3 repositorios
Awesome GitHub RepositoriesDiffusion Model Memory Optimizers
Frameworks that reduce the memory footprint of diffusion models for inference on consumer hardware.
Distinct from Diffusion Weight Optimizers: Focuses on inference memory optimization and precision recovery rather than training weight optimization.
Explore 3 awesome GitHub repositories matching artificial intelligence & ml · Diffusion Model Memory Optimizers. Refine with filters or upvote what's useful.
Awesome Diffusion Model Memory Optimizers GitHub Repositories
Encuentra los mejores repositorios con IA.Buscaremos los repositorios que mejor coincidan usando IA.
- picsart-ai-research/text2video-zero
Picsart-AI-Research/Text2Video-Zero
4,244Ver en GitHubText2Video-Zero es un modelo de difusión de texto a video y un framework diseñado para sintetizar secuencias de video temporalmente consistentes a partir de prompts textuales. Funciona como un generador de video zero-shot, reutilizando modelos de difusión de imágenes preentrenados para crear contenido de video sin requerir entrenamiento adicional en conjuntos de datos de video. El sistema incluye un sintetizador de video condicional que permite la generación guiada utilizando mapas de profundidad, bordes o poses para controlar el diseño estructural y el movimiento. También proporciona capacidades de edición de video basadas en texto para modificar el estilo o el contenido de clips de video existentes mediante instrucciones en lenguaje natural. Para gestionar los requisitos computacionales, el proyecto implementa inferencia optimizada para memoria de GPU. Esto se logra mediante técnicas como la fusión de tokens y la fragmentación de fotogramas para reducir el uso de VRAM durante el proceso de generación.
Optimizes GPU memory usage during video generation to enable inference on consumer-grade hardware.
Pythonvideo-editingvideo-generation - nunchaku-ai/nunchaku
nunchaku-ai/nunchaku
3,883Ver en GitHubNunchaku is a 4-bit model quantization library and diffusion model inference engine designed to run large-scale neural networks on consumer GPUs. It functions as a GPU-accelerated optimizer that reduces VRAM usage and increases inference speed through weight compression and memory management. The project utilizes low-rank weight decomposition and SVD weight quantization to compress models to four-bit precision while maintaining visual fidelity. It employs kernel-level operator fusion to minimize data movement and hardware-aware precision mapping to adjust numerical precision based on the unde
Provides a specialized inference engine for running large-scale diffusion models with reduced memory overhead on consumer GPUs.
Pythoncomfyuidiffusion-modelsflux - city96/comfyui-gguf
city96/ComfyUI-GGUF
3,291Ver en GitHubComfyUI-GGUF is a memory optimizer and model loader for ComfyUI that enables the execution of large transformer-based generative models using quantized weights. It provides a system for loading GGUF formatted weights within a node-based diffusion interface to reduce GPU memory consumption. The project includes a quantization tool for converting standard model checkpoints into compressed binary formats and a tensor fixer to restore missing keys and correct architectures in binary model files. These utilities ensure that compressed models remain functional during inference on hardware with limi
Provides a framework for running large transformer-based generative models using quantized weights.
Python
