Simon-max2020
Follow
Stars
Reverse engineer anything with agents, from app behavior down to native binaries.
Download videos from 90+ Chinese educational platforms. Single binary, yt-dlp style.
JetBrains Mono NerdFont + LXGW WenKai = Perfect 2:1 CJK monospace font with NerdFont icons
一个按路线图逐步实现的 RAG 学习项目：从 Naive RAG → Advanced RAG → Agentic RAG → GraphRAG & Fine-tuning → RAGOps，在本地用 LangChain + 向量库把“文档问答”跑通，并逐步加入语义分块、混合检索、重排序、路由、自反思与工具调用等能力。 从最简单的 Demo 进化到企业级、甚至科研级的 RAG 系统。
Resume Sender is a lightweight, self-hosted bulk email engine built on PHP and MySQL. From job applications to marketing campaigns, newsletters, and cold outreach, it handles any one-to-many email …
MoonTVPlus 是基于 MoonTV v100 二次开发的增强版影视聚合播放器。它在原版基础上新增了外部播放器支持、视频超分、弹幕系统、评论抓取等实用功能，提供更强大的观影体验。
Data driven agentic landscapes and insights. Produced by Ant Open Source and inclusionAI.
go-doudou（doudou pronounce /dəudəu/）is OpenAPI 3.0 (for REST) spec and Protobuf v3 (for grpc) based lightweight microservice framework. It supports monolith service application as well.
A modern GUI client based on Tauri, designed to run in Windows, macOS and Linux for tailored proxy experience
MedicalGPT: Training Your Own Medical GPT Model with ChatGPT Training Pipeline. 训练医疗大模型，实现了包括增量预训练(PT)、有监督微调(SFT)、RLHF、DPO、ORPO、GRPO。
llm theoretical performance analysis tools and support params, flops, memory and latency analysis.
[ACL 2025 Findings] The official code for "Beyond Surface-Level Patterns: An Essence-Driven Defense Framework Against Jailbreak Attacks in LLMs".
Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.
本项目旨在分享大模型相关技术原理以及实战经验（大模型工程化、大模型应用落地）
RapidLLM is an LLM inference framework with continuous batching, tensor and data parallelism, quantized weights, and with pluggable Triton/CUDA kernels.
llm deploy project based mnn. This project has merged into MNN.
A C++ header-only HTTP/HTTPS server and client library
高性能计算相关知识学习笔记，包含学习笔记和相关知识的代码demo，在持续完善中。 如果有帮助的话请Star一下，对作者帮助很大，谢谢！
CUDA C 编程权威指南代码实现 包含了书上第二章到第八章的大部分代码实现和作者笔记，全由作者本人手动实现，难免有错误的地方，请大家谨慎参考，非常欢迎对错误的指正。 如果有帮助的话请Star一下，对作者帮助很大，谢谢！
Towards Safe LLM with our simple-yet-highly-effective Intention Analysis Prompting
Official repo for GPTFUZZER : Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts
FlagGems is an operator library for large language models implemented in the Triton Language.
校招、秋招、春招、实习好项目，带你从零动手实现支持LLama2/3和Qwen2.5的大模型推理框架。
This method achieves SOTA single model accuracy of 73.70 % on FER2013 without using extra training data.