py是官方代码的简化版, model.py是在官方代码上做的修改以适配训练和推理。