Twitter Beğeni Hilesi

Message Queues in System Design levelup.gitconnected.com/message-queues… < queues? That's super boring tech that's also in use more than anything AI-related. Always good to learn/refresh the fundamentals!

I'm loving the new landing page from Hugging Face for Google Cloud. Learn how to train and deploy models with their Deep Learning Containers. huggingface.co/docs/google-cl… Also, the Hugging Face GitHub repo for DLC on Google Cloud: github.com/huggingface/Go…

I just tried this. Very slick. Compare prompt results between Gemini 1.0, 1.5, and Gemma versions. Try it right now in Google AI Studio. ai.google.dev/aistudio

I don't think we have a cloud database that's missing vector support at this point. Now, Memorystore for Redis Cluster (cloud.google.com/memorystore/do…) and Memorystore for Valkey (cloud.google.com/memorystore/do…).

What's Ray? And why would you run Ray on Kubernetes? If you care about scaling Python or AI workloads, keep reading. Kaslin Fields has a good post that answers your questions. I learned something! kaslin.rocks/ray-on-k8s/

Seroter Daily Reading List – September 13, 2024 (#397): Today’s links look at how Google builds engineering teams, why Cursor is catching attention for AI coding assistance, and what a platform engineering-ready Kubernetes looks like. seroter.com/2024/09/13/dai…

Just had a special experience with Google DeepMind Gemini 1.5 Pro's long-context capabilities! 🤯 I extracted the audio from Alex Albert 76-minute video on prompting at Anthropic, turning the video into a 140k token input. Before watching, I thought, "Let's see what AI can do

How might you build an LLM-powered app using Go? Eli Bendersky offers up three solid approaches: direct APIs, LangChain for Go, and Genkit for Go ... go.dev/blog/llmpowered

In which Hope Scrollin says that "open source foundations are a good thing" and brings receipts. redmonk.com/jgovernor/2024… Relatedly, CNCF added more members, which is great: cncf.io/announcements/…

Compare multimodal LLM results? Last week, Logan Kilpatrick shared that Google AI Studio (ai.google.dev/aistudio) now had a "compare" feature. I just tried comparing Gemini 1.5 Pro and Flash results using an arch diagram and "best practices" PDFs. Cool to see side by side!

Did you know that the gaming industry was the target of 46% of DDoS attacks? I did not. Here's a story on how EA is protecting themselves with a unique managed service ... cloud.google.com/blog/products/…

We recently hit a milestone: in the last 30 days over one million Colab users accessed our ✨Gemini-powered Code AI features✨. Learn more from our team here! youtube.com/watch?v=V7RXyq…

A few hints? A couple of best practices? Nope, this post goes for broke with *fifty* tips for better Angular development. The gauntlet has been laid down. developer.mescius.com/blogs/the-top-…

Customer complaints are a GOOD thing? Be HAPPY when you're drowning in feature requests? Your competitive landscape is LIT UP and you should celebrate? This First Round piece offers non-obvious signs you're getting early traction: review.firstround.com/non-obvious-si…

We announced DataGemma last week, an open model with grounding in Data Commons. This engineering post goes deeper into "retrieval interleaved generation (RIG)" and how it works: research.google/blog/grounding…

✨📒 Calling the Gemini API in Colaboratory, in 25 lines of code or less: colab.sandbox.google.com/notebooks/snip…

I can't say I've seen a lot written about unit testing your data pipelines. This post from Svetak Sundhar about testing Apache Beam may give you some ideas ... beam.incubator.apache.org/blog/unit-test…

How Google Makes Custom Cloud Chips That Power Apple AI And Gemini youtube.com/watch?v=MucuOt… < TPUs are training our models, and Apple's too. Fun CNBC segment with a first look at our chip lab.

A few lines of code to create embeddings, update a corpus of data, and generate relevant answers from an LLM? This post shows off a managed Google Cloud experience for LlamaIndex that looks super approachable. medium.com/google-cloud/g…

Seroter Daily Reading List – September 16, 2024 (#398): Today’s links look at building LLM-powered apps in Go, why open source foundations are helpful, and how you can predict developer attrition. seroter.com/2024/09/16/dai…