ranguna

30 ms·

1. 541.▲
   by ranguna 3y ago

   Looks pretty neat!
2. 542.▲
   by ranguna 3y ago

   I've been using librewolf on desktop and mull on android for around a year now and it has been great! Occasionally I have to switch to Firefox because some websites hard code agents and flag librewolf as not being Firefox, but that&#x2
3. 543.▲
   by ranguna 3y ago

   So you don't use a browser?
4. 544.▲
   by ranguna 3y ago

   Interesting benchmark! Would mosaic mpt be useful to benchmark as well?
5. 545.▲
   by ranguna 3y ago

   Can't wait to see the end result of all of this.
6. 546.▲
   by ranguna 3y ago

   Will chatgpt even know? This is past 2021.
7. 547.▲
   by ranguna 3y ago

   I know people that follow the Mediterranean diet who are perfectly healthy and those who are not. I know people who follow the vegan diet who are perfectly healthy and those who are not. It seems studies around diets like to spin the wheel
8. 548.▲
   by ranguna 3y ago

   Uncouple Russian modules and burn them ?
9. 549.▲
   by ranguna 3y ago

   When this all started, I had the same argument of hindering technical progress. But then I realised that all companies have to do is support usb c, it's fair game if they want to add another plug alongside it.
10. 550.▲
    by ranguna 3y ago

    Ah that makes more sense.
11. 551.▲
    by ranguna 3y ago

    https://news.ycombinator.com/newsguidelines.html > Please don't do things to make titles stand out, like using uppercase or exclamation points, or saying how great an article is. It's implicit in submitting som
12. 552.▲
    by ranguna 3y ago

    Is it just me, or is this comment AI generated? > and i was shocked. At first I though (...), but its actually (...)! > The (...) installed everywhere (...) each other nonstop, sometimes with real > There's social pyramid star
13. 553.▲
    by ranguna 4y ago

    I'm on a stack that also uses lambdas an event bridge, I think the OP means that the stresses originating from k8s and Kafka are gone when using lambdas and event bridge.
14. 554.▲
    by ranguna 4y ago

    The tuned 7B model is around 33 GBs, so you'll need a PC with that much VRAM or RAM. I haven't tried to load it on text generation ui though.
15. 555.▲
    by ranguna 4y ago

    Sorry, I should have added /s in the end there
16. 556.▲
    by ranguna 4y ago

    Ah yes, it fails to count letters but successfully explains quantum mechanics. It's useless.
17. 557.▲
    by ranguna 4y ago

    I pay 0 for either.
18. 558.▲
    by ranguna 4y ago

    I don't get where your question is coming from, you can already run LLMs on a single machine. Checkout llama.cpp, tabby, text generation webui, gpt4all, AI Dungeon open source models like clover-edition, and know this we gpu based app.
19. 559.▲
    by ranguna 4y ago

    This might be good for simple questions or complex ones that have been asked 1000s times online, but it starts to fail a bit for nich stuff: > 403 error when updating firebase runtime config using service account This gave me an answer w
20. 560.▲
    by ranguna 4y ago

    Really? Just that quote alone seemed pretty clear to me, and it becomes even clearer as you read the rest of the prompt.
21. 561.▲
    by ranguna 4y ago

    Does invoicing actually take more time per month than writing? I'm not saying invoicing is easier than writing, what matters here is the human time it takes. If the time you spend writing is reduced by 90%, then you'll have more t
22. 562.▲
    by ranguna 4y ago

    Are these like javascript promises or goroutines?
23. 563.▲
    by ranguna 4y ago

    We won't know until it happens, I for one think it's gonna get much better. All that needs to happen is openai model leaking and then it'll get into highway speed once it's in the hands of the people. Take a look at llam
24. 564.▲
    by ranguna 4y ago

    I think the ruling was that if it's transformative enough, you can copyright it. So as long as you put your own twist 8n it, maybe you can own the work. But I'm not a lawyer, so don't quote me on that. As for the overhead, ma
25. 565.▲
    by ranguna 4y ago

    But you won't do it alone, you'll have the most advanced AI next to you.
26. 566.▲
    by ranguna 4y ago

    It's 1000s of times better comprared to what we had a year ago. Imagine what we'll have in 3,5,10 years.
27. 567.▲
    by ranguna 4y ago

    So this person lost his job to an AI because it's way cheaper and slightly worse. Why not hire chatgpt yourself? Instead of resigning because of chatgpt, use it for your advantage. Instead of taking 10 clients, take 100 at 10x less you
28. 568.▲
    by ranguna 4y ago

    Tabby infers on the gpu and is slow, I can only imagine how slow truepilot is on the cpu. If people want auto complete, it needs to be super fast. For slow inference, a better application would be a chatbot that reads your code and answers
29. 569.▲
    by ranguna 4y ago

    You are comparing apples with chairs and the lack of examples as to why you don't like it doesn't make your comment useful.
30. 570.▲
    by ranguna 4y ago

    Ah so the OP lied, great.