A fast AI Video Generator for the GPU Poor. Supports Wan 2.1/2.2, Qwen Image, Hunyuan Video, LTX Video and Flux.

[ICCV 2025 Highlights] Large-scale photo-realistic virtual worlds for embodied AI

A pipeline parallel training script for diffusion models.

Diffusion attentive attribution maps for interpreting Stable Diffusion for image-to-image attention.

Wan2.2-Lightning: Speed up wan2.2 model with distillation

Make self forcing endless. Add cache purging. Add prompt controllability.

Awesome curated collection of images and prompts generated by gemini-2.5-flash-image (aka Nano Banana) state-of-the-art image generation and editing model. Explore AI generated visuals created with…

🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

Inference Llama 2 in one file of pure C

Fork of Implementation of 💍 Ring Attention, from Liu et al. at Berkeley AI, in Pytorch. Modifed to support tree attention for query length > 1

Official Pytorch implementation of "Vision Transformers Don't Need Trained Registers"

Embodied Chain of Thought: A robotic policy that reason to solve the task.

Implementing FlowEdit, maybe other inversion techniques for the Wan video generation model

Modern Shadowsocks in Go

IRS: A Large Synthetic Indoor Robotics Stereo Dataset for Disparity and Surface Normal Estimation

Large Gaussian Reconstruction Model for Efficient 3D Reconstruction and Generation

Marigold adapted for video estimation

FlashFFTConv: Efficient Convolutions for Long Sequences with Tensor Cores

Lifting ControlNet for Generalized Depth Conditioning

Stable diffusion for inpainting

Support BlenderProc2 with multi-GPU batch rendering and 3D visualization for 3D-Front

A curated list of resources dedicated to text summarization

Official PyTorch implementation of "Extract Free Dense Labels from CLIP" (ECCV 22 Oral)

assistant tools for attention visualization in deep learning

Sparse Single Sweep LiDAR Point Cloud Segmentation via Learning Contextual Shape Priors from Scene Completion (AAAI 2021)

A Chinese Translation of Stanford CS229 notes 斯坦福机器学习CS229课程讲义的中文翻译

openpilot is an open source driver assistance system. openpilot performs the functions of Automated Lane Centering and Adaptive Cruise Control for over 100 supported car makes and models.

Tooling for professional robotic development in C++ and Python with a touch of ROS, autonomous driving and aerospace: https://freerobotics.tools/