Favorite tweets

how does steve jobs always manage to make me cry https://t.co/BqUE18knPv

— shaurya (@shauseth) Jul 31, 2023

from Twitter https://twitter.com/shauseth

July 31, 2023 at 06:42AM
via IFTTT

Favorite tweets

Finally got a chance to spend some time with the 77-page Llama 2 paper (https://t.co/KNRiVzVJ0J). Here are some takeaways at a glance. Appreciate that someone finally did a comprehensive supervised finetuning vs RLHF evaluation! (Lower right) https://t.co/CXKyIQw0oI

— Sebastian Raschka (@rasbt) Jul 31, 2023

from Twitter https://twitter.com/rasbt

July 31, 2023 at 05:51AM
via IFTTT

Favorite tweets

"Attention is all you need ." https://t.co/NvZA8DzJ5z via @FT

— Scott Kupor (@skupor) Jul 30, 2023

from Twitter https://twitter.com/skupor

July 30, 2023 at 03:04PM
via IFTTT

Favorite tweets

🔧Anthropic Functions A wrapper around @AnthropicAI models to expose support for function calling in the same way as @OpenAI Uses XML encoding under the hood, as per docs thats what Claude is best at. Added some examples for tagging and extraction https://t.co/hTfabV929X https://t.co/mWyKSRWX4H

— Harrison Chase (@hwchase17) Jul 30, 2023

from Twitter https://twitter.com/hwchase17

July 30, 2023 at 08:04AM
via IFTTT

Favorite tweets

Second day of #ICML2023 workshop! We have H2O poster at ES-FoMo workshop (https://t.co/p0PVKe9U8h) that reduces KV cache usage by 80% without increasing perplexity in LLM inference! We have SurCo presentation (@ 10:15 am HST) in SODS workshop (https://t.co/CWwkAwSD3n). It… https://t.co/oHc4qe8jCu

— Yuandong Tian (@tydsh) Jul 29, 2023

from Twitter https://twitter.com/tydsh

July 29, 2023 at 10:59AM
via IFTTT

Favorite tweets

@jandreini1 It was Colony Capital and the titanium in the stadium roof https://t.co/SGoxKh25U0

— Oscar Azocar (@oscarazocar50) Jul 29, 2023

from Twitter https://twitter.com/oscarazocar50

July 29, 2023 at 10:25AM
via IFTTT

Favorite tweets

There’s a lot of vector dbs out there - they all share similarities, but what are the differences between them? This has been a big ask from users (LlamaIndex supports 20+ vector dbs) We’ve created a table that clarifies diffs wrt LlamaIndex! https://t.co/9WOYO2iSGl https://t.co/ibIUmW5tpt

— LlamaIndex 🦙 (GPT Index) (@llama_index) Jul 29, 2023

from Twitter https://twitter.com/llama_index

July 29, 2023 at 09:28AM
via IFTTT

Favorite tweets

MIT university is offering FREE education in Data Science. Courses cover: - Python - Data Science - Machine Learning - Statistics - Linear Algebra & - Deep Learning Learn from the best at a free of cost! Thread🧵👇 https://t.co/9eCx9Qr7JT

— Sumanth 🚀 (@Sumanth_077) Jul 29, 2023

from Twitter https://twitter.com/Sumanth_077

July 29, 2023 at 07:42AM
via IFTTT

Favorite tweets

In 2001, Warren Buffett gave a talk at the University of Georgia. He asked them the most Warren Buffett question ever: • If you could invest in a friend and get 10% of their income for life -- who would you pick? Once the students answered the question, he then asked this:… https://t.co/KIf4soqu0d https://t.co/FQdWZyNrxC

— George Mack (@george__mack) Jul 29, 2023

from Twitter https://twitter.com/george__mack

July 29, 2023 at 07:18AM
via IFTTT

Favorite tweets

I've worked with superconductors for the better part of a decade now in different contexts, from STM condensed matter labs, to particle accelerators, and now fusion. Time for a deep dive on what exactly this miracle-technology unlocks for us a species: 🧵 https://t.co/sRALBgYbkc

— Andrew Cote (@Andercot) Jul 29, 2023

from Twitter https://twitter.com/Andercot

July 28, 2023 at 05:43PM
via IFTTT

Favorite tweets

We just released LLaMA-2-7B-32K, a 32K context model that can be fine-tuned for tasks like doc understanding, summarization & QA! Built with Position Interpolation & our data recipe/optimizations, run inference & fine-tune with up to 3x speedup. Thread👇 https://t.co/qaxFI0Xb9M

