paulosalem

6 ms·

1. 1.▲
   by paulosalem 1mo ago

   Hi folks, WeaveMark is an open-source specification language for prompts: you specify abstract intent (using Markdown + special directives) and an LLM-based processor makes it concrete (either by "compilation" or - optionally - ac
2. 2.▲
   Show HN: WeaveMark – a specification language for reusable prompts(github.com)

   2 points by paulosalem 1mo ago|1 comments
3. 3.▲
   TinyTroupe: An LLM-Powered Multiagent Persona Simulation Toolkit (OSS Paper)(arxiv.org)

   3 points by paulosalem 1y ago|1 comments
4. 4.▲
   by paulosalem 1y ago

   We've recently released a preprint paper about our TinyTroupe open-source project developed at Microsoft. This is still very much under research and development, but the paper is meant to consolidate much of the principles and findings
5. 5.▲
   TinyTroupe, a new LLM-powered multiagent persona simulation Python library(github.com)

   143 points by paulosalem 2y ago|48 comments
6. 6.▲
   by paulosalem 4y ago

   I shared my tutorial (with code) on the topic before, but here's it again (just sent a PR to this repo too). https://medium.com/data-science-at-microsoft/building-gpt-3-... In short, Prompt Engineering is just one
7. 7.▲
   by paulosalem 4y ago

   Indeed. I do know Azure has special provisions for compliant computing for a couple of industries. Regarding GPT-3, besides the original OpenAI offer, it is now possible to consume it from Azure instead. My understanding is that this is not
8. 8.▲
   by paulosalem 4y ago

   Hey folks, FYI, I'm the author.
9. 9.▲
   by paulosalem 4y ago

   My take on what data scientists or software engineers eager to use GPT-3 should know to succeed.