Izzet Mustafaiev

A ... Software Engineer ... @EPAMSystems

[Alex Cheema @alexocheema]

AGI at home

Running DeepSeek R1 across my 7 M4 Pro Mac Minis and 1 M4 Max MacBook Pro.

Total unified memory = 496GB.

Uses @exolabs distributed inference with 4-bit quantization.

Next goal is fp8 (requires >700GB)

[Jeremy Howard @jeremyphoward]

Here, in full directly on Twitter, is "A Hackers' Guide to Language Models". This 90 minute tutorial is designed to be the one place I point coders at when they ask "hey, tell me everything I need to know about LLMs!"

It covers both @OpenAI models and open source ones in depth.

[Andrew Ng @AndrewYNg]

New short course on Fine-tuning LLMs! Many developers are moving beyond only prompting, to also fine-tuning LLMs - that is, taking a pre-trained model and training it further on your own data, which can deliver superior results inexpensively. In this course, @realSharonZhou, CEO of Lamini (disclosure: I’m a minor shareholder) shows you how to recognize when fine-tuning can be help, and how to train an open-source LLM on your own data. I hope you enjoy the course! https://t.co/3MDfIvmw6t

[Alvaro Cintas @dr_cintas]

This paper is going viral.

Why?

Authors were able to find a way to enable the Recurrent Memory Transformer to retain information across up to 2 million tokens 🤯

In simpler words, it can process and remember vast amounts of data, significantly more than before.

Just so you have an idea, GPT-4 handles 32K tokens (~50 pages of documents), while the entire Harry Potter series is ~1.5M tokens.

Imagine the potential implications of this:

• AI could write entire novels, not just blog posts
• It could aid in complex scientific research by analyzing vast quantities of data
• Enhance customer service by retaining years of interaction history
• It could store and recall the entirety of your life experiences!!

[Andrej Karpathy @karpathy]

🔥 New (1h56m) video lecture: "Let's build GPT: from scratch, in code, spelled out."
https://t.co/2pKsvgi3dE
We build and train a Transformer following the "Attention Is All You Need" paper in the language modeling setting and end up with the core of nanoGPT.

[Vlad Mihalcea @vlad_mihalcea]

Donald Knuth's "premature optimization is the root of all evil" is about micro-optimizations.

You still need to design your application so that it reads and writes data efficiently. That's not premature optimization. That's software engineering.

[The Micronaut Framework @micronautfw]

Don't miss @graemerocher's 2-hour #Micronaut training workshop at #OracleDevLive — #Java next week!

https://t.co/Bfwff0QlpR

[Stephan Janssen ☕️🧠🧞‍♂️ @Stephan007]

I have started an open source Sports Video Analysis project using #MachineLearning. If you want to learn, join or make suggestions then please read this article @ https://t.co/v9QlW6rcce #Basketball #OSS 🏀💪🏻

[Bilgin Ibryam @bibryam]

Great books should be #free for all, like open source software! Monetized indirectly, through alternative means and sponsors!
This new wave is coming, no more excuses for not reading!

[Sid Jayakumar @sidfix]

Really excited about our latest work showing that large Transformer-XLs can be used in RL agents. We show SoTA performance on DMLab with gated transformers and a few small changes. Led by Emilio as an internship project! @DeepMindAI

[Greg Brockman @gdb]

When we started working on robotics in 2016, there was controversy about how to make robots that learn.

Gather experience from *many* physical robots, or maybe *somehow* transfer knowledge from simulation?

@woj_zaremba bet on sim, and it's worked better than any of us imagined.

[Chris Olah @ch402]

Do you formally know Monte-Carlo and TD learning, but don't intuitively understand the difference? This is for you.

https://t.co/2DR75rK40u (with @samgreydanus)

[Miles Brundage @Miles_Brundage]

Totally crazy that you can squeeze BERT down to 7 MB without much drop in performance - "Extreme Language Model Compression with Optimal Subwords and Shared Projections," Zhao et al.: https://t.co/bGvBqneqFf

[Chuck Svoboda @Chuckernetes]

I'm not sure who made this, but this is amazing. #kubernetes

[eileencodes @eileencodes]

Guess what? @github has been running on Rails 6.0 in production since last Monday. We had 0 customer exceptions during testing. I'm so proud of how solid Rails 6.0 is and that our engineers sent over 100 PRs to this version. Blog post on @github's blog coming soon!

[François Chollet @fchollet]

The TensorFlow 2.0 release candidate has landed: https://t.co/nvt4h3fHFS

[AI at Meta @AIatMeta]

Facebook AI researchers are sharing an all-attention layer to simplify the Transformer model and an adaptive attention span method to make it more efficient. Even with a much simpler architecture, these methods match or improve state-of-the-art results. https://t.co/wCbK4P7Bgv

[Flavio Brasil @fwbrasil]

That's what latency looks like when you unleash Netty's full power! Repeat with me:
I will never use netty threads to execute application code
I will never use netty threads to execute application code