apolinario 🌐
@multimodalart
ML for Art and Creativity, working @HuggingFace ([email protected])
Excited to introduce LEDITS++, a novel way to edit real images with precision ✏️ - Multiple edits ✂️🔁 - Automagic free masking 🪄🎭 - 🆕 DPM-Solver fast inversion 🔀⚡ 🤗 Try it: huggingface.co/spaces/editing… 🔗 Project: leditsplusplus-project.static.hf.space/index.html 📝 Paper huggingface.co/papers/2311.16…
Sharing InScene, a Flux Kontext LoRA that lets you generate shots from within an input image's scene! It allows for very precise instructions while keeping characters, scenes and styles consistent! Great for anchor images + coverage shots. Link to model + dataset below!
follow this thread for incredibly under hyped open source gems
Big events bring big moves. WAIC, one of China’s top AI events, starts tomorrow. Got a feeling we’ll see a wave of new models and fresh AI moves in the coming days👀
✨ Second day of WAIC: SmallThinker🔥 a no GPU needed on-device MoE language models, built from the ground up for local AI, by @sjtu1896 and Zenergize AI. huggingface.co/PowerInfer/Sma… huggingface.co/PowerInfer/Sma… ✨ 4B (0.6B active)/ 21B (3B active) ✨ Blazing-fast CPU inference…
my weekend hyperfocus: Wan VACE Fast 💨 14B VACE, 3 control modes, 6 steps ⚡️ now on Spaces ✨ demo link 👇
I'm building a demo for Wan 2.1 VACE + Phantom FusionX LoRA using 🧨diffusers let's you run Wan VACE with multiple conditions using only 6-8 steps! 💨💨 look at this gpu poor little dancing reachy
It was missing, so I added @AnthropicAI Opus 4 Thinking and @OpenAI o3 benchmark results to the comparison mix chart 🆚🔎 Vibe check pending, but on benchmarks it seems that we got an open model competitive with Opus 4 / o3 / Gemini 2.5 🤯
🚀 We’re excited to introduce Qwen3-235B-A22B-Thinking-2507 — our most advanced reasoning model yet! Over the past 3 months, we’ve significantly scaled and enhanced the thinking capability of Qwen3, achieving: ✅ Improved performance in logical reasoning, math, science & coding…
I'm building a demo for Wan 2.1 VACE + Phantom FusionX LoRA using 🧨diffusers let's you run Wan VACE with multiple conditions using only 6-8 steps! 💨💨 look at this gpu poor little dancing reachy
Wow the new qwen reasoner at only 232B params is as good as the top closed frontier lab models Big day for OS
It was missing, so I added @AnthropicAI Opus 4 Thinking and @OpenAI o3 benchmark results to the comparison mix chart 🆚🔎 Vibe check pending, but on benchmarks it seems that we got an open model competitive with Opus 4 / o3 / Gemini 2.5 🤯
great lora
As promised, I’ve open-sourced the Tattoo LoRA! If you’re a dev who wants to integrate it into your app, or you just wanna see how a tattoo would look on yourself, feel free to use it however you want. Would appreciate it if you give credit too ofc. 🖤
Fast LoRA inference for Flux with Diffusers and PEFT 🚨 There are great materials that demonstrate how to optimize inference for popular image generation models, such as Flux. However, very few cover how to serve LoRAs fast, despite LoRAs being an inseparable part of their…
this is all you need to use the newest Qwen3-Coder-480B + cli using @huggingface Inference Providers on benchmarks, it's competitive with Claude Code, now it's vibe check time ✨
Who's gonna take the next-generation open image generation crown? 👑 Both @reveimage and @recraftai could be leading image generation now and be the top-of-mind imo. They could've 10-100x the impact of their "red panda" 🐼 and "halfmoon" 🌙 leading the charts... if they had open…
why do all llms converge to making purple gradient websites? 🤔
ADOS Exhibit — a night of AI art, digital worlds, and interactive experiments. Presented by @banodoco & @LTXStudio in collaboration with 204.ai & RnA Studio Big Thanks to @hannahsubmarine , @peteromallet and to everyone who explored with us!