Flux 2 Klein 4B param, dropped at 2 steps... with much higher FPS! I also added a couple of LORA's, I'll mess around with. Such a good diffusion model!show more

🆕 Image super-resolution model just dropped! Superior results even... with a single sampling step. 🔥InvSR: Arbitrary-steps Image Super-resolution via Diffusion Inversion.show more

CharacterSheet LoRA by Alisson Pereira just dropped on HF... and is so good 🎨 feed it one image of a character, photo or illustration, and this FLUX.2-klein LoRA generates a full model sheet: portrait, front, side and back, identity intact ▶️ on Spacesshow more

🎨 We're open sourcing 4 new LoRAs trained with... new fal FLUX.2 [klein] 4B Trainer! 🖼️ Outpaint - Expand your images beyond borders 🔍 Zoom - Select an area and regenerate it enlarged 🧹 Object Remove - Clean removal of unwanted objects 🎭 Background Remove - Seamless background extraction Model weights are on Hugging Faceshow more

Microsoft Asia just dropped Mage-Flow on Hugging Face a... smol 4B model for image generation and editing that matches much larger models in quality and it's fast! 4 steps in < 1s at 1024x1024 (model goes up to 4K) try out on spacesshow more

I might've had WAY too much fun with oatmeal... and a real mess afterwards... Such bulk... 🫠show more

The usual suspect threatens customers at a Naples bar... with scissors and gets a good dose of education. In Naples, you don't mess around.show more

Introducing Jan v3, with updates to Jan Desktop v0.7.6... 💛 Jan v3 is our first v3 model, a 4B base built for fine-tuning and fast local use, with stronger math and coding. This release also includes a small Jan Desktop update, starting with a UI refresh as we move toward a more unified Jan experience. Try it: - Jan v3 is available in Jan Desktop and at - Get the latest desktop app at Model: - Jan-v3-4B: - Jan-v3-4B-GGUF: Thanks Qwen for the base model and Georgi Gerganov for llama.cpp 💛show more

Text-to-image diffusion transformer models learn to align text and... image representations as a byproduct of their conditional denoising task. By taking the dot product between the text and image representations of a DiT model (like Flux 2), you can create rich saliency maps.show more

Last year when Krunal pandya tried to mess with... MS Dhoni by bowling a bouncer ❌ MS Dhoni responded with 2 massive sixes of the next 2 balls 😮‍💨☠️ Never mess with MSD i repeat never ever 😎show more

Along with Season 2 we added Impossibles to skate.... Here’s a look at just a few of the ways you can do it!show more

Clarity Upscaler now works with Flux! 🥳 Clarity started... 1 year ago as an upscaler to convert images with Stable Diffusion 1.5 into high-resolution, photorealistic images It took awhile to make it work with Flux but now it does. With Flux it's even better at upscaling faces while preserving identity and resemblance (a common problem with other upscalers) LoRAs are also supported, so upscaling with a specific style or face works perfectly Now available on Clarity AI and through my own API Link in the reply!show more

🐟🐠NEW MODEL REVEAL!!! 🐠🐟 As we set sail for... the SEA-son 2, I would much rather show the process than hype up an invisible model for you to see at Season 2! Meet the your long forgotten imaginary friend once again with a new look. With a new raggedy/toylike feel 🧸show more

2:30 am and I just started a movie with... a 9 am class in a couple hoursshow more

i dont think i'll ever get over the way... sasuke looked at naruto for such a long time with such an intense yearning expressionshow more

One of my favorite days of camp with the... Packers. Davante Adams, Funchess, Lazard. 3 different types of throws needed for higher probability of completion. I was obsessed with mastering fade types. I also knew I couldn’t give Jaire a chance to get his head around.show more

First few tests with Seed Audio 1.0 from ByteDance... and I'm loving it. In initial tests voice acting and foley seems much better than native audio from SeeDance 2. Also very cheap to experiment with. A 15 second audio test was only a couple of cents. Definitely something I could integrate into my workflow.show more

NVidia just released PiD: super resolution in pixel space... directly from model latents 🔎 4X resolution for any generated image, FAST! 🏎️💨 FLUX.1, 2 and Z-Image (Qwen Image coming) of course, i built a demo: generate 4K images with Z-Imageshow more

I do love .stl tricks basically making voids in... the model forcing the slicer to generate walls to reinforce specific areas of a model also maxing out that Prusa3D Core one L bed space with 2 astromech partsshow more

I have playing with Black Forest Labs Black Forest... Labs FLUX 3 for a few days now. (Thank you for the early access!) And I played with it... a lot. Some may say, too much. 😅 Here are some of my favorite gens. Needless to say, I love this model. Audio on! More in 🧵show more