Henok | Neural Nets
1.61K subscribers
233 photos
20 videos
13 files
157 links
Download Telegram
Forwarded from Dagmawi Babi
Media is too big
VIEW IN TELEGRAM
This's so impressive wtfruits!!! πŸ”₯

Lucy β€” Multilingual AI Voice Assistant & Chatbot for Ethiopians
β€’ https://www.linkedin.com/posts/zemenu_lucy-next-generation-ai-ugcPost-7298736880336920576-HD8B

I was impressed at the Amharic TTS and also when bro wrote amharic in english "endezih" and it understood. Not to mention it scrapes telegram channels and understands major Ethiopian languages,
πŸ”₯21πŸ‘1
How can a startup says they work on AI, blockchain, nanotech, quantum computing etc and all while being Africa-based. They're not disrupting the market, they're defying physics.
🀣27😭2😁1πŸ™Š1
If United win today, I don't see any reason why AGI won't happen soon.
😁11🀣6πŸ’―2
Henok | Neural Nets
This is interesting, how is tensorflow beating Jax in a normal 10,000 Γ— 10,000 matrices multiplication(element-wise)? Anway, if you only need per-element operations, use multiply() (very fast) and tf will be faster than jax. If you need actual matrix computations…
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 JAXπŸ˜‰
❀9πŸ‘2
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
πŸ”₯15
Forwarded from Beka (Beka)
Hey guys good news :)

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

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!
⚑13πŸ”₯7πŸ‘1
Congrats @beka_cru on thisπŸŽ‰, let's not make fun of him at least for today πŸ˜‚
😁22🀣5
OpenAI is the company that's going to take us to the next chapter !!!
πŸ”₯14😈1
Btw how can I make my papers title like this, or something quirky like "Attention is all you need" πŸ˜‚
😁15
100k is a lotttt.
😁8πŸ€”1
This media is not supported in your browser
VIEW IN TELEGRAM
Hasab AI πŸ”₯

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

Take a look at the demo.

https://www.hasab.ai/
πŸ”₯28
Why is the field of robotics so slow in the past 15 years?

Boston Dynamics is the only one I can think of.
πŸ’―6
Let's have fun😁
🀣27😁7
take it easy bro, where is the fun :)
😁3
My favorite open source model series. The Llama 4 herd is out

https://ai.meta.com/blog/llama-4-multimodal-intelligence/
πŸ”₯11
What are some *practical* ways for reducing input tokens (e.g. chunking, summarizing, selective filtering) ? Just looking for something that worked well.