Stars
A project structure aware autonomous software engineer aiming for autonomous program improvement. Resolved 37.3% tasks (pass@1) in SWE-bench lite and 46.2% tasks (pass@1) in SWE-bench verified with…
[NeurIPS 2025] A Graph-based LLM Framework for Real-world SE Tasks
This is a cleanroom deobfuscation of the official Claude Code npm package.
Being-VL-0.5: Unified Multimodal Understanding via Byte-Pair Visual Encoding
Neural Code Intelligence Survey 2024-25; Reading lists and resources
verl: Volcano Engine Reinforcement Learning for LLMs
[ACL 2024] Long-Context Language Modeling with Parallel Encodings
VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo
Various extensions for the Eino framework: https://github.com/cloudwego/eino
本仓库包含对 Claude Code v1.0.33 进行逆向工程的完整研究和分析资料。包括对混淆源代码的深度技术分析、系统架构文档，以及重构 Claude Code agent 系统的实现蓝图。主要发现包括实时 Steering 机制、多 Agent 架构、智能上下文管理和工具执行管道。该项目为理解现代 AI agent 系统设计和实现提供技术参考。
gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI
DeepSeek Native Sparse Attention pytorch implementation
Our paper about robust LLM fingerprints.
Official code implementation of General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model
用于从头预训练+SFT一个小参数量的中文LLaMa2的仓库；24G单卡即可运行得到一个具备简单中文问答能力的chat-llama2.
整理开源的中文大语言模型，以规模较小、可私有化部署、训练成本较低的模型为主，包括底座模型，垂直领域微调及应用，数据集与教程等。
Doing simple retrieval from LLM models at various context lengths to measure accuracy
The official implementation of LIFT: Improving Long Context Understanding of Large Language Models through Long Input Fine-Tuning
人工精调的中文对话数据集和一段chatglm的微调代码
A quick guide (especially) for trending instruction finetuning datasets
This is a continuously updated handbook for readers to easily track the latest Text-to-SQL techniques in the literature and provide practical guidance for researchers and practitioners.
LangGPT: Empowering everyone to become a prompt expert! 🚀 📌 结构化提示词（Structured Prompt）提出者 📌 元提示词（Meta-Prompt）发起者 📌 最流行的提示词落地范式 | Language of GPT The pioneering framework for structured & meta-prompt…
Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
The human face subset of LAION-400M for large-scale face pretraining.