I'm back to building! Excited to share I've started a company with Paril Jain and Luke Holoubek. We're building bots that do chores so you don't have to. Everyone is busy. Bots can help.

So many things compete for our time - commutes, longer working hours, and the complexities…

yo database forums are the craziest places on the internet

like bro what do you mean 'create table is 1000x slower when the table name starts with a number, in v3.43.3 compared to v3.45.2.' how did you even notice that

sqlite.org/forum/forumpos…

Linkedin cannot be real

bro on his way to design data-intensive applications

Introducing Meta Llama 3: the most capable openly available LLM to date.

Today we’re releasing 8B & 70B models that deliver on new capabilities such as improved reasoning and set a new state-of-the-art for models of their sizes.

Today's release includes the first two Llama 3…

Exactly a year ago, I pushed the first commit for a side project I was working on with pradhit and omar. Today, we accepted a check that closes our seed round after having just finished the most thrilling life experience Y Combinator. Get after it kids.

Today, Boom announced the successful flight of XB-1, the world's first independently developed supersonic jet, at the Mojave Air & Space Port in Mojave, California. Flown by Chief Test Pilot Bill 'Doc' Shoemaker, XB-1's inaugural flight marks a major milestone toward the return…

YC W24's @Reform_hq is building Retool for logistics software. It makes it easy to build the exact Transportation Management System your logistics business needs at a fraction of the time and cost.

ycombinator.com/launches/KYm-r…

Congrats on the launch Saif Abuhashish omar & pradhit!

sama: i need money to make AI chips

VC: how much?

sama: $7 trillion

VC: what??????

sama: i will buy Taiwan

VC:

🚀Introducing The LLM Inference Provider Leaderboard leaderboard.withmartian.com - a live-updated, unbiased eval of API Inference products.

Featuring: Abacus.AI, Anyscale, DeepInfra, Decart, Fireworks AI, Lepton AI, Together AI, Perplexity, Replicate, as…

It’s rare but powerful for a company name to become a verb:

Googling = searching
Ubering = getting a ride
Zooming = videoconferencing
Boeing = falling apart

I’m very excited to share our work on Gemini today! Gemini is a family of multimodal models that demonstrate really strong capabilities across the image, audio, video, and text domains. Our most-capable model, Gemini Ultra, advances the state of the art in 30 of 32 benchmarks,…

gpt-4-pro-max

ok this is outrageous and I don't know how they're getting away with it every year 😤

Added some new components (fine-tuning embeddings, lexical search, reranking, etc.) to our production guide for building RAG-based LLM applications. Combination of these yielded significant retrieval and quality score boosts (evals included).

Blog: anyscale.com/blog/a-compreh…

⛓️Chain of Verification

A great new paper from Meta on a prompting technique to reduce hallucinations

🦜🔗Sourajit Roy Chowdhury implemented this in LangChain **along with some improvements**

📃And he wrote a blog on it

🧵Lets dive in (this is why I love the LC community!)

A cool thread yesterday used GPT4 ($50), a 500-word ReAct prompt, and ~400 lines of code to finetune Llama2-7B to get 26% HotPotQA EM.

Let's use 30 lines of DSPy—without any hand-written prompts or any calls to OpenAI ($0)—to teach a 9x smaller T5 (770M) model to get 39% EM!

🧵

Michael Ehrmantraut Um.

Dude.

I invented the UML.

One major way to improve your RAG system is to fine-tune your embedding model ⚙️

We’ve created a full repo/guide (Simon Suo) on fine-tuning embeddings over any unstructured text (no labels needed) 🌟

5-10% improvement 📈 in evals + runs on your MacBook!

github.com/run-llama/fine…