We (Finegrain) had a small documentation sprint on our Open Source model adaptation microframework Refiners, and now we have a website up: refine.rs If you attend #FOSDEM come meet us on Sunday! fosdem.org/2024/schedule/…

Stickers time @ #fosdem2024 #curl

Thrilled to present Refiners at #FOSDEM's AI and ML devroom last week-end! 🎉 Check out our brand new docs (refine.rs) and the release of v0.3.0 🚀 #AI #ML

We were impressed by StyleAligned from Andrey Voynov and team at Google AI - so we added it to Refiners: blog.finegrain.ai/posts/implemen…. Check it out!

Just published a post on Finegrain's blog about achieving faster inference with Refiners using Latent Consistency Models and SDXL Lightning. blog.finegrain.ai/posts/faster-d…

Made a Hugging Face space to demonstrate one of the models we trained at Finegrain: the Object Eraser. Pretty happy everyone can try it publicly at last. :) huggingface.co/spaces/finegra…

FLUX.1 schnell / Finegrain Enhancer x2

🤯 Photoshop, but with your words Finegrain just dropped a new model to create HD cutouts for any object in an image using just a text prompt ⚗️ Synthetic data-trained 📜 MIT license Tested on a busy pic—mind-blowing results! #AI #ImageEditing

Finegrain's take on background removal: controllable (what is background vs. foreground often depends on users) and Open Source (MIT License)! Feel free to give it a try on our dedicated Hugging Face space: huggingface.co/spaces/finegra…

Object Cutter Create high-quality HD background removal for ANY object in your image with a text prompt or bounding boxes!

Inspiring Hacktober hackathon at the @Shopware HQ. I was getting all the feels during the demos. So great to have so many committed (😉) community members and current & ex-employees spend two days learning & coding. Open source is organicly fantastic & inherently empowering!

The goal of Finegrain has always been to *automate* image edition. One way to do this is by using our API with LLM tool calling. Here is an example of what you can do with ChatGPT in just a few days (the code is open source: github.com/finegrain-ai/f…)

A big part of training image models is managing image data: curating and annotating datasets, evaluating model outputs... Watch my colleague MBrenon present how we (Finegrain) self-host Voxel51 and CVAT to achieve this: youtube.com/watch?v=3UGRRd…

✨ Lights on, AI activated ✨ Upload ANY photo with lamp(s). Watch this app instantly illuminate it. No Photoshop. No editing skills needed. Just pure AI magic.🔥 Try Finegrain Light Switcher now.

torch2caffe by Andrew Tulloch at Facebook github.com/facebook/fb-ca… (caffe2torch already exists via github.com/szagoruyko/loa… )

New blog post: “Generating unque IDs: an easy and reliable way” => antirez.com/news/99

Google's Inception-v3 model for imagenet classification now available in Torch format, thanks to Moodstocks Inc. buff.ly/1SA36Xl

I'm excited to announce OpenFace 0.2.0! Check out the details in my blog at bamos.github.io/2016/01/19/ope… #Python #Torch #DeepLearning

Twitter releases their distributed deep learning secret sauce that makes AI training super fast. blog.twitter.com/2016/distribut…

Just answered this LuaRocks survey: bit.ly/luarocks