Comment on DeepSeek's distilled new R1 AI model can run on a single GPU | TechCrunch

fogetaboutit@programming.dev ⁨4⁩ ⁨weeks⁩ ago

• T156@lemmy.world ⁨4⁩ ⁨weeks⁩ ago

  The censorship only exists on the version they host, which is fair enough. If they’re running it themselves in China, they can’t just break the law.

  If you run it yourself, the censorship isn’t there.

  source

  • jaschen@lemm.ee ⁨4⁩ ⁨weeks⁩ ago

    Untrue, I downloaded the vanilla version and it’s hardcoded in.

    source
  • MonkderVierte@lemmy.ml ⁨4⁩ ⁨weeks⁩ ago

    Yeah, i think the censoring in the LLM data itself would be pretty vulnerable.

    source
• Mwa@lemm.ee ⁨4⁩ ⁨weeks⁩ ago

  You can self host it right??

  source

  • jaschen@lemm.ee ⁨4⁩ ⁨weeks⁩ ago

    The self hosted model has hard coded censored content.

    source
  • fogetaboutit@programming.dev ⁨4⁩ ⁨weeks⁩ ago

    if the model is censored… then what, retraining it? Or doing it from scratch like what open-r1 is doing?

    source