Introducing Kitten TTS, a SOTA tiny text-to-speech model. - Just 15M parameters .- Runs without a GPU.- Model size less than 25 MB.- Multiple high-quality voices.- Ultra-fast - even runs on low-end edge devices. Github and HF links below

I've left Meta Reality Labs. Miss working with the best team for realistic avatars. Damn, I definitely miss using thousands of GPUs and running super large-scale training experiments. Now back to being GPU poor, but excited to start something new: building tiny AI models for

Repo link :

We launched on possibly the worst timing ever - right when OpenAI, Google DeepMind, and Anthropic all had major releases. Somehow we still made it to #1 on Hacker News. Thanks to everyone who checked us out! 🙏

Also, we would love to collaborate with y’all. If you want to potentially use this model somewhere, shoot me a DM.

Someone deployed it on a browser within few hrs!

link:

The open-source community is amazing. We released a tiny TTS model under 25MB. And someone made a web-browser demo within few hours. Yes, its running locally in the web-browser . Thanks clowerweb!

Github:  Huggingface: 

link :

Open source speech AI models are getting incredibly powerful. So we built a free Mac app that makes them easy to use! 🎤✨

ARC-AGI-3 preview is out. It consists of a few games which AI agents can play. It's a good test of few shot learning, generalization and adaptation to new scenarios. > No instructions given to AI.> Easy for humans, but hard for AI. For perspective: most RL algos have to play

MLX can now run on CUDA devices! 🔥. A game changer for cross platform local AI

Right now a lot of LLM usage is for very basic tasks. eg. Extracting info from text . We will soon have several small task specific LMs which are faster, cheaper and more accurate!.

Would love to hear your thoughts:

Introducing NamiGen, a fully local Mac App to run state-of-the-art speech models locally. - Easy to install and use.- Local alternative to ElevenLabs.- High quality TTS.- Free to use

I was able to create a high quality animated video from a pencil sketch under 5 minutes . 🤯🤯🤯. First turn your sketch to a Pixar style image using ControlNet in @diffusionbee . Then use @LumaLabsAI to turn it into a video. Wow 😮

Right now what we see is the GPT-2 of Video Generation. It's truly impressive, but just wait for GPT-3 and GPT-4.

Building actors is much harder. All these models are really good in interpolating between training data points. It’s much harder to do that for a set of actions, because the space of valid action sequences is far smaller and sparse.