Youtu-Embedding is an industry-leading, general-purpose text representation model developed by Tencent Youtu Lab.

A lightweight LMM-based Document Parsing Model

[Pytorch] The repo contains the code for "FORGE: Forming Semantic Identifiers for Generative Retrieval in Industrial Datasets"

Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels

Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 500+ LLMs (Qwen3, Qwen3-MoE, Llama4, GLM4.5, InternLM3, DeepSeek-R1, ...) and 200+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, Llava, GLM4v, Ph…

Benchmarking Recommendation Abilities for Large Language Models

Tongyi Deep Research, the Leading Open-source Deep Research Agent

Open Agent Coding CLI, Koding with GLM, Qwen, Kimi, DeepSeek etc.（welcome to use Kode to summit PR)

RecBase: Generative Foundation Model Pretraining for Zero-Shot Recommendation

Toolkit for linearizing PDFs for LLM datasets/training

Transforms complex documents like PDFs into LLM-ready markdown/JSON for your Agentic workflows.

[KDD 2025] Quadratic Neural Networks for Click-through Rate Prediction

The official repo for “Dolphin: Document Image Parsing via Heterogeneous Anchor Prompting”, ACL, 2025.

中文对话0.2B小模型（ChatLM-Chinese-0.2B），开源所有数据集来源、数据清洗、tokenizer训练、模型预训练、SFT指令微调、RLHF优化等流程的全部代码。支持下游任务sft微调，给出三元组信息抽取微调示例。

使用open-webui中的pipelines技术在open-webui中调用ragflow的agent实现基于知识库的智能对话，并拥有美观的界面。

Pipelines: Versatile, UI-Agnostic OpenAI-Compatible Plugin Framework

RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs

Production-ready platform for agentic workflow development.

🚀🚀 「大模型」2小时完全从0训练26M的小参数GPT！🌏 Train a 26M-parameter GPT from scratch in just 2h!

“AI-Compass”将为社区指引在 AI 技术海洋中航行的方向，无论你是初学者还是进阶开发者，都能在这里找到通往 AI 各大方向的路径。旨在帮助开发者系统性地了解 AI 的核心概念、主流技术、前沿趋势，并通过实践掌握从理论到落地的全过程。

Source code for Twitter's Recommendation Algorithm

Examples for Recommenders - easy to train and deploy on accelerated infrastructure.

A privacy-first, open-source platform for knowledge management and collaboration. Download link: http://github.com/logseq/logseq/releases. roadmap: http://trello.com/b/8txSM12G/roadmap

verl: Volcano Engine Reinforcement Learning for LLMs

A decoder-only llm-based generative recommendation framework that integrates endogenous and exogenous behavioral and semantic information in a non-intrusive manner

This is the repository for ”ROMA: Recommendation-Oriented Language Model Adaptation Using Multi-Modal Multi-Domain Item Sequences“ in KDD 2025

📰 Must-read papers and blogs on LLM based Long Context Modeling 🔥