Karlozkiller

14 ms·

91.
by Karlozkiller 11y ago
Soon after this got on HN, the source seems to have been taken down? I 404 both on Github and npm.
92.
by Karlozkiller 11y ago
I realize there is a problem of underdiagnosis, however there have been double blind-test studies where people who claim they get better on gluten-free diets without a diagnose did feel better if they thought they were eating gluten-free wh
93.
by Karlozkiller 11y ago
I was diagnosed 1994, and I do not live in the US. Back then I will say it was harder to find gluten-free stuff. But it progressed a lot long before the 'diet-thing' happened. Bread mix has always been around here, however more al
94.
by Karlozkiller 11y ago
Yeah, fuck people who are medically unfortunate. Or does PITA mean something other than I think here? Either way, there is a legitimate condition for which this could be a useful thing, I don't see why you should be disrespectful to pe
95.
by Karlozkiller 11y ago
As a celiac myself I welcome the new technology. However I doubt that I would use this myself, as it is often quite easy to figure out if something is gluten free, in my experience. But of course it could be beneficial in some cases. Also,
96.
by Karlozkiller 11y ago
There's actually some research on compressing ANN's to the size that it could be embedded in all sorts of devices. I think I saw something about it on HN a few months back?
97.
by Karlozkiller 11y ago
I have had a problem with using the speech_recognition library in that it does not stop listening when silence occurs. After trying to tweak the threshold parameters without success I just figured I'd add a custom key-command to break
98.
by Karlozkiller 11y ago
There might still be some substance here, but it is all written in a very sensationalistic way. Like, we already knew the brain was modular, so the local networks were a given. We also knew the brain has hierarchical networks aggregating in
99.
by Karlozkiller 11y ago
No, the most annoying thing is when they don't tell you the rules in advance, and you start typing a password that fits most standard rules. Now either you get through but don't know if you could've used a better password. Or
100.
by Karlozkiller 11y ago
Ah, thank you for pointing that out. I guess I got confused in all the papers I've read on the topic recently. It's hard to get into. However, I would still not agree that the comment-linked article explaining negative sampling re
101.
by Karlozkiller 11y ago
I found that explanation to focus almost entirely on the negative sampling, not explaining too much of the actual Skip-Gram and CBOW-models. However as I've understood it, the negative-sampling is a big part in why those models are so
102.
by Karlozkiller 11y ago
Huh, I wrote a few pages on neural networks for Natural Language Processing just a few days ago. Too bad I didn't have access to this. It seems to mention all the different kinds of networks I figured to be relevant to mention, and it
103.
by Karlozkiller 11y ago
I really don't get this to work as I feel it should. Initially I get like 4 articles showing up, sometimes two articles have the same pos-value. The list of articles then continues to be populated very slowly (I'm up to 12 now aft
104.
by Karlozkiller 12y ago
You could always run 'man ori' to see some end-user documentation. I agree that it would be nice with some documentation on the home page but man is nice if you found none.