Stars
An AI-powered custom node for ComfyUI designed to enhance workflow automation and provide intelligent assistance
Lightweight PyAV-powered ComfyUI nodes to load/save multimedia
Make images smaller using best-in-class codecs, right in the browser.
收集全国各高校招生时不会写明，却会实实在在影响大学生活质量的要求与细节
[ICCV 2025] 🔥🔥 UNO: A Universal Customization Method for Both Single and Multi-Subject Conditioning
A fast image processing library with low memory needs.
[CVPR 2024] Real-Time Open-Vocabulary Object Detection
Faster generation with text-to-image diffusion models.
T-GATE: Temporally Gating Attention to Accelerate Diffusion Model for Free!
Tensors and Dynamic neural networks in Python with strong GPU acceleration
An extremely fast Python package and project manager, written in Rust.
https://wavespeed.ai/ Context parallel attention that accelerates DiT model inference with dynamic caching
https://wavespeed.ai/ [WIP] The all in one inference optimization solution for ComfyUI, universal, flexible, and fast.
Scalene: a high-performance, high-precision CPU, GPU, and memory profiler for Python with AI-powered optimization proposals
A debugging and profiling tool that can trace and visualize python code execution
Quantized Attention achieves speedup of 2-5x and 3-11x compared to FlashAttention and xformers, without lossing end-to-end metrics across language, image, and video models.
Terminal image viewer with native support for iTerm and Kitty
Official implementation of "A Noise is Worth Diffusion Guidance", code and weights will be available soon.
[ICCV2025] The code of our work "Golden Noise for Diffusion Models: A Learning Framework".
This repo contains the code for 1D tokenizer and generator
Support for miscellaneous image models. Currently supports: DiT, PixArt, HunYuanDiT, MiaoBi, and a few VAEs.