About 200 t/s prompt processing and 10-20 t/s with MTP.
Greatly depends on the task, predictable things like code generates at 18-20 t/s. Creative writing more like 10-17 t/s.
About 200 t/s prompt processing and 10-20 t/s with MTP.
Greatly depends on the task, predictable things like code generates at 18-20 t/s. Creative writing more like 10-17 t/s.
Yes, I got a Strix Halo machine before the RAM price hike and use it to run all my ML stuff on it.
Currently using llama-swap with llama.cpp/ComfyUI and opencode/Open WebUI as frontend.
I’m running Qwen3.6-27b, Voxtral Mini 4b, Piper and Qwen Image. Also, some embedding and reranking models.
I use them for:
If you have trouble with outgoing mails, you can use a hybrid approach.
Receive mails directly to your server but use a mail service to relay your outgoing mails. Configuration for that is very simple in mailcow and there are a few dozen (free) transactional email providers (e.g. Scaleway).
That way you can keep receiving your mails privately and only have to give up some privacy when sending mails.


That looks pretty good. Looks like Portainer is getting replaced this weekend.


I share photos and videos through my Matrix/Signal bridge all the time just fine.
Calls don’t work but that’s about it.
Do you actually train the LLM or use RAG? I have been looking for a local LLM + Wikipedia RAG solution for a while now.
For now I just have kiwix-serve + searxng doing a simple search but the Kiwix search is…questionable.


I wrote an application which runs on my server and monitors my favorites on Tidal/Deezer/Qobuz. It downloads them in bulk whenever I have a premium account with one of them. Usually I purchase a month of premium every few months, at which point I get nice clean FLACs for local use.
The FLACs are moved to Jellyfin and I stream them using Finamp, which also supports transcoding, so I keep 128 kbps Opus files for offline playback and stream the raw FLAC files when bandwidth is no concern.
I have amassed a huge music library over the last decades, so even if all streaming websites go under tomorrow, I have enough music locally to last me a lifetime.


If we’re talking about online editing, Collabora has web editors based on LibreOffice but with a modern UI: https://www.collaboraonline.com/
They are really great and can be self hosted (e.g. with Nextcloud).
For offline editing, as already mentioned, LibreOffice has an optional ribbon UI and OnlyOffice looks pretty modern as well.


Is it a Surface laptop?


Haven’t had any Vertex explosions or shader compiling issues in Wilds but I also assume that’s Nvidia related.
Do you have those issues in their other titles like their newer Resident Evil games as well?


I had the same issue with PINCE not restoring the correct memory addresses on start.
Although I think I’m doing something wrong and the memory in modern games is just dynamic so the correct location can’t be found with just the memory addresses. Haven’t looked if it is possible yet but I assume you need some pattern matching to find the right address, not sure if PINCE can do that yet.


What’s the problem?
Played Wilds on launch and had pretty much no issues other than the game freezing for a second or two every hour or so.
On the other hand, my friend on Windows would crash from time to time, which I didn’t experience.
Although it should be noted that neither Wilds nor Dragon’s Dogma are technological marvels. They run bad everywhere.


Cheat Engine is a thing on Linux!
Game Conqueror is bundled for a lot of distros but PINCE is my favorite.
Really? I did a 365 day streak in Japanese (and most of that time was spent learning Hiragana and Katakana) but I still had some basic phrases down at the end.
Kanji broke me in the end though, Duolingo really doesn’t do it well.


Helldivers 2, and will try Arc Raider
Helldivers 2 and Arc Raiders both work fine on Linux. PUBG does not.
potentially Marathon next year
Marathon is unlikely to work since Destiny also doesn’t work.


C# by muhammad-sammy.
Doesn’t have the fancy project manager that the Microsoft one has but since I’m used to the dotnet CLI, I don’t mind that much.


Fedora Kinoite with VSCodium (Flatpak), both for work and my own stuff.
Also a few toolboxes with different compiler versions for some older projects.
I mostly do .NET and PHP stuff.
We need a fourth one for “User error”.


I don’t really want automatic updates, I want a notification once a month with all images that have a newer :latest available or if versionised, when a image with a newer version is available.
Q8 from unsloth.
My go to model for knowledge. Definitely much faster at Q5 but it lacks the tool calling quality of the Qwen3.6 models. Really hoping we see a Qwen3.6-122b soon…