Last week I needed a quick way to share a CSV with images in them so .... ... we turned this into a weekend project with Julien Blanchon 😅 deployed it on Vercel this morning :) check it out -> sharetable.xyz/f/6338c93a-185… table component -> Maximilian Kaske 🏓 🙏 ui -> shadcn 🔥

this is so promising 😍 I want to build on this

Anyone running batch jobs for image captioning ? I am testing gpt-4o on Microsoft Azure to get a baseline and I keep getting blocked by their "content management policy" 😂 Apparently a picture of a watch can be flagged as sexual content of "medium severity"... their nsfw algorithm

Several airdrops were launched last week, many are still ongoing, so I created another tool to check the latest airdrops It's free, no need to connect anything, and it helps ensure you don’t miss any of them! Just use it here 👇 checkdrop.byzantine.fi

Found this gem tonight: arxiv.org/pdf/2303.11306 They propose a new way to create synthetics segmentation dataset leveraging self-attention maps. Currently, the go to solution for synthetics dataset creation is currently 3D generation, but it's quite hard to generate a diverse

Change my mind

Introducing Recraft V3 — a revolutionary AI model that thinks in design language. It delivers unprecedented quality in text generation, outperforming models from Midjourney, OpenAI, and others. It’s more than an image generator, it’s a powerful tool with enhanced text

When we look at the AI bubble in a few years and try to figure out when was the top … 🤔 Well that was probably it 🎃

Speaking of bubbles … this one keeps coming back 🤦

In-Context LoRA for Diffusion Transformers ali-vilab.github.io/In-Context-LoR… Apparently it's suitable for training four-panel comics

Automatically disbelieving authorities isn't independent-mindedness. It's just conformism with the sign bit flipped.

Zheng Zeng Hey, this paper is amazing. I've made some Spaces with GPU zero for people to try out: - huggingface.co/spaces/blancho… - huggingface.co/spaces/blancho…

[1/10] 🚨 We present our recent Snap Inc. project: Stable Flow --- A training-free method that performs various types of image editing operations (e.g., non-rigid editing, object addition and replacement) using flow models. Project page: omriavrahami.com/stable-flow

Je sais reconnaître un bâton de mage qui inflige des dégâts de glace, une restauration de mana et 50% de chance de coups critiques 😁 #NotreDameDeParis #RetrouvonsNotreDame

Gandalf essayant d'ouvrir les portes de la Moria