Stars
Implementation of "EasyControl: Adding Efficient and Flexible Control for Diffusion Transformer"
ComfyUI-Manager is an extension designed to enhance the usability of ComfyUI. It offers management functions to install, remove, disable, and enable various custom nodes of ComfyUI. Furthermore, th…
LRCLIB server written in Rust with Axum and SQLite3 database
PyTorch half precision gemm lib w/ fused optional bias + optional relu/gelu
Official inference repo for FLUX.1 models
A high-throughput and memory-efficient inference and serving engine for LLMs
F491A powerful tool that translates ComfyUI workflows into executable Python code.
https://wavespeed.ai/ Best inference performance optimization framework for HuggingFace Diffusers on NVIDIA GPUs.
PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
Kandinsky x Deforum — generating short animations
Better Aligning Text-to-Image Models with Human Preference. ICCV 2023
Official repo for consistency models.
Kandinsky 2 — multilingual text2image latent diffusion model
Examples and guides for using the OpenAI API
Optimized primitives for collective multi-GPU communication
Repository for NeurIPS 2020 Spotlight "AdaBelief Optimizer: Adapting stepsizes by the belief in observed gradients"
Train transformer language models with reinforcement learning.
A new one shot face swap approach for image and video domains
[CVPR2023] The code for 《Position-guided Text Prompt for Vision-Language Pre-training》
Implementation of aspect ratio bucketing for training generative image models as described in: https://blog.novelai.net/novelai-improvements-on-stable-diffusion-e10d38db82ac
RuTransform: python framework for adversarial attacks and text data augmentation for Russian
Versatile Diffusion: Text, Images and Variations All in One Diffusion Model, arXiv 2022 / ICCV 2023
Linear Algebra Course being taught in HSE in 2021/2022 (in russian)
A tiny scalar-valued autograd engine and a neural net library on top of it with PyTorch-like API