On another episode of Keeping up with the Language Models...

Connor looks into Gemini Pro 1.5 and Gemini Flash

Gemini Pro 1.5 and Gemini Flash!! ✨🔥

I am SUPER excited about long context LLMs! I put together a quick video walking through a Lost in the Middle / Needle in the Haystack test, using Gemini for Re-Ranking, and something I think is really exciting and gamechanging ... 🥁 ...

I think the most exciting announcement from Google I/O is Gemini Pro 1.5's 2 million token input window! 📚

Amongst many things, long input LLMs are going to be a game changer for search!

Search is typically a two stage process between retrieval and ranking. Beginning with

That’s a wrap! This week, our team attended the Gartner Application Innovation & Business Solutions Summit in Las Vegas. Bob van Luijt and Arjun gave a talk about how Cisco delivers Generative AI applications using Weaviate, and we had plenty of interesting conversations with

|￣￣￣￣￣￣￣￣|
Join the webinar
|＿＿＿＿＿＿＿＿|
\ (•◡•) /
\ /
——
| |
|_ |_

Join Job Nijenhuis from Neople and Erika Cardenas from Weaviate for a webinar on May 29th at 1 pm ET / 7 pm CET

The key takeaways:
• How vector databases work and how to use vector search for quick retrieval
• All about ingesting and chunking your data (feedback documents, FAQ

The future of AI is here, and we're building it. Join our team!

Apply today: weaviate.io/company/careers

Open roles:
• Partner Manager - System Integrator: careers.weaviate.io/jobs/4467330-p…
• Partner Solution Engineer: careers.weaviate.io/jobs/4467290-p…
• Compliance & Security Expert:

Team Weaviate • vector database will be in NYC for our AI [in Prod] event on May 21st 🥳

What you can expect:
• Lessons for running Weaviate in production
• Demos from Google Cloud, FactSet and Weaviate
• Hands-on workshop
• Panel discussion
• Food
• Excellent vibes (🤌)

Register here:

What is Vector Quantization, and how does it work?

Learn all about it from Etienne Dilocker, CTO and Co-Founder of Weaviate, as he gave a guest lecture at Brown University's CSCI, diving into the various Vector Quantization Techniques.. He

Watch the full lecture on YouTube:

API achieved internally

Gemma 2 is coming next month, and PaliGemma (Vision) is available today, open models for the win!

Open source RAG? Ollama integrations? NEW VERBA RELEASE?? 🤩

Run a state-of-the-art RAG application locally on your computer with Verba’s new ollama integration and use all the fantastic open source models like llama3, Mistral AI's 7B, or Command R from cohere ✨

The evolution of open-source versus hosted models on GCP

Dai Vu from Google Cloud discusses how they rely on customer choice for GCP, whether it is a 1st party model or a 3rd party open-source model.

In the full podcast, @cshorten30, Bob van Luijt, and Dai Vu discuss open-source

Open Source RAG with Ollama! 🥳

YouTube: youtube.com/watch?v=swKKRd…

The newest Verba v1.0.0 update is out, and it’s...

✨ B-E-A-U-T-I-F-U-L ✨ 

We added many exciting things to Verba, including open-source models powered by Ollama, Frontend Customization, and many Quality of

Here's a summary of my personal favorites at #GoogleIO , in case you missed it.

1. Agent World: True personal assistants are coming. Project Astra ✨ is super powerful as the name 'Astra' suggests.
Amazing answers to complex prompts involving instruction following and

Finally added a React frontend to the RAG application. Still using DSPy, FastAPI, and Weaviate in the backend. It is amazing to develop a full-stack web application powered by DSPy.🔥🔥🔥🔥🔥

Link to module in second tweet !

Routing between multiple databases for a RAG application, very cool and useful. Thank you for sharing this 🙏

Can’t wait to see you all in NYC!

Great example of how distributed data can be seamlessly integrated for robust RAG applications!
Leveraging BigQuery and Weaviate with DSPy is a game-changer for efficient data querying and context fusion.
This approach optimizes the use of diverse datasets, making it possible

For people based in The Netherlands 🇳🇱 Check this out: