redrove

6 ms·

1. 1.▲
   by redrove 29d ago

   I'm sorry but that's like saying communism is 'user error'. If the ideal scenario never happens in practice, it's a flawed concept to begin with.
2. 2.▲
   by redrove 29d ago

   YC gives DO credits for startups.
3. 3.▲
   by redrove 29d ago

   Could’ve just used fewer tokens and redirected to the Codex signup page. ba dum tsss (sorry couldn’t help myself)
4. 4.▲
   by redrove 1mo ago

   I’m not so sure I agree, given the overwhelming concentration of capital and regulatory capture they have, I just don’t see them going anywhere; becoming more niche rather than even more of a standard is “going away” to a certain extent as
5. 5.▲
   by redrove 1mo ago

   Thanks for the very detailed answer! Would you say ChatGPT does everything you need today or do you still see some gaps? i.e. things you think it should be able to do but currently doesn’t, or things that still take too much effort on your
6. 6.▲
   by redrove 1mo ago

   I would also be interested in what you use AI for. Is it marketing? bureaucracy?
7. 7.▲
   by redrove 1mo ago

   I don’t disagree but saying “cloud will change in the next 8-20y-ish” is a bit of a non-argument, you’re not really stating any thesis to speak of; Change is a given over that time frame.
8. 8.▲
   by redrove 1mo ago

   I think you're supposed to use AI to digest it
9. 9.▲
   by redrove 1mo ago

   So..pseudo code?
10. 10.▲
    by redrove 1mo ago

    Oh this looks very interesting, both for personal and work; I’ve been looking for something similar for quite a while. However, no OpenAI API support (just Anthropic + openai.com) means I can’t use it for either.
11. 11.▲
    by redrove 1mo ago

    I’ve tried using this as a self hosted instance and it’s been a little rough around the edges with Hermes.
12. 12.▲
    by redrove 1mo ago

    Could it just be dosage? Commonly found dosage is 10-20x what you'd actually secrete and IIRC absorption is actually quite good.
13. 13.▲
    by redrove 1mo ago

    Is there a screenshot and a feature matrix of some sort? I couldn’t find on the docs site and I don’t understand what this tool is capable of, even though I’m familiar with SIEM as a general concept.
14. 14.▲
    by redrove 1mo ago

    Exactly, sounds to me like OP is the one with the LLM skill issue.
15. 15.▲
    by redrove 1mo ago

    No worries! Would love to run K3 but lack the hardware for a 2.8T model, like most people.
16. 16.▲
    by redrove 1mo ago

    I’d be happy to test this out if you want to publish an alpha branch or something.
17. 17.▲
    by redrove 2mo ago

    The FP8 version from DeepSeek themselves [0], around 1800 tps prefill and 45 tokens per second decode. I’ve been running a custom VLLM image with b12x as well as nvfp4_ds_mla. I would say it’s quite fantastic in day to day, I use it mostly
18. 18.▲
    by redrove 2mo ago

    Been running this on a few Asus GX10 machines with k3s on top, it’s been great. I’m running the new deepseek. Thank you for your work!
19. 19.▲
    by redrove 3mo ago

    No, it doesn’t seem like it.
20. 20.▲
    by redrove 3mo ago

    I’m sorry but I can’t take a European Commission link seriously about training SOTA LLMs.
21. 21.▲
    by redrove 3mo ago

    Do you have any sources for this? Genuinely interested in reading more.
22. 22.▲
    by redrove 3mo ago

    No it’s not. You’re generalizing, DACH != the entire EU.
23. 23.▲
    by redrove 3mo ago

    > My dream would be a local model that can do, say, 80% of the day to day tasks I need; "how does X Handler connect to Y storage?", "commit that feature, but leave out the bits that relate to billing" etc. Qwen 3.6 27
24. 24.▲
    by redrove 3mo ago

    They’re still debating that, they’ll get back to us soon I’m sure.
25. 25.▲
    by redrove 3mo ago

    I read the entire thing and it seems to me like they started with the conclusion and tried to find proof, like a lot of psychology papers. Thoroughly unreliable.
26. 26.▲
    by redrove 4mo ago

    It’s about bang for buck. That high a score for 5B params is pretty good, nigh unbelievable a short while ago. It is my belief that smaller models will get better and better, and even cloud SOTA models will shrink. Yet another reason the cu
27. 27.▲
    by redrove 4mo ago

    Wow, if you you could've just done docker compose up. I guess we'll never know.
28. 28.▲
    by redrove 4mo ago

    So the theory goes. Can you point out one place where this is working out?
29. 29.▲
    by redrove 4mo ago

    Don’t vibe code your auth path folks.
30. 30.▲
    by redrove 4mo ago

    I’m genuinely interested in what you find to be the problem with that? It’s a landing page.