A neat feature of KerasHub: you can create KerasHub pretrained components straight from the base classes (tokenizers/backbones/classifiers/etc). Write code once, and switch the model later. Similar vibe to Transformer autoclasses and pipelines, but unlike pipelines these
OPEN SOURCE
-
Meta FAIR Releases New Models, Benchmarks, and Datasets
By
–
Announcing the newest releases from Meta FAIR. We’re releasing new groundbreaking models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience.
— AI at Meta (@AIatMeta) 14 mai 2025
1️⃣ Open Molecules 2025 (OMol25): A dataset… pic.twitter.com/PAmnNgTVnBAnnouncing the newest releases from Meta FAIR. We’re releasing new groundbreaking models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience. Open Molecules 2025 (OMol25): A dataset
-
Stable Audio Open: Weights, Code, and Research Paper Released
By
–
Weights, code, and paper @HuggingFace
: https://
huggingface.co/stabilityai/st
able-audio-open-small
… @Github
: https://
github.com/Stability-AI/s
table-audio-tools
… @arXiv
: https://
arxiv.org/abs/2505.08175 -
Stable Audio Open Small: On-Device Music Generation for Smartphones
By
–
Today we’re open-sourcing Stable Audio Open Small, a 341M-parameter text-to-audio model optimized to run entirely on @Arm CPUs.
— Stability AI (@StabilityAI) 14 mai 2025
This means 99% of smartphones can now generate music-production samples in seconds, right on-device with no internet required.
Built for fast,… pic.twitter.com/LryqPlfwjjToday we’re open-sourcing Stable Audio Open Small, a 341M-parameter text-to-audio model optimized to run entirely on @Arm CPUs. This means 99% of smartphones can now generate music-production samples in seconds, right on-device with no internet required. Built for fast,
-
Self-Organizing Teams Drive Modern HR Digital Transformation
By
–
Adaptable and self-organizing teams are the secret sauce for modern HR leaders. We share what we've learned firsthand https://
bit.ly/4iVMcSJ #foss #drupal #digitaltransformation -
ChatGPT Reveals Its System Prompt Blueprint
By
–
ChatGPT just revealed its internal system prompt. It’s the blueprint that powers everything it does and now you can study it, tweak it, and build better prompts. I broke it all down in this System Prompt Guide. Like + comment “System” and I’ll DM it to you.
-
Alibaba Open-Sources Wan2.1-VACE Video Creation Model
By
–
Alibaba is open-sourcing Wan2.1-VACE tonight (local time)! Judging by the teaser poster, it’s an all-in-one model for video creation and editing. Big news incoming!
-
ByteDance DanceGRPO advances vision generation; Qwen3 tech report released
By
–
May 14 – Today’s Highlights (May 14) ByteDance applies GRPO to vision generation for the first time. The new DanceGRPO method enables faster convergence and higher visual fidelity. The code will be released. https://
arxiv.org/abs/2505.07818 Qwen3’s full tech report is out. -
Dr. Lance Martin Leads LangGraph 101 Workshop at Interrupt
By
–
Our own Dr. @RLanceMartin kicking off workshops at Interrupt with LangGraph 101! Great session demystifying concepts like workflows vs agents + human in the loop, and shows off LangGraph Studio and how it can help devs iterate on reliable agents. Next workshops in
-
Open-Source AI Image Editor ‘IceEdit’ Runs Locally and for Free
By
–
Une IA open source qui modifie les images mieux que Gemini ou GPT-4o ?
— VISION IA (@vision_ia) 13 mai 2025
IceEdit transforme n’importe quelle photo avec un simple prompt. Suppression d’objets, changement de style, d’éclairage, d’expressions… en 10 secondes.
Le plus fou ? Ça tourne en local. Gratuitement.… pic.twitter.com/WR2E5jObqLAn open-source AI that edits images better than Gemini or GPT-4o? IceEdit transforms any photo with a simple prompt. Object removal, style changes, lighting, expressions… in 10 seconds. The craziest part? It runs locally. For free. Explanations: