buyucu

26 ms·

1. 571.▲
   by buyucu 2y ago

   they go together. once you have a walled garden, the temptation to moderate/censor it is too large. censorship was practically impossible in the old internet before social media.
2. 572.▲
   by buyucu 2y ago

   I'm willing to bet that Satya loves this. Microsoft's business model relies on AI getting cheaper and commoditized. Paying gazillions to OpenAI can hardly be fun for Microsoft.
3. 573.▲
   by buyucu 2y ago

   no someone forked an older janus model. pro is not released yet
4. 574.▲
   by buyucu 2y ago

   no someone forked an older janus model. pro is not released yet
5. 575.▲
   by buyucu 2y ago

   Americans are having a crisis of confidence as their technological lead is crumbling before their eyes. Expect more tantrums of this kind.
6. 576.▲
   by buyucu 2y ago

   I used more to fine-tune SDXL and it looked horrible.
7. 577.▲
   by buyucu 2y ago

   the evidence is that facebook is blocking this content.
8. 578.▲
   by buyucu 2y ago

   this is why content moderation is a really bad idea. the false positives are going to dominate any moderation you do.
9. 579.▲
   by buyucu 2y ago

   aren't tesla cars (in)famous for their very bad build quality?
10. 580.▲
    by buyucu 2y ago

    Banking is a heavily regulated sector. Regulation effectively blocks any kind of disruption and perpetuates the established firms.
11. 581.▲
    by buyucu 2y ago

    I doubt the journalist actually has 25 million to pay, even if the CEO wins the lawsuit.
12. 582.▲
    by buyucu 2y ago

    It would be such a shame if this went viral. Before today I did not know who Maury Blackman was. I now know that he was arrested for domestic violence in 2021. Here is the original news report: https://jackpoulson.substack.com&#
13. 583.▲
    by buyucu 2y ago

    they are distillations of r1, and work fairly well given the modest hardware they need.
14. 584.▲
    by buyucu 2y ago

    first, this is amazing! second, how does one increase the context window without requiring obscene amounts of RAM? we're really hitting the limitations of the transformer architecture's quadratic scaling...
15. 585.▲
    by buyucu 2y ago

    nice try cia bot
16. 586.▲
    by buyucu 2y ago

    Deepseek R1 now has almost 1M downloads in Ollama: https://ollama.com/library/deepseek-r1 That is a lot of people running their own models. OpenAI is probably is panic mode right now.
17. 587.▲
    by buyucu 2y ago

    what makes you think they can't do it? they had no trouble doing everything else. they'll be fine: https://www.msn.com/en-us/news/technology/huawei-smic-to-bui...
18. 588.▲
    by buyucu 2y ago

    meta just replaced its public policiy officer to pander to the new administration. american companies work hard to align with american government.
19. 589.▲
    by buyucu 2y ago

    I'm running the 32b version on a home server (Ryzen 7) using llama.cpp, which uses Vulkan for hardware acceleration on the iGPU. For hobbyist inference, getting a iGPU with lots of system ram is probably better than getting a dedicate
20. 590.▲
    by buyucu 2y ago

    Try asking ChatGPT about the genocide Israel is committing. Then you'll see what censorship looks like.
21. 591.▲
    by buyucu 2y ago

    I just tried asking ChatGPT how many civilians Israel murdered in Gaza. It didn't answer.
22. 592.▲
    by buyucu 2y ago

    I just asked ChatGPT how many civilians Israel killed in Gaza. It refused to answer.
23. 593.▲
    by buyucu 2y ago

    One of Meta's policy officials (Jordana Cutler) is a former Israeli government official who was censoring anti-genocide content online. American models are full of censorship. Just different stuff.
24. 594.▲
    by buyucu 2y ago

    Mastodon seems to be doing pretty good so far on scaling up. Pixelfed is smaller than Mastodon, so it will be fine.
25. 595.▲
    by buyucu 2y ago

    they seem to be doing fine so far. every day we wake up to more success stories from china's AI/semiconductory industry.
26. 596.▲
    by buyucu 2y ago

    I had no problems running the 32b at q4 quantization with 24GB of ram.
27. 597.▲
    by buyucu 2y ago

    isn't this what yt-dlp does?
28. 598.▲
    by buyucu 2y ago

    I'm impressed by not only how good deepseek r1 is, but also how good the smaller distillations are. qwen-based 7b distillation of deepseek r1 is a great model too. the 32b distillation just became the default model for my home server.
29. 599.▲
    by buyucu 2y ago

    I'm running the 7b distillation on my laptop this very moment. It's an insanely good model. You don't need reddit to judge how good a model is.
30. 600.▲
    by buyucu 2y ago

    Why would Deepseek lie? They are in China, American export laws can't touch them.