Guardrails AI reposted this
Creating LLM apps is easy. Creating production LLM apps is hard -- but it can be easy. We've paired up with Guardrails AI so you can guard your LLMs and view when those guards trigger! Check out the docs for how to get started with both: https://lnkd.in/gxryJhZ3 🙌 Sign up for IUDEX AI here: https://lnkd.in/g3K5FKxz Sign up for Guardrails AI here: https://lnkd.in/gg-QYm2Q
Our co-founder, Safeer Mohiuddin, will be in NYC for a chat on Sep 26th with the Arize AI team! Are you one of our open-source users? Are you building a GenAI platform for your company? We would love to meet up during the week around NYC or at the event on the 26th! https://lnkd.in/gGwVpA5z
Another week, another release! Guardrails v0.5.8 is out! This week’s updates center around the Guardrails CLI. Most importantly, we enabled installing multiple validators at once. Full changelog here: https://lnkd.in/gZagp8KX
Choosing the right method for generating structured outputs is like making orange juice—effort, quality, and trade-offs matter. From function calling to plain-old-prompting, it’s all about balancing precision and performance. 🍊 We wrote a new blogpost about tradeoffs of different strategies of generating structured data https://lnkd.in/gdSE9NzP #AI #StructuredData #NLP
Excited to announce an addition to the Guardrails Hub through a collaboration with BespokeLabsAi! Bespoke’s minicheck is fantastic at detecting hallucinations by identifying sentences in LLM outputs that are not supported by a given context. Use it today by running guardrails hub install hub://bespokelabs/bespoke_minicheck This validator works by invoking the MiniCheck api after splitting outputs into sentences. You can create a free BespokeLabs api key here https://lnkd.in/gBYF_vH5 See the validator here https://lnkd.in/g22dvMge
Guardrails v0.5.7 is out! The main addition to the library is an experimental function to fetch a Guard that runs on the server. We've also done some maintenance - we updated our hub manifest interface and some dependencies. View the full changelog here https://lnkd.in/gEzy73qi
New blog post out by the Bespoke AI team! Hallucination has become a blanket term. This blog post helps clarify and reframe LLM hallucinations in the context of more traditional NLP tasks. We’ve both been thinking this problem for a bit. We’ve been using Provenance-LLM as our primary hallucination detection validator, however we suspect that Minicheck performs better. @bespokelabsai's grounded factuality detector, Bespoke-Minicheck-7B, tops the LLM AggreFact leaderboard. This lightweight model outperforms larger foundation models, including GPT-4 and Mistral-Large 2 on grounded factuality detection. https://lnkd.in/g4tPCAKU
Last chance to sign up to our free webinar tomorrow: https://lnkd.in/gEyxbTuj Don't miss out!
Sign up for our free webinar this Wednesday! 👉 https://lnkd.in/gEyxbTuj Learn how to build a standalone centralized guardrails service and get to ask your questions to our Head of Engineering 🙂
Our new Guardrails Server allows developers to create a standalone, centralized guardrails service that can protect all GenAI use cases across an organization. With parallelization, streaming and inference endpoints, the Guardrails Server is the fastest validation server. Join us for an interactive session on migrating and spinning up the new Guardrails Server! https://lnkd.in/gEyxbTuj
Founder @ Iudex | Former Staff SWE @ Scale | Former SWE @ GoogleX
4wThere's a lot of features to unpack but happy to chat with anyone who is interested in learning best practices