dvrp

13 ms·

1. 181.▲
   by dvrp 1y ago

   See also: https://news.ycombinator.com/item?id=20849148
2. 182.▲
   I find journaling indispensable (2019)(jon.bo)

   1 points by dvrp 1y ago|1 comments
3. 183.▲
   Tell HN: Hacking AIs is similar to hacking humans

   1 points by dvrp 1y ago|0 comments
4. 184.▲
   by dvrp 1y ago

   Hi there! Thank you for the glowing review! I'm the cofounder of Krea and I'm glad you liked Sangwu's blog post. The team is reading it. You'll probably get a lot of replies around how this model is a just a fine-tune an
5. 185.▲
   by dvrp 1y ago

   It is not just a fine-tune.
6. 186.▲
   by dvrp 1y ago

   We've noticed that Wan 2.2 (available on Krea) + Krea 1 refinement yields _beautiful_ results. Check this from our designer, for instance: https://x.com/TitusTeatus/status/1952645026636554446 (Disclaimer: I a
7. 187.▲
   by dvrp 1y ago

   Woah, I didn't know about Packed SIMD/SIMD within a register (SWAR).
8. 188.▲
   by dvrp 1y ago

   i know you said this is an experiment but are there goals behind the project or is it more a “let’s see what happens” ??
9. 189.▲
   by dvrp 1y ago

   krea.ai | Distributed Systems Eng. & Data Engineers | San Francisco (on-site), CA | $200K-$400K I'm the cofounder of Krea ( https://www.krea.ai ), a startup in San Francisco building browser-based AI creative tools to imp
10. 190.▲
    by dvrp 1y ago

    nit: was it Sequoia or Tiger Global who led your last round? On your Series C announcement it says it was Tiger Global.
11. 191.▲
    by dvrp 1y ago

    Can you point to a URL with the tests you’ve done? Also, FWIW, this model focus was around aesthetics rather than strict prompt adherence. Not to excuse the bad samples, but to emphasize what was one of the research goals. It’s a thorny t
12. 192.▲
    by dvrp 1y ago

    Haha-classic! It’s simple: hackability and recruiting! The open-source community hacking around it and playing with it PLUS talented engineers who may be interested in working with us already makes this release worth it. A single talented d
13. 193.▲
    by dvrp 1y ago

    You can find details about the license here: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/mai... In a nutshell, it follows the same license as BFL Flux-dev model.
14. 194.▲
    by dvrp 1y ago

    Copying and pasting Sangwu’s answer: We used two types of datasets for post-training. Supervised finetuning data and preference data used for RLHF stage. You can actually use less than < 1M samples to significantly boost the aesthetics.
15. 195.▲
    by dvrp 1y ago

    Yeah, there are still imperfections. But, it’s surprising to us how much the quality can be improved without the need of a whole pre-training (re-) run.
16. 196.▲
    by dvrp 1y ago

    Haha thanks! Do you happen to have a use-case for it?
17. 197.▲
    by dvrp 1y ago

    Yup... We have the following: https://www.krea.ai/pricing Though we wanted to keep this technical blogpost free from marketing fluff, but maybe we over-did it. However, sometimes it's hard to give an exact price per im
18. 198.▲
    by dvrp 1y ago

    We actually have a GitHub repository to help with inference code. Check this out: https://github.com/krea-ai/flux-krea Let me see if we can add more details on the blog post and thanks for the flag!
19. 199.▲
    by dvrp 1y ago

    Relevant links: - GitHub repository: https://github.com/krea-ai/flux-krea - Model Technical Report: https://www.krea.ai/blog/flux-krea-open-source-release - Huggingface model card: https:/&#
20. 200.▲
    by dvrp 1y ago

    Hey thanks! I’ll ask Sangwu to hop here to answer this and give a more research-oriented answer
21. 201.▲
    by dvrp 1y ago

    Thanks! From a business point of view, there are many use-cases. Here's a list in no particular order: - You can quickly generate assets that can be used _alongside_ more traditional tools such as Adobe Photoshop, After Effects, or May
22. 202.▲
    by dvrp 1y ago

    OMG. Thank you! I had to setup a CDN-level redirect and I was so confused as to why when I asked others to help, their submissions were flagged as [dupe] or [dead] Thank you so much! I knew that HN software was advanced, but I didn’t know y
23. 203.▲
    by dvrp 1y ago

    thx evan!
24. 204.▲
    by dvrp 1y ago

    The engineering heart in me is happy in the sense that these people have clearly way more leverage than NBA or professional soccer/football players and should be compensated as such. But it's definitely a change.
25. 205.▲
    by dvrp 1y ago

    Relevant links: - Model Technical Report: https://www.krea.ai/blog/flux-krea-open-source-release - Huggingface model card: https://huggingface.co/black-forest-labs/FLUX.1-Krea-dev - Black Forest L
26. 206.▲
    FLUX.1 Krea: post-trained text-to-image model from Black Forest Labs and Krea(github.com)

    9 points by dvrp 1y ago|6 comments
27. 207.▲
    by dvrp 1y ago

    Related: https://news.ycombinator.com/item?id=44745555
28. 208.▲
    by dvrp 1y ago

    For Dang or HN-mods: I noticed that the URL for this submission is wrong: I tried to submit the correct URL ( https://www.krea.ai/blog/flux-krea-open-source-release ) but, for some reason, the submission gets flagged as
29. 209.▲
    by dvrp 1y ago

    Hello everyone. I’m the Co-founder and CTO of Krea. We’re excited because we wanted to release the weights for our model and share it with the HN community for a long time. My team and I will try to be online and try to answer any questions
30. 210.▲
    by dvrp 1y ago

    I think about it from an entropy POV: how much signal is the code/text transmitting? You can tell it’s AI for a surprisingly high amount of LLM outputs. If you feel it’s regurgitating what you or your team already know, it’s slop. This