sorted by: new top controversial old
NUC recommendations please in c/selfhosted@lemmy.world
An update on my FOSS downloader extension recommended by Mozilla! in c/opensource@lemmy.ml
TypeType: a self-hosted, open-source video app for YouTube, NicoNico and BiliBili (1.9.0 is out!!) in c/opensource@lemmy.ml
TypeType: a self-hosted, open-source video app for YouTube, NicoNico and BiliBili (1.9.0 is out!!) in c/opensource@lemmy.ml
Looking for a durable and repairable Linux laptop for business use in c/linux@lemmy.ml
[–] 10 points 1 month ago (1 child)

Las Lenovo Thinkpad son lo mas recomendado con razón, pero suelen ser bastantes caras aquí en Latinoamerica. Una buena opción más económica son las Lenovo Thinkbook, tengo una con Debían instalado y funciona todo.

  • permalink
  • fedilink
  • source
    •
    llama-server ignoring webui flags in c/localllama@sh.itjust.works
    [–] [S] 1 point 1 month ago

    Solved. I was using a config file generated using the Export Settings button, but llama-server --webui-config-fle doesn’t understand it because it has extra params, only the config entry of the saved json file should be passed to llama-server webui flag. No errors show up in the browser’s console, qwen pointed me out the conflict.

  • permalink
  • fedilink
  • source
  • parent
  • context
    • llama-server ignoring webui flags in c/localllama@sh.itjust.works
    [–] [S] 1 point 1 month ago

    Thanks! yeah, the configuration has passed through and I can see the correct settings there, but they don't show up in the Settings and I also checked that they are no "loaded", e.g. the model ignores the system prompt of the config file, something that doesn't happen when I load the config file manually.

  • permalink
  • fedilink
  • source
  • parent
  • context
    • what your favorite custom rom in c/degoogle@lemmy.ml
    [–] 7 points 1 month ago

    LineageOS, good enough to degoogle and if it's possible for you to use F-Droid-only apps you are out of the surveillance machinery (with the exception of the TSP if you use a SIM card) and it's a huge step from stock Android in terms of privacy.

  • permalink
  • fedilink
  • source
    •
    6
    llama-server ignoring webui flags (lemmy.ml)
     

    It's been many weeks since --webui-config and --webui-config-file don't work for me. I was expecting that they were bugs that would be quickly fixed, but here I am... I use Chromium in Linux, and also tested it in Librefox. I don't have a github account and don't want to make it just to report this issue, but I doubt it's a bug anyway, someone else would have already reported it. Any clue about what could be happening?

    Command examples in case you want to test it:

    llama-server --models-preset ~/models/presets.ini --webui-config '{"systemMessage": "You are a helpful assistant.", "showSystemMessage": false, "showMessageStats":"true"}' --model ~/models/Qwen3.6-35B-A3B/Qwen3.6-35B-A3B-UD-MTP-Q4_K_XL.gguf --temp 0.6 --min-p 0 --top-k 20 --top-p 0.95 --mcp-servers-config ~/config/llama-mcp-servers.json

    or

    llama-server --models-preset ~/models/presets.ini --webui-config-file ~/config/llama_settings_2026-07-28.json --model ~/models/Qwen3.6-35B-A3B/Qwen3.6-35B-A3B-UD-MTP-Q4_K_XL.gguf --temp 0.6 --min-p 0 --top-k 20 --top-p 0.95 --mcp-servers-config ~/config/llama-mcp-servers.json

    The webui config json file is loaded in this last command, I know it because if I make a wrong json file it throws an error, but the configurations don't show up in Settings window an the are completely ignored, I have to load it manually using the Import/Export tab in the Settings.

    Your Computer Might Demand Your Age Soon, No Matter Where You Live in c/privacy@lemmy.ml
    Privacy-focused mail providers (FOSS preferred)? in c/opensource@lemmy.ml
    A Severe Misalignment of AI in Mathematics in c/fuck_ai@lemmy.world
    [–] 0 points 1 month ago

    You shouldn't use an LLM to do your homework but you can certainly use it to learn and clarify things you don't understand. I teach math and physics and tested local LLMs for learning and I can say they really help, at least in these subjects. The best option is to learn from books and making your homework, but when you're stuck with something that you don't understand, ask a good LLM to clarify things, give examples and context. I tested myself while learning about differential geometry and relativity, it really speeds up the learning process and understanding. But don't use them as a primary source, nothing beats a good book and practice on that. I wish there was something like that when I was in college, I wouldn't have carried some misunderstandings so many years. Being just a tool it's up to you if you use it responsibly or not, as any other tool.

  • permalink
  • fedilink
  • source
  • parent
  • context
    • Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses in c/localllama@sh.itjust.works
    13
    Swapping storage between an Intel laptop and an AMD mini PC (lemmy.ml)
     

    I have a laptop with an Intel i5-1335 CPU and I'm about to receive a mini PC with a Ryzen 8845HS, which is going to be my main computer now. If I just install the SSD M.2 of the laptop on the mini PC, is there any software I need to install that was not installed when first installed Linux on the SSD while being in the laptop? Or something that I need to change in the configuration concerning the new architecture? is it OK to do that? In other words, can Linux deal with the change without any issue or misconfiguration? Just trying to see if I can avoid the work of installing Linux from zero and all the software that I already have on the laptop. I'm using Debian Trixie.

    1
    Cmake ignoring option and compiling with just one thread (lemmy.ml)
    submitted 2 years ago by to c/cpp@lemmy.ml
    comment fedilink
     

    Using the -j option is ignored. About two weeks ago I found out that cmake was compiling with only one thread but everything was ok a couple of months ago. I'm using Debian Trixie which is still in testing so I thought it was a bug, but after many system updates I didn't saw cmake being updated. I couldn't find anything after a some searches. Another related issue is that I recently compiled a new version of a qt app with an edit box to set the number of threads for running LLMs, after the last compilation it doesn't allow to set more than one. Any clue?

    view more: next ›