WKQ

Mini-LLM. This project aims to replicate mainstream open-source model architectures with limited computational resources, implementing mini models with 100-200M parameters.

transformer-from-scratch. 手撕transformer并完成一个简单的机器翻译。

mini-llama3-from-scratch. 从零实现一个迷你版llama3，并进行预训练和微调

wrap-openai. Wrap custom generate function as an OpenAI SDK compatible API service.

WKQ9411.github.io. My personal blog.