🚀Excited to announce F Lite: a new open-source text-to-image model by @freepik and @FAL! The first at this scale that’s both open-source and trained exclusively on licensed, high-quality data.🧵
MAI-Thinking-1: el primer modelo fundacional de Microsoft SuperIntelligence (MSI). El equipo de MSI Madrid ha participado activamente en su diseño y entrenamiento! Y buscamos más AI engineers excepcionales para crear los próximos modelos de Microsoft desde España! DM para+info 🚀
microsoft MAI tech report is a gold mine, one of the most transparent for a model at this scale.
this model uses zero synthetic data or distillation from previous models. this means reasoning, agentic behavior, tool use are all learned fully during post-training with no cold start. bold choice that makes it harder and requires more iterations to reach sota, but you get FULL control over your model series and it proves they are serious about being a frontier lab.
the tech report is insanely detailed and precise about numbers. to give an example, they give the exact MFU across all the iterations of the model, with the exact changes etc. they also share the full scaling ladder recipe, to my knowledge this is the first time i've seen this in a tech report at this scale
let's look at all of this in this likely very long thread 🧵
Where AI meets creativity. For the fourth time
Upscale Conf returns to San Francisco, June 3–4
Two days to unpack what AI is changing about creative work
Talks. Hands-on workshops. The space where the industry connects
Grab your ticket ⬇️
Es este de los mejores artículos técnicos que he leído en meses?
No por lo avanzado, que también, pero por lo práctico, útil, cercano y por el pragmatismo que transmite especialmente estos meses que la mayoría de contenido es vacío.
Si Mercadona estuviese en SF estaríamos dando palmas con las orejas
https://t.co/sGz32G2SLH
Seedance 2.0 is NOW on Freepik, the world's biggest AI creative suite
Business and Enterprise accounts can start generating today, individual users are next
Verification is required for Business. Available in 150+ countries, except US and some exceptions
Get access today 💥
Your next 3D photo shoot will be done with AI
3D Scenes generates full environments from any image
→ Place your objects in the scene
→ Move the camera like a real shoot
→ Consistent lightning and detail across every angle
Available now on Freepik 👇
Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc
@JawsVibe@HoidBroom@freepik@pzf_ai Yes, it makes sense. We will work to increase limits in future releases. Best way to estimate: I would upload a 1 minute video clip, set the parameters you want, and then check the credits. Then multiply this credits by the amount of total minutes your entire video has.