Open-source Distilled Stable Diffusion 100% speedup

Fast Segmind-Vega with 2-8 inference steps.

Segmind-Vega Model is a distilled version of SDXL, offering a 70% reduction in size and an 100% speedup

Stable Diffusion XL - Refiner

ToolForcer: Predicting the sequence of tools so solve a query

Source: v1olet/v1olet_marcoroni-go-bruins-merge-7B ✦ Quant: TheBloke/v1olet_marcoroni-go-bruins-merge-7B-AWQ ✦ Merge AIDC-ai-business/Marcoroni-7B-v3 and rwitz/go-bruins-v2 using slerp merge

MotionDirector: Motion Customization of Text-to-Video Diffusion Models

Source: mistralai/Mistral-7B-Instruct-v0.2 ✦ Quant: TheBloke/Mistral-7B-Instruct-v0.2-AWQ ✦ Improved instruct fine-tuned version of Mistral-7B-Instruct-v0.1

generate product images - PixMiller.com

Stable Diffusion 2.1 - Realistic Vision 5.1 - VAE

GoogleAI: Style Aligned Image Generation via Shared Attention

SDXL Image Blending

A quantized 13B parameter language model from NexusFlow for SOTA zero-shot function calling

Realistic Vision V6.0 B1 Model (Text2Img, Img2Img and Inpainting)

Meina Mix V11 Model (Text2Img, Img2Img and Inpainting)

DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors

Generating object-level shape variations with Stable Diffusion

SDXL LCM with multi-controlnet, lora loading, img2img, inpainting

Kandinsky 3.0 Model (Text2Img and Img2Img)

A quantized 33B parameter language model from Deepseek for SOTA repository level code completion

Upscale Portrait Images with ControlNet Tile

Source: Q-bert/MetaMath-Cybertron-Starling ✦ Quant: TheBloke/MetaMath-Cybertron-Starling-AWQ ✦ Merge Q-bert/MetaMath-Cybertron and berkeley-nest/Starling-LM-7B-alpha using slerp merge

Source: umd-zhou-lab/claude2-alpaca-13B ✦ Quant: TheBloke/claude2-alpaca-13B-AWQ ✦ This model is trained by fine-tuning llama-2 with claude2 alpaca data

Source: rwitz/go-bruins-v2 ✦ Quant: TheBloke/go-bruins-v2-AWQ ✦ Designed to push the boundaries of NLP applications, offering unparalleled performance in generating human-like text

The Mixtral-8x7B Large Language Model (LLM) is a pretrained generative Sparse Mixture of Experts.

Depth Controlnet, but fuzzier

Image to Image enhancer using DemoFusion

Video to OpenPose

MagicAnimate using an OpenPose input video

🦙 LaMa Image Inpainting, Resolution-robust Large Mask Inpainting with Fourier Convolutions, WACV 2022

Stable Diffusion 2.1 - NSFW - Supabase

Convert scanned or electronic documents to markdown, very very very fast

Create driving poses for magic-animate

Editable image generation with MasaCtrl-SDXL

BSHM通用抠图

BSHM 人像抠图

A fine-tuned SDXL based on GTA V art

Kosmos-G: Generating Images in Context with Multimodal Large Language Models

SVD + in-painting

A ~7B parameter language model from Deepseek for SOTA repository level code completion

Source: Pclanglais/MonadGPT ✦ Quant: TheBloke/MonadGPT-AWQ ✦ What would have happened if ChatGPT was invented in the 17th century?

Playground v2 is a diffusion-based text-to-image generative model trained from scratch by the research team at Playground

Source: fblgit/una-cybertron-7b-v2-bf16 ✦ Quant: TheBloke/una-cybertron-7B-v2-AWQ ✦ A 7B MistralAI based model, best on it's series. Trained on SFT, DPO and UNA (Unified Neural Alignment) on multiple datasets

Generate plasma shader equations

Translate audio while keeping the original style, pronunciation and tone of your original audio.

Convert your videos to DensePose and use it with MagicAnimate