— Together AI (@togethercompute) Jul 28, 2023

from Twitter https://twitter.com/togethercompute

July 28, 2023 at 03:05PM
via IFTTT

Favorite tweets

Programming is changing. Fast! The attached code example uses Gorilla, an open-source Large Language Model that specializes in writing API calls. Gorilla kicks GPT-4's butt at this task. It's also much better than ChatGPT and Claude. The team claims the model is very reliable… https://t.co/EZAt7rk0NT https://t.co/25eF0JSpw2

— Santiago (@svpino) Jul 28, 2023

from Twitter https://twitter.com/svpino

July 28, 2023 at 09:42AM
via IFTTT

Favorite tweets

🚨We found adversarial suffixes that completely circumvent the alignment of open source LLMs. More concerningly, the same prompts transfer to ChatGPT, Claude, Bard, and LLaMA-2…🧵 Website: https://t.co/ja2FPw9aad Paper: https://t.co/1q4fzjJSyZ https://t.co/SQZxpemCDk

— Andy Zou (@andyzou_jiaming) Jul 28, 2023

from Twitter https://twitter.com/andyzou_jiaming

July 27, 2023 at 08:22PM
via IFTTT

Favorite tweets

Back in April, I set out to write an explainer on how large language models work. First I had to learn how they worked, and that was harder than I expected. This article is the culmination of 2+ months of in-depth research. I hope people find it useful. https://t.co/qyRDYN1mzl https://t.co/JvXWm7Gl9s

— Timothy B. Lee (@binarybits) Jul 27, 2023

from Twitter https://twitter.com/binarybits

July 27, 2023 at 04:32AM
via IFTTT

Favorite tweets

"Building Generative AI Applications with Gradio" A brand new https://t.co/XKVB4VFqmr mini-course covering image generation, LLMs, and more 🥳 Taught by @AndrewYNg and @multimodalart 😎 Take it for free at: https://t.co/pGZBLnDqyk https://t.co/VAIALNSU09

— Abubakar Abid (@abidlabs) Jul 26, 2023

from Twitter https://twitter.com/abidlabs

July 26, 2023 at 04:21PM
via IFTTT

Favorite tweets

First fine-tuned version of Llama 2 70B are beating @OpenAI GPT-3.5 (70) on the MMLU benchmark. 👉 https://t.co/CK0GYjjtjX https://t.co/8hAAhD7S3L

— Philipp Schmid (@_philschmid) Jul 26, 2023

from Twitter https://twitter.com/_philschmid

July 26, 2023 at 02:21PM
via IFTTT

Favorite tweets

The best introduction to LLM Agents! 🤖 @OfficialLoganK has written one of the best introductions to Large Language Model powered agents. AutoGPT and BabyAGI have been very popular, but there is still some magic to them. This write up provides an ELI-5 introduction to the… https://t.co/NtuU5qIScz https://t.co/aZvjD45v5W

— Sanyam Bhutani (@bhutanisanyam1) Jul 26, 2023

from Twitter https://twitter.com/bhutanisanyam1

July 26, 2023 at 06:17AM
via IFTTT

Favorite tweets

If you are building an LLM application that uses RAG , poor retrieval can be detrimental to its UX. Phoenix now supports passing in your knowledge base as a corpus dataset so that you can inspect how your retrieval system is querying for relevant documents from your vector store. https://t.co/r5W7y7kIJo

— arize-phoenix (@ArizePhoenix) Jul 26, 2023

from Twitter https://twitter.com/ArizePhoenix

July 25, 2023 at 06:32PM
via IFTTT

Favorite tweets

Listen to @jmrphy's discussion with apex accelerationist @Outsideness! https://t.co/BBpoEP7cki https://t.co/zwpemdckEC

— Marc Andreessen -- e/acc (@pmarca) Jul 25, 2023

from Twitter https://twitter.com/pmarca

July 25, 2023 at 12:48PM
via IFTTT

Favorite tweets

A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis Presents WebAgent, an LLM-driven agent that can complete the tasks on real websites following natural language instructions. https://t.co/dI70POSNZr https://t.co/mg16dZoVIU

