Is @AMD's MI300X GPU the best pick for LLM inference on a single GPU ❓
We compared Nvidia's H100 and AMD's MI300X GPU and found that MI300X can be a better fit thanks to its larger memory and higher memory bandwidth
Read the full benchmark: https://t.co/xMRhF6nORh
🎉 We’re excited to announce our partnership with @OVHcloud, Europe’s leading cloud provider.
Our shared goal is to accelerate ML development by combining:
☁️ Scalable and secure cloud environments
⚙️ Hybrid-cloud pipeline automation
Learn more at: https://t.co/skwhBquMIM
What’s worse than waiting for an expensive computation to perform ❓
Waiting for the exact same computation to perform for the second time…
We've built a new functionality so you don't need to wait and pay for redundant computations!
Learn more at: https://t.co/DQqHtqhVvm
📢 New feature alert… Make it two… Make it three!
We’ve just released multiple significant improvements to the Valohai MLOps platform, designed to help you further accelerate time-to-market and optimize costs.
Learn more: https://t.co/L9Hsms8C3z
🔎 Here’s what comes with Valohai’s Model Registry, a centralized hub for managing model lifecycle:
🔸 Model overview with threads
🔸 Model states for internal approvals
🔸 Model versions with complete lineage and reproducibility
And more: https://t.co/RLfjLStmxw
Hilla 2.3 introduced the new AutoGrid feature. See how it simplifies coding a robust grid with sorting, filtering, and lazy loading in just one line of code.
https://t.co/DIiydpw2D2
The @vaadin Flow team is planning on addressing a big pile of small (but important!) issues in their next 6-week project. Check out the list and let us know in the comments if something important to you is missing. https://t.co/L5uxmnLXDg
🚂Trains go choo-choo!
Hilla goes two-two!
@hillaframework 2.2 is out, adding form support. Define validation rules on your Java model, reuse them in the browser.
Full release notes: https://t.co/2kwNNRqhMf
We are planning to ramp up @hillaframework development significantly this fall 🚀
🚢 We're going to release A LOT more often so we can get more features and fixes in your hands faster
🫵We want to hear from you: What do you want to see in Hilla? How can we make it better for you?
🤼♂️ We have a planning session next week with the Hilla team and members from the DevRel, product, CS, and design teams to hash out, hack, and prioritize features to work on. I'll share more info after that!
We're excited to announce the new Vaadin AI Form Filler, an experimental feature for #VaadinFlow. 🎉 This new add-on can automatically fill in forms using unstructured text in various languages. Learn more: https://t.co/H8bdwMo3En
@flozano@simas_ch As a PM at Vaadin, I'm interested in why you perceive Spring as a safe option to depend on, but not Vaadin or JOOQ (though the latter two have extensive and free open-source capabilities)?