Stars
[ICML 2026] Coloring the Noise: Adversarial Sobolev Alignment for Faithful Image Super Resolution (Official Implementation)
All my self trained & released AI upscaling models. After gathering and applying over 600 different upscaling models, I learned how to train my own models, and these are the results.
A tool to snap pixels to a perfect grid. Designed to fix messy and inconsistent pixel art generated by AI.
Official repo for paper "Sparse Representation and Construction for High-Resolution 3D Shapes Modeling".
From Images to High-Fidelity 3D Assets with Production-Ready PBR Material
[NeurIPS 2025] PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers
[ICCV 2025] Official code of DeepMesh: Auto-Regressive Artist-mesh Creation with Reinforcement Learning
[SIGGRAPH Asia 2025] DreamO: A Unified Framework for Image Customization
High-Resolution 3D Assets Generation with Large Scale Hunyuan3D Diffusion Models.
Official repo for paper "Structured 3D Latents for Scalable and Versatile 3D Generation" (CVPR'25 Spotlight).
SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformer
A high-throughput and memory-efficient inference and serving engine for LLMs
Fully fine-tune large models like Mistral, Llama-2-13B, or Qwen-14B completely for free
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
[CVPR 2024 Highlight] Putting the Object Back Into Video Object Segmentation
[ECCV 2022] XMem: Long-Term Video Object Segmentation with an Atkinson-Shiffrin Memory Model
An open-source project dedicated to tracking and segmenting any objects in videos, either automatically or interactively. The primary algorithms utilized include the Segment Anything Model (SAM) fo…
Track-Anything is a flexible and interactive tool for video object tracking and segmentation, based on Segment Anything, XMem, and E2FGVI.
The TinyLlama project is an open endeavor to pretrain a 1.1B Llama model on 3 trillion tokens.
[MLSys 2024 Best Paper Award] AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration
QLoRA: Efficient Finetuning of Quantized LLMs
Official implementation of Würstchen: Efficient Pretraining of Text-to-Image Models
OpenLLaMA, a permissively licensed open source reproduction of Meta AI’s LLaMA 7B trained on the RedPajama dataset
🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support
Large-scale, Informative, and Diverse Multi-round Chat Data (and Models)
We unified the interfaces of instruction-tuning data (e.g., CoT data), multiple LLMs and parameter-efficient methods (e.g., lora, p-tuning) together for easy use. We welcome open-source enthusiasts…
Modern spell checking library - accurate, fast, multi-language
Finetune ModelScope's Text To Video model using Diffusers 🧨