— Aran Komatsuzaki (@arankomatsuzaki) Jul 25, 2023

from Twitter https://twitter.com/arankomatsuzaki

July 24, 2023 at 09:22PM
via IFTTT

Favorite tweets

Releasing LLongMA-2 13b, a Llama-2 model, trained at 8k context length using linear positional interpolation scaling. The model was trained in collaboration with @theemozilla of @NousResearch and @kaiokendev1. https://t.co/KGkEkEcFTx

— Enrico Shippole (@EnricoShippole) Jul 24, 2023

from Twitter https://twitter.com/EnricoShippole

July 24, 2023 at 09:07AM
via IFTTT

Favorite tweets

Enjoy! https://t.co/fV2QMDVwUk

— Marc Andreessen -- e/acc (@pmarca) Jul 23, 2023

from Twitter https://twitter.com/pmarca

July 23, 2023 at 02:11PM
via IFTTT

Favorite tweets

The weirdest WW2 coincidence: Japan's random balloon bombing campaign hit the Manhattan Project plutonium production site in Washington state, knocking out the cooling system. If it wasn’t for conservative engineering, the reactor might have melted down before the Trinity test! https://t.co/maJZqirS01

— Ethan Mollick (@emollick) Jul 23, 2023

from Twitter https://twitter.com/emollick

July 23, 2023 at 02:03PM
via IFTTT

Favorite tweets

Everywhere I look I see policies put in place to eliminate the disruption & weirdness that AI brings. These policies are not going to work. And, even worse, the benefits of AI are going to be greatly reduced by trying to pretend AI is a normal technology https://t.co/hNvAPkhyri

— Ethan Mollick (@emollick) Jul 23, 2023

from Twitter https://twitter.com/emollick

July 23, 2023 at 08:55AM
via IFTTT

Favorite tweets

The definitive guide to Multimodal deep learning! 🙏 Since the GPT-4 demo, multimodal has become one of the coolest domains in our field. This is a 240 page no-nonsense book to the domain, it starts from the basics of individual modalities upto the key details of the domain.… https://t.co/pfPXFNUh6T https://t.co/bkFuT6A2EH

— Sanyam Bhutani (@bhutanisanyam1) Jul 23, 2023

from Twitter https://twitter.com/bhutanisanyam1

July 23, 2023 at 06:13AM
via IFTTT

Favorite tweets

There are some critical data considerations that you must take into account to make your LLM application production-ready 📣 Using naive RAG techniques (naive text chunking, simple top-k retrieval -> LLM) is fine for hackathons, but will lead to lots of failure cases. ⛔️ By… https://t.co/KOfyGIRCYO

— Jerry Liu (@jerryjliu0) Jul 22, 2023

from Twitter https://twitter.com/jerryjliu0

July 22, 2023 at 07:53AM
via IFTTT

Favorite tweets

Runway Gen-2 can now generate videos from a starting image. And the results are incredible! Here are the best 8 examples I’ve found (PLUS tutorial at the end): https://t.co/S4FviOFxHV

— Alvaro Cintas (@dr_cintas) Jul 22, 2023

from Twitter https://twitter.com/dr_cintas

July 22, 2023 at 06:33AM
via IFTTT

Favorite tweets

Easily the best paper on current State of LLMs! 🙏 A 50 page read but it’s not “just another” survey paper, that only documents facts. The authors actually add very useful commentary capturing all aspects of building Large Language Models. Hence the result is a collection of… https://t.co/fHJP6NMvN3 https://t.co/SWfR27rNSX

— Sanyam Bhutani (@bhutanisanyam1) Jul 22, 2023

from Twitter https://twitter.com/bhutanisanyam1

July 22, 2023 at 06:18AM
via IFTTT

Favorite tweets

The most powerful open source model is out!! 🔥 Nous-Hermes-LLaMA2 🔥 - https://t.co/QBRtcPRhZG - GGML: https://t.co/UXJ3DQA9Rx - GPTQ: https://t.co/Qie7n8zzEG war.

