【[9星]mini_llm：GPT-2 和 Llama 的最小 PyTorch 实现，旨在简化代码以便更容易理解和使用，并且能够在短时间内训练出性能良好的自然语言生成系统】'Minimal PyTorch implementation of GPT2 and Llama. This repo simplifies the GPT-2 and LLaMA code for easier understanding and usage. This repo trains a natural language generation system with 1 million characters and converges in less than 5 minutes.' GitHub: github.com/qiuqiangkong/mini_llm
