gsuuon

20 ms·

1. 271.▲
   by gsuuon 3y ago

   This site has been the best programming education site I've encountered in terms of real, pragmatic concepts taught. My favorite is the concept of making invalid states unrepresentable[1] which I try to apply now regardless of language
2. 272.▲
   by gsuuon 3y ago

   How would the licensing for something like that work?
3. 273.▲
   by gsuuon 3y ago

   I'm not sure if I'd consider it the worst, but github ui (with its own issues) definitely feels way more friendly than gitlab
4. 274.▲
   by gsuuon 3y ago

   > you'll either be the only one doing what you're doing or among just a few people Agreed.. I've definitely Googled an issue I got stuck on only to land back on my own post (lol). You need fewer people for the same amount
5. 275.▲
   by gsuuon 3y ago

   This looks neat! Have you thought of putting a demo video on the frontpage?
6. 276.▲
   by gsuuon 3y ago

   I've wanted exactly this before - would be great to easily tell a story in the code review. Or an easier way to just see each commit in isolation (with a previous/next button).
7. 277.▲
   by gsuuon 3y ago

   I had it starred! It looks like a nice set of tools for building a multi-modal AI app - I'll give it a try when I flesh out a discord AI gaming app I was working on. Is the multi-modal aspect (image, audio, language) the main focus? Ma
8. 278.▲
   by gsuuon 3y ago

   This is a neat idea! I had a similar idea but with github repos instead of gists - it'd be possible to use tags to just check out blogs you're interested in. I built mine as a neovim plugin rather than a standalone site though. Gi
9. 279.▲
   by gsuuon 3y ago

   I think a single take-away from this is to generally overcommunicate rather than under, especially around the social context stuff (e.g. expectations from either side).
10. 280.▲
    by gsuuon 3y ago

    I used to do this as well, maybe a bit _too_ much. After reading an article somewhere (I forget which) that had just an absolutely excessive amount of exclamations I've toned it back down a lot. But for the folks who don't ever us
11. 281.▲
    by gsuuon 3y ago

    Why do you say that?
12. 282.▲
    by gsuuon 3y ago

    This is hilarious!
13. 283.▲
    by gsuuon 3y ago

    I really wish filtering could optionally include matching processes' children
14. 284.▲
    by gsuuon 3y ago

    This sounds like IQ testing which I thought was illegal (in the US)?
15. 285.▲
    by gsuuon 3y ago

    This is neat! I'd say there could be some improvement on selecting which dice to roll, but I'm not sure what. It feels clunky to adjust the dice with the drawer and -+ buttons. I wonder if this works on mobile - have you considere
16. 286.▲
    by gsuuon 3y ago

    I plug it directly into my editor (via https://github.com/gsuuon/llm.nvim ) and have it fill out code for me. I write what I want with comments and ask it to fill the rest - if it's straightforward enough it basica
17. 287.▲
    by gsuuon 3y ago

    Which people and hashtags? Trying to check it out but struggling to find relevant content. Is there a tech community somewhere? The ones I found appeared to be dead.
18. 288.▲
    by gsuuon 3y ago

    Yup, here's their web stable diffusion repo: https://github.com/mlc-ai/web-stable-diffusion The input is a model (weights + runtime lib) compiled via the mlc-llm project: https://mlc.ai/mlc-llm&#x2
19. 289.▲
    by gsuuon 3y ago

    Does syntax highlighting or lsp stuff work for things like v-for?
20. 290.▲
    by gsuuon 3y ago

    According to this [1] article (current top of hn) memory bandwidth is typically the limiting factor, so as long as your batch size isn't huge you probably aren't losing too much in performance. https://finbarr.ca/h
21. 291.▲
    by gsuuon 3y ago

    I'm really curious to see where that investigation leads - have you done any comparisons between Llama 2 and the embedding focused models? I wonder if it'll be better able to provide more 'intuitively correct' similariti
22. 292.▲
    by gsuuon 3y ago

    Even the 7B model is shockingly good! I've been hacking on a project also built on MLC (but the web runtime) and the completions I'm seeing from Llama 2 7B, just running on my laptop's browser, have been really impressive. Th
23. 293.▲
    by gsuuon 3y ago

    Generating an FSM over the vocabulary is a really interesting approach to guided sampling! I'm hacking on a structured inference library ( https://github.com/gsuuon/ad-llama ) - I also tried to add a vocab preproces
24. 294.▲
    by gsuuon 3y ago

    I've always thought it was interesting that in English there are lots of different words which mean the same thing but a different scale - like pebble / rock / boulder. It could easily just be little rock, rock, big rock. In
25. 295.▲
    by gsuuon 3y ago

    > I argue from data that the Data Scientist role is poorly differentiated and I speculate that its responsibilities are being eroded by better specified roles such as ML Engineer and Data Scientist. I'm already struggling to parse t
26. 296.▲
    by gsuuon 3y ago

    You can probably do this with the typescript compiler API https://github.com/microsoft/TypeScript/wiki/Using-the-Compi...
27. 297.▲
    by gsuuon 3y ago

    I imagine a photon as being like a glider in the game of life, except the relevant rules are electromagnetism. https://conwaylife.com/wiki/Glider
28. 298.▲
    by gsuuon 3y ago

    I wonder if you could train a model generally across a lot of languages, then specialize for a specific one with a different tokenizer / limited vocabulary? Here's the reference I've been using for llama 2 tokens: https:&#x2
29. 299.▲
    by gsuuon 3y ago

    Congrats Junru! I'm not on AMD but love seeing progress in this project. Excited for batched inference -- I didn't think it'd be useful for me but I've realized batched inference is also useful for a single user / e
30. 300.▲
    by gsuuon 3y ago

    FWIW I did get the CUDA backend running via WSL2