Spent the last 2 weeks exploring JAX(I've some prior experience) for no particular reason, for most of the time Numpy outperforms it for a few things I tried esp CPU and low GPU requiring tasks, but I'll try to train/inference in big models like Gemma to see the difference.

If you hated PyTorch for not being functional, then go for JAX.

I really want to know if anyone worked with JAX extensively and see how your experience was.

Here is a very good notebook to learn more by some friends: Notebook

Oh and Deepmind uses JAX😉

Better Auth is 500 stars away from 10k stars ✨ could you please give us a star if you haven't ;)

https://github.com/better-auth/better-auth

Llama 3.2 400M Amharic

This is a smaller version of the Meta's Llama-3.2-1B decoder transformer model pretrained from scratch for 23 hours using a single A100 40GB GPU and 274 million tokens of Amharic text.



https://huggingface.co/rasyosef/Llama-3.2-400M-Amharic

Hey guys good news :)

Better Auth has been accepted into Y Combinator's Spring 2025 batch (X25)! 🎉

Myself and @kinfishfarms, will be part of YC's first spring batch. Super excited and thanks everyone here for being part of my journey so far :)) but a lot more to come!

Congrats @beka_cru on this🎉, let's not make fun of him at least for today 😂

OpenAI is the company that's going to take us to the next chapter !!!

Btw how can I make my papers title like this, or something quirky like "Attention is all you need" 😂

100k is a lotttt.

Hasab AI 🔥

Here is a great practical use case of ML in Ethiopia. The inference is really optimized 🔥.

Take a look at the demo.

https://www.hasab.ai/

Why is the field of robotics so slow in the past 15 years?

Boston Dynamics is the only one I can think of.

Started reviewing applicants today. So many great people.

Let's have fun😁

Text to Bark, what a breakthrough in human civilization.

https://x.com/elevenlabsio/status/1907014022009876508

This got me fooled tbh

take it easy bro, where is the fun :)

My favorite open source model series. The Llama 4 herd is out

https://ai.meta.com/blog/llama-4-multimodal-intelligence/

What are some *practical* ways for reducing input tokens (e.g. chunking, summarizing, selective filtering) ? Just looking for something that worked well.

I'm craving a PhD

Anthropic’s Evaluation of Chain-of-Thought Faithfulness

https://www.marktechpost.com/2025/04/05/anthropics-evaluation-of-chain-of-thought-faithfulness-investigating-hidden-reasoning-reward-hacks-and-the-limitations-of-verbal-ai-transparency-in-reasoning-models/

I'm not signing a doc to access a so called "open source dataset", it's also very clear that i won't be able to develop a model with such small data, let alone use for commercial purposes.