微舆：人人可用的多Agent舆情分析助手，打破信息茧房，还原舆情原貌，预测未来走向，辅助决策！从0实现，不依赖任何框架。

[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo

A unified inference and post-training framework for accelerated video generation.

Codebase of GRPO: Implementations and Resources of GRPO and Its Variants

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer

Pixel-Level Reasoning Model trained with RL [NeuIPS25]

🚀 Efficient implementations of state-of-the-art linear attention models

📚LeetCUDA: Modern CUDA Learn Notes with PyTorch for Beginners🐑, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.🎉

Dexbotic: Open-Source Vision-Language-Action Toolbox

dLLM: Simple Diffusion Language Modeling

The best ChatGPT that $100 can buy.

[arXiv 2025] Generative View Stitching

Native Multimodal Models are World Learners

A Curated List of Awesome Works in World Modeling, Aiming to Serve as a One-stop Resource for Researchers, Practitioners, and Enthusiasts Interested in World Modeling.

rCM: SOTA Diffusion Distillation & Few-Step Video Generation

Nano vLLM

LongLive: Real-time Interactive Long Video Generation

Lumina-Image 2.0: A Unified and Efficient Image Generative Framework

ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

xDiT: A Scalable Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism

[SIGGRAPH 2025] Official code of the paper "FlexiAct: Towards Flexible Action Control in Heterogeneous Scenarios"

BranchGRPO: Stable and Efficient GRPO with Structured Branching in Diffusion Models

MOSAIC: Multi-Subject Personalized Generation via Correspondence-Aware Alignment and Disentanglement

codes for R-Zero: Self-Evolving Reasoning LLM from Zero Data (https://www.arxiv.org/pdf/2508.05004)

[NeurIPS 2025] Image editing is worth a single LoRA! 0.1% training data for fantastic image editing! Surpasses GPT-4o in ID persistence~ MoE ckpt released! Only 4GB VRAM is enough to run!

The simplest, fastest repository for training/finetuning small-sized VLMs.

Open-source unified multimodal model