What We’ve Learned From A Year of Building with LLMs – Applied LLMs
A practical guide to building successful LLM products, covering the tactical, operational, and strategic. Also published on O’Reilly Media in three parts: Tactical, Operational, Strategic (podcast). Also translated to Japanese (by Kazuya Kanno) It’s an exciting time to build with large language models (LLMs). Over the past year, LLMs have become “good enough” for real-world applications. And they’
Mixture-of-Agents Enhances Large Language Model Capabilities
Recent advances in large language models (LLMs) demonstrate substantial capabilities in natural language understanding and generation tasks. With the growing number of LLMs, how to harness the collective expertise of multiple LLMs is an exciting open direction. Toward this goal, we propose a new approach that leverages the collective strengths of multiple LLMs through a Mixture-of-Agents (MoA) met
M1MacでOllamaを試したら爆速で驚いた
はじめに いつもNVIDIAが載っているWindowsで楽しくLLMを動かしたり生成AIライフを楽しんでいますが、今回はMacOSでOllamaを入れてLlama3を動かしてみました。 スペック： Apple M1 Pro(16 GB) 少し前だとCUDAのないMacでは推論は難しい感じだったと思いますが、今ではOllamaのおかげでMacでもLLMが動くと口コミを見かけるようになりました。 ずっと気になっていたのでついに私のM1 Macでも動くかどうかやってみました！ 結論、爆速で推論できていたのでとても驚きました。OSS開発に感謝です！ Ollamaとは OllamaとはローカルでLLMを動かすことができるアプリケーションです。 以下からダウンロードできます。 MacOSとLinuxで使うことができます。Windowsもプレビュー版があるみたいです。 #いざ推論 ダウロードができたらシ
8 Google Employees Invented Modern AI. Here’s the Inside Story
Eight names are listed as authors on “Attention Is All You Need,” a scientific paper written in the spring of 2017. They were all Google researchers, though by then one had left the company. When the most tenured contributor, Noam Shazeer, saw an early draft, he was surprised that his name appeared first, suggesting his contribution was paramount. “I wasn’t thinking about it,” he says. It’s always
Efficient Streaming Language Models with Attention Sinks
Deploying Large Language Models (LLMs) in streaming applications such as multi-round dialogue, where long interactions are expected, is urgently needed but poses two major challenges. Firstly, during the decoding stage, caching previous tokens' Key and Value states (KV) consumes extensive memory. Secondly, popular LLMs cannot generalize to longer texts than the training sequence length. Window att
SolidGoldMagikarp (plus, prompt generation) — LessWrong
UPDATE (14th Feb 2023): ChatGPT appears to have been patched! However, very strange behaviour can still be elicited in the OpenAI playground, particularly with the davinci-instruct model. More technical details here. Further (fun) investigation into the stories behind the tokens we found here. Work done at SERI-MATS, over the past two months, by Jessica Rumbelow and Matthew Watkins. TL;DR Anomalou
Multimodal generative AI search | Google Cloud Blog
What is Multimodal Search: "LLMs with vision" change businesses What if large language models (LLMs) had "vision", the ability to understand the meaning of images? Just like we have seen the innovation with LLMs with chatbots and text data, the ability would make another huge impact on businesses by letting LLMs look at and organize millions of images in enterprise IT systems. In this post, we wil
リリース、障害情報などのサービスのお知らせ
最新の人気エントリーの配信
j次のブックマーク
k前のブックマーク
lあとで読む
eコメント一覧を開く
oページを開く
https://www.reuters.com/technology/major-tech-firms-develop-ai-networking-standard-without-major-player-nvidia-2024-05-30/
Mastering GenAI ML System Design Interview (2): Design ChatGPT Memory Feature
Charting the Graphical Roadmap to Smarter AI
ML-at-Scale '23 - Go Beyond GPT-4: Harnessing the Power of LLMs in Industry
GitHub - openai/prm800k: 800,000 step-level correctness labels on LLM solutions to MATH problems
GitHub - langchain-ai/opengpts
機械学習と自動微分 (2023)
Pruning vs Quantization: Which is Better?
Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolation
In search of dispersed memories: Generative diffusion models are associative memory networks
PFN Releases PLaMo-13B Open-Source Large Language Model in Japanese and English for Research and Commercial Use - Preferred Networks, Inc.
Two Ways to Download and Access Llama 2 Locally
These glitch tokens are all near the centroid of the token embedding space. That... | Hacker News