— Yam Peleg (@Yampeleg) Jul 21, 2023

from Twitter https://twitter.com/Yampeleg

July 21, 2023 at 12:03PM
via IFTTT

Favorite tweets

This week LLaMa2 got released by @MetaAI 🔥 The week has been busy with lots of great examples, tutorials, insights, and playgrounds for Llama 2! 🤯 👉 https://t.co/AQnNhKu6CH https://t.co/AQnNhKu6CH

— Philipp Schmid (@_philschmid) Jul 21, 2023

from Twitter https://twitter.com/_philschmid

July 21, 2023 at 08:55AM
via IFTTT

Favorite tweets

New to #LLMs? Check out this awesome Beginner’s guide to @OpenAI API by my talented colleague @thedataprof! 🔥 ✔️ Build your own LLM chatbot in @Streamlit using OpenAI's 🐍#Python library ✔️ Create blog outlines, #ChatGPT-like chatbots, and more! 🔗 https://t.co/R2Lr0kd9yx https://t.co/V1vo89j01q

— DataChazGPT 🤯 (not a bot) (@DataChaz) Jul 21, 2023

from Twitter https://twitter.com/DataChaz

July 21, 2023 at 03:46AM
via IFTTT

Favorite tweets

Big News: We've just added Evaluation Functions to Humanloop! This gives you a powerful new set of tools to evaluate model performance, allowing you to build LLM applications with much greater confidence than before. Here’s why this is a big deal 🧵 https://t.co/uUV1QJkEAU

— Humanloop (@humanloop) Jul 20, 2023

from Twitter https://twitter.com/humanloop

July 20, 2023 at 10:50AM
via IFTTT

Favorite tweets

We're launching Custom Instructions in ChatGPT — persistent instructions & context that are applied to every conversation. Gives you more control over ChatGPT & makes it possible for it to remember facts about you: https://t.co/JT86v5BfKs

— Greg Brockman (@gdb) Jul 20, 2023

from Twitter https://twitter.com/gdb

July 20, 2023 at 10:34AM
via IFTTT

Favorite tweets

Possible tip on prompting Llama-2. Try special tokens from llama's generation code (<<SYS>>, <</SYS>>, [INST], [/INST]). Answers seem better w/ them. LangSmith trace w/o tokens linked (also, image left): https://t.co/2pMnVZqyNy w/ tokens (right): https://t.co/s4HYgw3NIX https://t.co/1VQ0G9Vh0e

— Lance Martin (@RLanceMartin) Jul 20, 2023

from Twitter https://twitter.com/RLanceMartin

July 19, 2023 at 09:10PM
via IFTTT

Favorite tweets

Simultaneously running the same chats across: - @openai GPT-4 + Code Interpreter - @Google Bard - @microsoft Bing - @anthropic Claude 2 - @huggingface Chat with Llama 2 SOTA models + full access to capabilities not released in API. https://t.co/heyFPe2O45 h/t @SeanOliver https://t.co/b5x2EweYyN

— swyx @ 🇸🇬 (@swyx) Jul 20, 2023

from Twitter https://twitter.com/swyx

July 19, 2023 at 05:42PM
via IFTTT

Favorite tweets

Private Chat / QA over docs at ~25 tokens / s with 13b Llama-v2 (on Mac M2 max gpu). Using @trychroma vectorDB, @nomic_ai GPT4all embeddings, LLama-v2 Full recipe added to @LangChainAI docs: https://t.co/amzJ9ZcfeE https://t.co/KNgNdxRDqB

— Lance Martin (@RLanceMartin) Jul 19, 2023

from Twitter https://twitter.com/RLanceMartin

July 19, 2023 at 11:33AM
via IFTTT

Favorite tweets

🚨BREAKING: Apple is building an AI chatbot called 'Apple GPT.' The company has developed a framework called "Ajax" and has deployed Apple GPT within internal systems. A consumer product is aimed for next year. Siri is about to get some massive upgrades 👀 https://t.co/ueRL69ICYV https://t.co/6FbKpt7KuZ

— Rowan Cheung (@rowancheung) Jul 19, 2023

from Twitter https://twitter.com/rowancheung

July 19, 2023 at 09:52AM
via IFTTT

Favorite tweets

How does Docker work? The diagram below shows the architecture of Docker and how it works when we run “docker build”, “docker pull” and “docker run”. There are 3 components in Docker architecture: 🔹 Docker client The docker client talks to the Docker daemon. 🔹 Docker host… https://t.co/GhxuKcfJME https://t.co/HXq1aaUiSV

— Alex Xu (@alexxubyte) Jul 19, 2023

from Twitter https://twitter.com/alexxubyte

July 19, 2023 at 07:40AM
via IFTTT