Waescher

The internet is dying

@CodeWeavers Will we get some benefits from macOS 26 an Metal 4 on monday?

New #CrossOver Preview released. Join in on the fun and create some delightful chaos. Read updates here: bit.ly/4mz0lbd Join the Preview Crew here: bit.ly/4laCNIX #crossoverpreview #codeweavers #macgaming #apple #gameonmac #mac

@DevExpress It has been 30 months without a new WinForms skin 🥲

WinForms — New WXI Skin #WinForms #Skins community.devexpress.com/blogs/winforms…

.NET 9 announcement features OllamaSharp 😊

@Technovangelist @pdevito3 @stevensanderson @ollama Oh and that .NET should be working perfectly fine on Mac if done right. I use .NET a lot on Mac and Windows. Just give the OllamaApiConsole (in the OllamaSharp repository) or my little llmaid a try 😊

@Technovangelist @pdevito3 @stevensanderson @ollama Really appreciate the mention, thanks a lot. My comment to the citation was that Microsoft’s Ollama implementation is a PoC, and they reached out if OllamaSharp could be the official unofficial implementation.

Introducing Microsoft.AI.Extensions - a library to streamline use of AI in .NET apps. Glad to see this go live today! The library is something I and a few smart people have been working on in recent months. devblogs.microsoft.com/dotnet/introdu…

@ollama @Microsoft OllamaSharp 4.0 now provides both: a full Ollama API implementation and a battle tested Ollama backend for Microsoft.Extensions.AI github.com/awaescher/Olla…

Preview of @Microsoft Extensions AI is here for .NET builders! The goal is to provide a standard implementation for caching, tool calling, telemetry, and other common tasks that work with any provider. - Unified API for consistent set of APIs for integrating AI services into .NET apps - Portable, not tied to a specific provider - Easy to experiment with different packages using the same abstractions - Componentization Currently it has support for: - Ollama ❤️ - OpenAI - Azure AI Inference 👇👇 devblogs.microsoft.com/dotnet/introdu… Thank you! cc. @stevensanderson @chanezon

@stevensanderson OllamaSharp 4.0 is now supporting Microsoft.Extensions.AI 🎉 github.com/awaescher/Olla…

@pdevito3 @Technovangelist @stevensanderson @ollama @Microsoft It truly is. Knowing .NET since the early 2000s but never been more in love with it since the reboot.

@Technovangelist @stevensanderson @ollama @Microsoft .NET is a completely different animal In the best possible way since then! Night and day in approachability and complexity. It’s such a delight these days. They’ve got some examples here too: github.com/dotnet/ai-samp…

@pdevito3 @stevensanderson @ollama @Technovangelist In case you're interested, there's a full implementation of the Ollama API in .NET which now provides a battle tested backend for Microsoft.Extension.AI, too. github.com/awaescher/Olla…

@stevensanderson Loving the native @ollama support 🚀 Cc @Technovangelist — maybe some new content for ya

500 GitHub stars for OllamaSharp 🤩

@brave Sadly, there's a significant difference between Brave hosted and self hosted models in terms of answer quality. (same model llama3.1:8b) Can we get this fixed?

Leo lets you summarize pages and videos, create new content, ask questions and more right in the browser. You can access Leo by pressing Ctrl+B (Cmd+B on macOS) or clicking the “Leo AI” icon in the toolbar. Or type a question into the URL bar and click "Ask Leo."

AI in Brave is now more private and customizable with BYOM (Bring Your Own Model)! Starting today with v1.69, Brave users on desktop can use their own AI models through the browser's built-in assistant Leo.

@neuecc Codename should have been nsync

async2 #とは async2 - The .NET Runtime Async experiment concludes steven-giesel.com/blogPost/59752…

⭐️ 1k for RepoZ 🤩 github.com/awaescher/RepoZ

@ErikVoorhees @AskVenice Well, Qwen2:72b and R+ were the top LLMs on almost every benchmark. Generally, the stronger the models the more powerful Venice is as a replacement for closed AI. A little outdated already with yesterdays release of llama 3.1 70b 😅 I am super happy with these!

@Waescher @AskVenice Tell us more... why do you want those?

Venice is live Try it: Venice.ai

OllamaSharp ships a full featured console demo to use @OLLAMA locally *and* on a remote machine. It supports model management, chats and even image chat with multimodal models right from the terminal.