Solo dev reverse-engineered Google's billion-dollar algorithm in 7 days
Google published the paper that crashed memory stocks worldwide. Then shipped zero code.
Tom Turney read the math, opened his terminal, and built the whole thing with Claude - then made it faster than Google promised.
Day 1-3: Core algorithms, 141 tests, Python prototype
Day 3-5: C port into llama.cpp, Metal GPU kernels
Day 5-7: Speed optimization from 739 to 2747 tok/s
That's a 3.7x speedup through pure engineering:
> fp32 → fp16 WHT
> half4 vectorized butterfly ops
> graph-side rotation
> block-32 storage layout
Then he added his own research on top:
> Sparse V: skip 90% of value decompressions at long context
> Asymmetric K/V: keep keys precise, compress values harder
> Temporal decay: old tokens get lower precision automatically
Result: 35B model running on a MacBook with 4.6x compressed cache.
613 GitHub stars in a week. Google still hasn't released their own code.
Check out this post on the Microsoft Tech Community : MVP Tech Gathering Station: A Platform to Spark Your Dreams - Microsoft Community Hub https://t.co/0dqmzdGClo
Passion for technology
Passion for sharing
Passion for making changes
That’s what MVP is all about ❤️
Nice Github repo - AI Agent in your terminal with local tools, and vision.
As the agent has access to tools so it can run shell commands, execute code, read/write files, and more, enabling them to assist in all kinds of development and terminal-based work.
introducing apollo, a new project i've been working on that visualizes topics or concepts in @3blue1brown style videos, all ai-generated.
@nextjs framework
@GroqInc inference
supports @AnthropicAI 3.5 sonnet & @OpenAI gpt-4o
integrated with @langchain
inspired by Chris Abey
This demo is insane.
A student shares their iPad screen with the new ChatGPT + GPT-4o, and the AI speaks with them and helps them learn in *realtime*.
Imagine giving this to every student in the world.
The future is so, so bright.
@tim_cook I think the ad would work much better if it was reversed. All the objects should be expanding out of the iPad rather than being crushed into it
made this edited version in five minutes (thanks iMovie!)
DREAM - a Distributed RAG Experimentation Framework 🧑🔬
Building RAG comes with a lot of knobs that you need to tune, and it’s important that you setup the right experimentation infrastructure to build production RAG.
This project by Aishwarya Prabhat provides a comprehensive full-stack blueprint for letting you run experimentation and evals in a distributed manner to pick the combination that works best.
Uses the following architecture:
✅ Ray (@anyscalecompute) - distributed compute
@llama_index - advanced RAG techniques
✅ Ragas (@Shahules786) - synthetic data + evals!
✅ MinIO - store data + artifacts
✅ @MLflow - experiment tracking
✅ Project Jupyter - interactive experimentation against Ray
✅ ArgoCD - deploying tooling to k8s cluster
There’s sample code so you can follow along yourself - check It out!
Full post here: https://t.co/twl5N4tAkM
Github project: https://t.co/N1NGkqsKgt