Introducing RobustSAM by Snap Inc. Research! 🚀 A robust version of the Segment Anything Model (SAM) with improved performance on low-quality images while maintaining zero-shot segmentation capabilities 🦾 Integration with OWLv2 allows RobustSAM to be text-promptable

🚀 Concept Sliders now support FLUX.1 models! 🧠💡 Interestingly, the same formulation that works for SDXL works for FLUX models too! Even though they work quite differently under the hood. Check out the code here: github.com/rohitgandikota… Project Page: sliders.baulab.info

🖼 Now we have Flux.1 Lora viewer and 👁‍🗨 Qwen2-VL-7B 🥳 Thanks to Qwen Team ❤ and Civitai open-source community! ❤ We can also use any small-size public #Flux1 #Lora with 🥪 🥪 Tost AI 🔥 please try it 🐣

StableNormal: Reducing Diffusion Variance for Stable and Sharp Normal 👇🧐 This enables “Stable-and-Sharp” normal estimation, which outperforms multiple baselines Install locally: 👇😋 github.com/Stable-X/Stabl…

Recently RunwayML deleted SDv1.5 from their HF /GitHUB Account.....😖 Lots of other model getting deleted.. Be prepare for future.. lots of Govt. restrictions.. Quick solution is here......😎 Clone any Hugging Face repo to your account via COLAB 😻👇 stablediffusiontutorials.com/2024/09/clone-…

Flux now in Realtime. available in Krea with hundreds of styles included. free for everyone.

Now FlUX LoRA Training KohyaSS 😍 -Support low VRAM-12/16/20 -Simple to use WebUI -Flux DEV and Schnell Available... Follow this tutorial :👇😎 stablediffusiontutorials.com/2024/08/flux-l…

Here they are...with SD3 medium fine tuning....😖👇

This one deserves 10K stars. github.com/chflame163/Com…

Tencents GameGen-O: 😵 Diffusion transformer for video game generation! -Generates high-quality characters, environments, actions -Provides interactive gameplay simulation -Two-stage training: Foundation pretraining + InstructNet tuning More details: gamegen-o.github.io

No GPU No Problem...😵😳 Train any SD models...with Dreambooth on Colab -SDv1.5 -SDXL -Flux Follow the workflow:👇🤯 stablediffusiontutorials.com/2024/03/dreamb…

🕹 GameGen-O: 🤯👇 Open-world Video Game Generation . . . . Code on GITHUB (coming soon): github.com/GameGen-O/Game…

ComfyUI-WhatsappNotifier :😵👇 Node to send messages via WhatsApp, used in a workflow, in the Queue of the final prompt. -Helps to get notified when u are running heavy workflows on COMFY..... Here is the workflow and installation guide: github.com/I210I/ComfyUI-…

PuLID-FLUX provides a tuning-free ID customization solution for FLUX.1-dev model 😵😃 . . . . More details on their REpo:👇😃 github.com/ToTheBeginning…

A quick update that we have release the inference code and model for DepthCrafter. Code: github.com/Tencent/DepthC… Model: huggingface.co/tencent/DepthC… Project page: depthcrafter.github.io

Fast Upscaler is the newest addition to our suite of upscale services. This feature boosts image resolution by 4x in just about 1 second using advanced predictive and generative AI—delivering fast, high-quality results without the need for extensive processing. It can be used

🖌 DrawingSpinUp: 3D Animation from Single Character Drawings 🛸 Thanks to Jie Zhou ❤ Chufeng Xiao ❤ Miu-Ling Lam ❤ Hongbo Fu ❤ 🌐page: lordliang.github.io/DrawingSpinUp/ 📄paper: arxiv.org/abs/2409.08615 🧬code: github.com/LordLiang/Draw…

Stable Assistant now enables you to apply specific styles when generating videos! With the "Generate with Style" feature, you can: 🎨 Apply uniform stylistic elements across different content types, from videos to images and 3D designs. ⏱️ Save time by automating style-based

KAN be strong! 🚀 KAN faced scalability issues. We tackle this with 3 simple tricks. We introduce the Kolmogorov-Arnold Transformer (KAT). By combining KAN with Transformers, we've built a much stronger model. 💪 paper: huggingface.co/papers/2409.10… code: github.com/Adamdad/kat

FLUX Gym training (AI tookit+KohyaSS) Update: 😵 Now publish your Flux LORA to your HF repo with 1-click 👏 Follow Flux training tutorial if you haven't yet....👇😃 stablediffusiontutorials.com/2024/08/flux-l…