s1: Simple test-time scaling

A streamlined and customizable framework for efficient large model evaluation and performance benchmarking

TransMLA: Multi-Head Latent Attention Is All You Need

Official PyTorch implementation for "Large Language Diffusion Models"

Official repository for the paper "LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code"

Qwen GRPO Graph Extraction RL Finetune

Easy-to-use LLM fine-tuning framework (LLaMA, BLOOM, Mistral, Baichuan, Qwen, ChatGLM)

Visual Studio Code

Argilla is a collaboration tool for AI engineers and domain experts to build high-quality datasets

The easiest tool for fine-tuning LLM models, synthetic data generation, and collaborating on datasets.

ClearML - Auto-Magical CI/CD to streamline your AI workload. Experiment Management, Data Management, Pipeline, Orchestration, Scheduling & Serving in one MLOps/LLMOps solution

Democratizing Reinforcement Learning for LLMs

Finetune Llama 3.3, DeepSeek-R1, Reasoning, Phi-4 & Gemma 2 LLMs 2x faster with 70% less memory

Utils for Unsloth

A repository aimed at pruning DeepSeek V3, R1 and R1-zero to a usable size

User-friendly AI Interface (Supports Ollama, OpenAI API, ...)

Open source codebase powering the HuggingChat app

🚀 Next Generation AI One-Stop Internationalization Solution. 🚀 下一代 AI 一站式 B/C 端解决方案，支持 OpenAI，Midjourney，Claude，讯飞星火，Stable Diffusion，DALL·E，ChatGLM，通义千问，腾讯混元，360 智脑，百川 AI，火山方舟，新必应，Gemini，Moonshot …

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

Collection of leaked system prompts

LIMO: Less is More for Reasoning

This is a replicate of DeepSeek-R1-Zero and DeepSeek-R1 training on small models with limited data

Open Thoughts: Fully Open Data Curation for Thinking Models

A lightweight reproduction of DeepSeek-R1-Zero with indepth analysis of self-reflection behavior.

一个LLM-Agent与PPT项目，支持基于对话式需求进行操作PPT的原生AI应用项目。模型为DeepSeek，基于LLM+VBA调用进行文档操作，轻量级ChatPPT服务。An LLM-Agent and PPT project that supports native AI application projects that operate PPT based on conversati…

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

Let your Claude able to think

Cache-Augmented Generation: A Simple, Efficient Alternative to RAG

The headless rich text editor framework for web artisans.