Sublime

"Compounding is magic. Look for it everywhere. Exponential curves are the key to wealth generation" ~Sama Senpai https://t.co/IufjomfRz9

The best paper about Quality Investing? The Equity Compounders by Morgan Stanley It tells you everything you need to know about Quality Investing in 8 pages: https://t.co/a9I4wF51dz

This is the fastest serving engine for LLMs! LMCache cuts time-to-first-token by 7x and slashes GPU costs dramatically. Think of it as a smart caching layer that remembers everything your LLM has processed before. 100% open-source, makes vLLM go brr...... See more

UPDATED: The world's highest quality compounders ⬇️⬇️⬇️ * Thanks to everyone who commented on my previous post. This version reflects your suggestions. https://t.co/joRGPNvdky

Financial RAG Evaluation 🕵️ I added reranking to the pipeline today. As expected, command-r performed even better. Main takeaways: • command-r excels at RAG • cohere reranking is seriously fast • gpt-3.5 slow at... See more

A tiny plug-in memory lets frozen LLMs adapt to domains quickly, with low cost and no retriever. Memory Decoder is a small transformer that learns the behavior of a retriever, then plugs into any base model that shares the tokenizer. It does not touch the base weights. At runtime, the base model and this memory each... See more