Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
maalber
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
31.
▲
by
maalber
2y ago
It is actually a really cool idea! Why i tried it out however it is not always clear to me which parts are the 'game interface' and which parts are your 'tooltips'. Not sure if this is intentional though I was also just
32.
▲
by
maalber
2y ago
Does this belong in Show HN?
33.
▲
by
maalber
2y ago
Can you elaborate on some of the technical details and motivation behind this service? Looks interesting, but I am unsure what you gain from it
34.
▲
by
maalber
2y ago
Very fun idea! Got way too invested and spend more time than I wanted to, but go there in the end:D I use MongoDB somewhat regularly, but I mostly have to ask GPT for help regarding the syntax. A couple of suggestions: - Gamify it a bit, a
35.
▲
by
maalber
2y ago
I love how that one post started a whole trend :D
36.
▲
by
maalber
2y ago
Thank you for elaborating, and interesting insight regarding the aspect ratio! I would be interested to know a bit more about the manual review process, e.g. how the invalid masks were classified and how many hours were dedicated to this. A
37.
▲
by
maalber
2y ago
This looks interesting. Quick comment on the format of the dataset though: I would suggest placing both the image and the segmentation mask as the same row with meaningful column names. I think this would make it easier to use. I think it c
38.
▲
by
maalber
2y ago
Cool! I appreciate that you made it openly available without signup! If I understand correctly, it first generates a bitmap using a text-2-image model and then converts that to svg? It might be cool to see the image alongside the svg and th
39.
▲
by
maalber
2y ago
Very cool idea! I think the UX needs some love. Zooming feels like too big steps and not very smooth, and for me it takes a while to load/update I like the concept a lot though - keep it up!
40.
▲
by
maalber
2y ago
Do you currently employ any measures to evaluate how well the model(s) perform? If yes, how? If no, what are the challenges that keep you from doing so?
41.
▲
by
maalber
2y ago
Thanks a lot for the reply, that does answer my questions. As a follow-up, I think I have something that could potentially help you, especially for the evaluation part. We have an API that can allow you to really quickly get feedback and p
42.
▲
by
maalber
2y ago
Have you trained/are you training your own model(s)? If so, I would be interested to hear what thoughts and challenges you had in that regard. In particular in regards to training data and evaluating whether your models improves
43.
▲
Show HN: We collected detailed annotations for text-to-image generation
(huggingface.co)
2 points
by
maalber
2y ago
|
0 comments
44.
▲
by
maalber
2y ago
This is quite cool actually. By which parameter do you determine how prominent the people are?
45.
▲
by
maalber
2y ago
The text color in the input field makes it very hard to see what you are typing. Not sure if that is intentional
46.
▲
by
maalber
2y ago
This is hilarious
47.
▲
by
maalber
2y ago
I was very excited when I saw the title. Text-to-CAD is something that I have had in mind for a while now - I think it could be a really cool use of transformers. Unfortunately, as also mentioned by dang, for a Show HN I was a bit disappoin
48.
▲
by
maalber
2y ago
Really neat idea actually! I love the landing page, both the interactive part and particularly the visual style. One thing that I would note though, is that scrolling down, I expected the images to be interactive since they blend in with
49.
▲
by
maalber
2y ago
Good idea, more focus on buy for life type products would definitely be a good change. I wonder what your criteria are for including products? Are there different categories/scales of how bfl certain products are? A little bit more i
50.
▲
by
maalber
2y ago
I agree with the others, the signup made me bounce immediately. I think it could be interesting to do as a versus game, where you could share a game link and compete with your friends.
51.
▲
by
maalber
2y ago
What is up with all these sound generators recently? That being said, nice sounds and nice layout
52.
▲
by
maalber
2y ago
Funny idea! A lot of the people I did not know, so it could be a cool feature to have the option to see a slightly longer summary either before or after the guess is submitted
53.
▲
by
maalber
2y ago
Also, is there a reason that you are highlighting child models and underwear models? To me that sends a weird vibe. However, I am also surely not your target audience
54.
▲
by
maalber
2y ago
Nice landing page! I especially like the small detail in the "Powerful Solution for Your Business" section, where the images nicely flow with the wavy line. I have a couple of questions: 1. I assume you train your own models, or a
55.
▲
by
maalber
2y ago
Cool, I really like the 'stress relief' setting! What is the intuition/settings behind each of the options?
56.
▲
by
maalber
2y ago
haha! I knew it was a perfect square but I did not know the sum thing, interesting!
57.
▲
by
maalber
2y ago
Interesting, could come in handy at some point. Can you shed some light on what goes on in the background? Is it mostly a GPT-wrapper and some prompt engineering?
58.
▲
by
maalber
2y ago
Thanks for elaborating! Please reach out once/if you start looking into further refinement (my email is in my profile). I am with a startup specializing in collecting preference data quickly and efficiently - so efficient that we even
59.
▲
by
maalber
2y ago
I love stuff like this. Thanks a lot!
60.
▲
by
maalber
2y ago
I spent way to long playing with this. Very simple and intuitive. Nice work!
More ›